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
32 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
33 No known attacks, no CVE, this is defensive hardening.
35 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
36 a queue-runner could start a delivery while other operations were ongoing.
37 Cutthrough delivery was a common victim, resulting in duplicate delivery.
38 Found and investigated by Tim Stewart. Fix by using the open message data
39 file handle rather than opening another, and not locally closing it (which
40 releases a lock) for that case, while creating the temporary .eml format
41 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
47 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
48 When redis_servers is set to a list of > 1 element, and the Redis servers
49 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
50 case of MOVED into a DEFER case instead, thus moving the query onto the
51 next server in the list. For a cluster of N elements, all N servers must
52 be defined in redis_servers.
54 GF/02 Catch and remove uninitialized value warning in exiqsumm
55 Check for existence of @ARGV before looking at $ARGV[0]
57 JH/01 Replace the store_release() internal interface with store_newblock(),
58 which internalises the check required to safely use the old one, plus
59 the allocate and data copy operations duplicated in both (!) of the
62 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
63 modifier. This matches the restriction on the commandline.
65 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
66 Previously only the last row was returned.
68 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
69 we assumed that tags in the header were well-formed, and parsed the
70 element content after inspecting only the first char of the tag.
71 Assumptions at that stage could crash the receive process on malformed
74 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
75 While running the DKIM ACL we operate on the Permanent memory pool so that
76 variables created with "set" persist to the DATA ACL. Also (at any time)
77 DNS lookups that fail create cache records using the Permanent pool. But
78 expansions release any allocations made on the current pool - so a dnsdb
79 lookup expansion done in the DKIM ACL releases the memory used for the
80 DNS negative-cache, and bad things result. Solution is to switch to the
81 Main pool for expansions.
82 While we're in that code, add checks on the DNS cache during store_reset,
83 active in the testsuite.
84 Problem spotted, and debugging aided, by Wolfgang Breyha.
86 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
87 When none of the hosts presented to a transport match an already-open
88 connection, close it and proceed with the list. Previously we would
89 queue the message. Spotted by Lena with Yahoo, probably involving
92 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
93 Previously a spurious "250 OK id=" response was appended to the proper
96 JH/08 The "support for" informational output now, which built with Content
97 Scanning support, has a line for the malware scanner interfaces compiled
98 in. Interface can be individually included or not at build time.
100 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
101 by the template makefile "src/EDITME". The "STREAM" support for an older
102 ClamAV interface method is removed.
104 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
105 rows affected is given instead).
107 JH/11 The runtime Berkeley DB library version is now additionally output by
108 "exim -d -bV". Previously only the compile-time version was shown.
110 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
111 SMTP connection. Previously, when one had more receipients than the
112 first, an abortive onward connection was made. Move to full support for
113 multiple onward connections in sequence, handling cutthrough connection
114 for all multi-message initiating connections.
116 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
117 routers. Previously, a multi-recipient message would fail to match the
118 onward-connection opened for the first recipient, and cause its closure.
120 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
121 a timeout on read on a GnuTLS initiating connection, resulting in the
122 initiating connection being dropped. This mattered most when the callout
123 was marked defer_ok. Fix to keep the two timeout-detection methods
126 JH/15 Relax results from ACL control request to enable cutthrough, in
127 unsupported situations, from error to silently (except under debug)
128 ignoring. This covers use with PRDR, frozen messages, queue-only and
131 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
133 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
134 metadata, resulting in a crash in free().
136 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
137 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
138 Broken also in d185889f4, with init system revamp.
140 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
141 Previously we abruptly closed the connection after reading a malware-
142 found indication; now we go on to read the "scan ok" response line,
145 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
146 ACL. Previously, a crash would result.
148 JH/19 Speed up macro lookups during configuration file read, by skipping non-
149 macro text after a replacement (previously it was only once per line) and
150 by skipping builtin macros when searching for an uppercase lead character.
152 JH/20 DANE support moved from Experimental to mainline. The Makefile control
153 for the build is renamed.
155 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
156 was allocated for every new TLS startup, meaning one per message. Fix
157 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
159 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
160 reported the original. Fix to report (as far as possible) the ACL
161 result replacing the original.
163 JH/23 Fix memory leak during multi-message connections using STARTTLS under
164 OpenSSL. Certificate information is loaded for every new TLS startup,
165 and the resources needed to be freed.
167 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
169 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
172 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
173 DATA response info to the (existing) per-recipient response info for
174 the "C=" log element. It can have useful tracking info from the
175 destination system. Patch from Simon Arlott.
177 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
178 length value. Previously this would segfault.
180 HS/02 Support Avast multiline protoocol, this allows passing flags to
181 newer versions of the scanner.
183 JH/28 Ensure that variables possibly set during message acceptance are marked
184 dead before release of memory in the daemon loop. This stops complaints
185 about them when the debug_store option is enabled. Discovered specifically
186 for sender_rate_period, but applies to a whole set of variables.
187 Do the same for the queue-runner and queue-list loops, for variables set
188 from spool message files. Do the same for the SMTP per-message loop, for
189 certain variables indirectly set in ACL operations.
191 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
192 as a multi-recipient message from a mailinglist manager). The coding had
193 an arbitrary cutoff number of characters while checking for more input;
194 enforced by writing a NUL into the buffer. This corrupted long / fast
195 input. The problem was exposed more widely when more pipelineing of SMTP
196 responses was introduced, and one Exim system was feeding another.
197 The symptom is log complaints of SMTP syntax error (NUL chars) on the
198 receiving system, and refused recipients seen by the sending system
199 (propating to people being dropped from mailing lists).
200 Discovered and pinpointed by David Carter.
202 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
203 replaced by the ${authresults } expansion.
205 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
207 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
208 allows proper process termination in container environments.
210 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
211 Previously the "final dot" had a newline after it; ensure it is CR,LF.
213 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
214 and "err_perm", deprecated since 4.83 when the RFC-defined words
215 "temperror" and "permerror" were introduced.
217 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
218 transport-filters or DKIM-signing. The restriction was lost in the
219 consolidation of verify-callout and delivery SMTP handling.
220 Extend the restriction to also cover ARC-signing.
222 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
223 in defer=pass mode supply a 450 to the initiator. Previously the message
226 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
227 tls_require_ciphers is used as before.
229 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
230 "pass_unscanned". Only tmpfails from the scanner are written to
231 the paniclog, as they may require admin intervention (permission
232 denied, license issues). Other scanner errors (like decompression
233 bombs) do not cause a paniclog entry.
235 JH/36 Fix reinitialisation of DKIM logging variable between messages.
236 Previously it was possible to log spurious information in receive log
239 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
240 triggered odd behaviour from Outlook Express clients.
242 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
245 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
246 since the IETF WG has not yet settled on that versus the original
247 "bare" representation.
249 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
250 Previously the millisecond value corrupted the output.
251 Fix also for syslog_pid=no and log_selector +pid, for which the pid
252 corrupted the output.
258 JH/01 Rework error string handling in TLS interface so that the caller in
259 more cases is responsible for logging. This permits library-sourced
260 string to be attached to addresses during delivery, and collapses
261 pairs of long lines into single ones.
263 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
264 during configuration. Wildcards are allowed and expanded.
266 JH/02 Rework error string handling in DKIM to pass more info back to callers.
267 This permits better logging.
269 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
270 do not close it down and have the child transport start it up again on
271 the passed-on TCP connection. Instead, proxy the child (and any
272 subsequent ones) for TLS via a unix-domain socket channel. Logging is
273 affected: the continued delivery log lines do not have any DNSSEC, TLS
274 Certificate or OCSP information. TLS cipher information is still logged.
276 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
277 identical IP addresses on different listening ports. Will also affect
280 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
281 add noisy ifdef guards to special-case this sillyness.
282 Patch from Bernd Kuhls.
284 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
285 than 255 are no longer allowed.
287 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
288 Disable the session-cache too, which might reduce our load. Since we
289 currrectly use a new context for every connection, both as server and
290 client, there is no benefit for these.
291 GnuTLS appears to not support tickets server-side by default (we don't
292 call gnutls_session_ticket_enable_server()) but client side is enabled
293 by default on recent versions (3.1.3 +) unless the PFS priority string
296 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
297 <https://reproducible-builds.org/specs/source-date-epoch/>.
299 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
300 the check for any unsuccessful recipients did not notice the limit, and
301 erroneously found still-pending ones.
303 JH/08 Pipeline CHUNKING command and data together, on kernels that support
304 MSG_MORE. Only in-clear (not on TLS connections).
306 JH/09 Avoid using a temporary file during transport using dkim. Unless a
307 transport-filter is involved we can buffer the headers in memory for
308 creating the signature, and read the spool data file once for the
309 signature and again for transmission.
311 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
312 4.77 as the kernel support then wasn't solid, having issues in 64bit
313 mode. Now, it's been long enough. Add support for FreeBSD also.
315 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
316 case where the routing stage had gathered several addresses to send to
317 a host before calling the transport for the first, we previously failed
318 to close down TLS in the old transport process before passing the TCP
319 connection to the new process. The new one sent a STARTTLS command
320 which naturally failed, giving a failed delivery and bloating the retry
321 database. Investigation and fix prototype from Wolfgang Breyha.
323 JH/12 Fix check on SMTP command input synchronisation. Previously there were
324 false-negatives in the check that the sender had not preempted a response
325 or prompt from Exim (running as a server), due to that code's lack of
326 awareness of the SMTP input buffering.
328 PP/04 Add commandline_checks_require_admin option.
329 Exim drops privileges sanely, various checks such as -be aren't a
330 security problem, as long as you trust local users with access to their
331 own account. When invoked by services which pass untrusted data to
332 Exim, this might be an issue. Set this option in main configuration
333 AND make fixes to the calling application, such as using `--` to stop
336 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
337 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
338 responses to those, into a single TLS record each way (this usually means
339 a single packet). As a side issue, smtp_enforce_sync now works on TLS
342 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
343 affects you only if you're dancing at the edge of the param size limits.
344 If you are, and this message makes sense to you, then: raise the
345 configured limit or use OpenSSL 1.1. Nothing we can do for older
348 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
349 cmdline element to get the documented default one. Previously it was
352 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
353 are unsuable for non-message-specific reasons - eg. connection timeout,
354 banner-time rejection.
356 JH/16 Fix logging of delivery remote port, when specified by router, under
359 PP/06 Repair manualroute's ability to take options in any order, even if one
360 is the name of a transport.
363 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
365 JH/17 Change the list-building routines interface to use the expanding-string
366 triplet model, for better allocation and copying behaviour.
368 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
369 Previously it was constructed the first time a possibly-matching string
370 was met in the configuration file input during startup; now it is done
373 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
374 compatible one, to avoid the (poorly documented) possibility of a config
375 file in the working directory redirecting the DB files, possibly correpting
376 some existing file. CVE-2017-10140 assigned for BDB.
378 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
379 cache-hot. Previously, although the result was properly cached, the
380 initial verify call returned a defer.
382 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
383 the main verify for receipient in uncached-mode.
385 JH/22 Retire historical build files to an "unsupported" subdir. These are
386 defined as "ones for which we have no current evidence of testing".
388 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
389 if present. Previously it was ignored.
391 JH/24 Start using specified-initialisers in C structure init coding. This is
392 a C99 feature (it's 2017, so now considered safe).
394 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
395 if was a fixed-sized field and bitmask ops via macros; it is now more
398 PP/07 GitHub PR 56: Apply MariaDB build fix.
399 Patch provided by Jaroslav Škarvada.
401 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
402 during Coverity cleanups [4.87 JH/47]
403 Diagnosis and fix provided by Michael Fischer v. Mollard.
405 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
406 the right size to place the terminating semicolon on its own folded
407 line, the header hash was calculated to an incorrect value thanks to
408 the (relaxed) space the fold became.
410 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
411 and confused the parent.
413 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
414 which could crash as a result. This could lead to undeliverable messages.
416 JH/28 Logging: "next input sent too soon" now shows where input was truncated
419 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
420 matters on fast-turnover and PID-randomising systems, which were getting
421 out-of-order delivery.
423 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
424 a possibly-overlapping copy. The symptom was that "Remote host closed
425 connection in response to HELO" was logged instead of the actual 4xx
428 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
429 Previously only that bufferd was discarded, resulting in SYMTP command
432 JH/32 DKIM: when a message has multiple signatures matching an identity given
433 in dkim_verify_signers, run the dkim acl once for each. Previously only
434 one run was done. Bug 2189.
436 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
437 "panic the current process" to "deliberately defer". The panic log is
438 still written with the problem list name; the mail and reject logs now
439 get a temp-reject line for the message that was being handled, saying
440 something like "domains check lookup or other defer". The SMTP 451
441 message is still "Temporary local problem".
443 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
444 A crafted sequence of BDAT commands could result in in-use memory beeing
445 freed. CVE-2017-16943.
447 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
448 from SMTP input. Previously it was always done; now only done for DATA
449 and not BDAT commands. CVE-2017-16944.
451 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
452 to the message (such as an overlong header line). Previously this was
453 not done and we did not exit BDAT mode. Followon from the previous item
454 though a different problem.
460 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
461 than -2003 did; needs libidn2 in addition to libidn.
463 JH/02 The path option on a pipe transport is now expanded before use.
465 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
466 Patch provided by "Björn", documentation fix added too.
468 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
469 missing a wire-to-host endian conversion.
471 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
472 close after a BDAT command line could be taken as a following command,
473 giving a synch failure. Fix by only checking for synch immediately
474 before acknowledging the chunk.
476 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
477 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
478 macro. Patches provided by Josh Soref.
480 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
481 Previously we did not; the RFC seems ambiguous and VRFY is not listed
482 by IANA as a service extension. However, John Klensin suggests that we
485 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
486 the dkim code may be unix-mode line endings rather than smtp wire-format
487 CRLF, so prepend a CR to any bare LF.
489 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
490 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
492 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
493 the first were themselves being wrongly included in the feed into dkim
494 processing; with most chunk sizes in use this resulted in an incorrect
495 body hash calculated value.
497 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
498 DKIM signature block, for verification. Although advised against by
499 standards it is specifically not ruled illegal.
501 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
503 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
504 missing a body hash (the bh= tag).
506 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
507 It seems that HAProxy sends the Proxy Protocol information in clear and
508 only then does a TLS startup, so do the same.
510 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
511 TCP connections (such as for Spamd) unless the daemon successfully set
512 Fast Open mode on its listening sockets. This fixes breakage seen on
513 too-old kernels or those not configured for Fast Open, at the cost of
514 requiring both directions being enabled for TFO, and TFO never being used
515 by non-daemon-related Exim processes.
517 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
518 endings, at least on the first header line. Try to canonify any that get
519 past that check, despite the cost.
521 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
522 now limited to an arbitrary five deep, while parsing addresses with the
523 strip_excess_angle_brackets option enabled.
525 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
526 instead leave the unprompted TLS handshake in socket buffer for the
527 TLS library to consume.
529 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
531 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
533 JH/16 Drop variables when they go out of scope. Memory management drops a whole
534 region in one operation, for speed, and this leaves assigned pointers
535 dangling. Add checks run only under the testsuite which checks all
536 variables at a store-reset and panics on a dangling pointer; add code
537 explicitly nulling out all the variables discovered. Fixes one known
538 bug: a transport crash, where a dangling pointer for $sending_ip_address
539 originally assigned in a verify callout, is re-used.
541 PP/06 Drop '.' from @INC in various Perl scripts.
543 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
545 PP/08 Reduce a number of compilation warnings under clang; building with
546 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
547 should be warning-free.
549 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
551 HS/01 Fix portability problems introduced by PP/08 for platforms where
552 realloc(NULL) is not equivalent to malloc() [SunOS et al].
554 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
555 chunk. This allows us to accept broken chunked messages. We need a more
556 general solution here.
558 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
559 already-broken messages in the queue.
561 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
563 JH/19 Fix reference counting bug in routing-generated-address tracking.
569 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
570 supports it and a size is available (ie. the sending peer gave us one).
572 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
573 years of being deprecated). The replacements are the ACLs
574 acl_smtp_mime and acl_not_smtp_mime.
576 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
577 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
578 or even an in-clear connection were permitted. Now, if the host lookup
579 was dnssec and dane was requested then the host is only used if the
580 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
581 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
582 if one fails this test.
583 This means that a poorly-configured remote DNS will make it incommunicado;
584 but it protects against a DNS-interception attack on it.
586 JH/04 Bug 1810: make continued-use of an open smtp transport connection
587 non-noisy when a race steals the message being considered.
589 JH/05 If main configuration option tls_certificate is unset, generate a
590 self-signed certificate for inbound TLS connections.
592 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
593 in rewrites and routers.
595 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
596 and logged a warning sing 4.83; now they are a configuration file error.
598 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
599 (lacking @domain). Apply the same qualification processing as RCPT.
601 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
603 JH/10 Support ${sha256:} applied to a string (as well as the previous
606 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
607 a cutthrough deliver is pending, as we always want to make a connection.
608 This also avoids re-routing the message when later placing the cutthrough
609 connection after a verify cache hit.
610 Do not update it with the verify result either.
612 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
613 when routing results in more than one destination address.
615 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
616 signing (which inhibits the cutthrough capability). Previously only
617 the presence of an option was tested; now an expansion evaluating as
618 empty is permissible (obviously it should depend only on data available
619 when the cutthrough connection is made).
621 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
622 the relevant preceding SMTP command did not note the pipelining mode.
624 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
625 Previously they were not counted.
627 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
628 as one having no matching records. Previously we deferred the message
629 that needed the lookup.
631 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
632 distinguished as "(=".
634 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
635 for missing MX records. Previously it only worked for missing A records.
637 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
639 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
640 after the data-go-ahead and data-ack. Patch from Jason Betts.
642 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
643 even for a "none" policy. Patch from Tony Meyer.
645 JH/22 Fix continued use of a connection for further deliveries. If a port was
646 specified by a router, it must also match for the delivery to be
649 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
650 When one of the recipients of a message was unsuitable for the connection
651 (has no matching addresses), we lost track of needing to mark it
652 deferred. As a result mail would be lost.
654 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
656 JH/25 Decoding ACL controls is now done using a binary search; the source code
657 takes up less space and should be simpler to maintain. Merge the ACL
658 condition decode tables also, with similar effect.
660 JH/26 Fix problem with one_time used on a redirect router which returned the
661 parent address unchanged. A retry would see the parent address marked as
662 delivered, so not attempt the (identical) child. As a result mail would
665 JH/27 Fix a possible security hole, wherein a process operating with the Exim
666 UID can gain a root shell. Credit to http://www.halfdog.net/ for
667 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
670 JH/28 Enable {spool,log} filesystem space and inode checks as default.
671 Main config options check_{log,spool}_{inodes,space} are now
672 100 inodes, 10MB unless set otherwise in the configuration.
674 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
675 Previously it only applied to the main-section connection policy
678 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
680 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
681 by me. Added RFC7919 DH primes as an alternative.
683 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
684 are not in the system include path.
686 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
687 GnuTLS, when a session startup failed (eg because the client disconnected)
688 Exim did stdio operations after fclose. This was exposed by a recent
689 change which nulled out the file handle after the fclose.
691 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
692 signed directly by the cert-signing cert, rather than an intermediate
693 OCSP-signing cert. This is the model used by LetsEncrypt.
695 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
697 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
698 an incoming connection.
700 HS/02 Bug 1802: Do not half-close the connection after sending a request
703 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
704 fallback to "prime256v1".
706 JH/34 SECURITY: Use proper copy of DATA command in error message.
707 Could leak key material. Remotely exploitable. CVE-2016-9963.
713 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
714 and 3.4.4 - once the server is enabled to respond to an OCSP request
715 it does even when not requested, resulting in a stapling non-aware
716 client dropping the TLS connection.
718 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
719 support variable-length bit vectors. No functional change.
721 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
722 The I= interface field on outgoing lines is now after the H= remote
723 host field, same as incoming lines. There is a separate
724 outgoing_interface log selector which allows you to disable the
727 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
728 If not running log_selector +smtp_connection the mainlog would be held
729 open indefinitely after a "too many connections" event, including to a
730 deleted file after a log rotate. Leave the per net connection logging
731 leaving it open for efficiency as that will be quickly detected by the
732 check on the next write.
734 HS/01 Bug 1671: Fix post transport crash.
735 Processing the wait-<transport> messages could crash the delivery
736 process if the message IDs didn't exist for some reason. When
737 using 'split_spool_directory=yes' the construction of the spool
738 file name failed already, exposing the same netto behaviour.
740 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
741 mime_regex ACL conditions.
743 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
744 to DSN fail messages (bounces): remote IP, remote greeting, remote response
745 to HELO, local diagnostic string.
747 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
748 log line to debug. Even when configured with a tls authenticator many
749 client connections are expected to not authenticate in this way, so
750 an authenticate fail is not an error.
752 HS/02 Add the Exim version string to the process info. This way exiwhat
753 gives some more detail about the running daemon.
755 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
756 matter for fast-change records such as DNSBLs.
758 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
759 retry record, even if constant. There may be multiple transports with
760 different interface settings and the retry behaviour needs to be kept
763 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
765 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
767 JH/10 Bug 840: fix log_defer_output option of pipe transport
769 JH/11 Bug 830: use same host for all RCPTS of a message, even under
770 hosts_randomize. This matters a lot when combined with mua_wrapper.
772 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
773 ${quote_pgsql:<string>} operator.
775 JH/13 Bug 1708: avoid misaligned access in cached lookup.
777 JH/14 Change header file name for freeradius-client. Relevant if compiling
778 with Radius support; from the Gentoo tree and checked under Fedora.
780 JH/15 Bug 1712: Introduce $prdr_requested flag variable
782 JH/16 Bug 1714: Permit an empty string as expansion result for transport
783 option transport_filter, meaning no filtering.
785 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
787 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
788 defaults to "*" (all hosts). The variable is now available when not built
789 with TLS, default unset, mainly to enable keeping the testsuite sane.
790 If a server certificate is not supplied (via tls_certificate) an error is
791 logged, and clients will find TLS connections fail on startup. Presumably
792 they will retry in-clear.
793 Packagers of Exim are strongly encouraged to create a server certificate
794 at installation time.
796 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
797 with the $config_file variable.
799 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
800 in transport context, after the attempt, and per-recipient. The latter type
801 is per host attempted. The event data is the error message, and the errno
802 information encodes the lookup type (A vs. MX) used for the (first) host,
803 and the trailing two digits of the smtp 4xx response.
805 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
806 to write to mainlog (or rejectlog, paniclog) in the window between file
807 creation and permissions/ownership being changed. Particularly affects
808 installations where exicyclog is run as root, rather than exim user;
809 result is that the running daemon panics and dies.
811 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
813 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
814 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
815 "pri" and "weight". Note that the previous implicit priority given by the
816 list order is no longer honoured.
818 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
821 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
822 by defining SUPPORT_SOCKS.
824 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
825 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
826 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
827 variables are renamed to proxy_{local,external}_{address,port}.
829 JH/27 Move Internationalisation support from Experimental to mainline, enabled
830 for a build by defining SUPPORT_I18N
832 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
833 of the query string, and make ${quote_redis:} do that quoting.
835 JH/29 Move Events support from Experimental to mainline, enabled by default
836 and removable for a build by defining DISABLE_EVENT.
838 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
840 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
841 cached by the daemon.
843 JH/32 Move Redis support from Experimental to mainline, enabled for a build
844 by defining LOOKUP_REDIS. The libhiredis library is required.
846 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
847 keys are given for lookup.
849 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
850 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
851 only supported when built with TLS support. The PolarSSL SHA routines
852 are still used when the TLS library is too old for convenient support.
854 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
855 openssl_options), for security. OpenSSL forces this from version 1.1.0
856 server-side so match that on older versions.
858 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
859 allocation for $value could be released as the expansion processing
860 concluded, but leaving the global pointer active for it.
862 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
863 and to use the domains and local_parts ACL conditions.
865 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
866 incorrectly not doubled on cutthrough transmission, hence seen as a
867 body-termination at the receiving system - resulting in truncated mails.
868 Commonly the sender saw a TCP-level error, and retransmitted the message
869 via the normal store-and-forward channel. This could result in duplicates
870 received - but deduplicating mailstores were liable to retain only the
871 initial truncated version.
873 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
875 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
877 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
878 we're in there, support oversigning also; bug 1309.
880 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
882 HS/04 Add support for keep_environment and add_environment options.
884 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
885 either intentional arithmetic overflow during PRNG, or testing config-
888 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
889 delivery resulted in actual delivery. Cancel cutthrough before DATA
892 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
893 rejects a recipient: pass the reject to the originator.
895 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
896 Many were false-positives and ignorable, but it's worth fixing the
899 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
900 for the new environment-manipulation done at startup. Move the routines
901 from being local to tls.c to being global via the os.c file.
903 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
904 an extract embedded as result-arg for a map, the first arg for extract
905 is unavailable so we cannot tell if this is a numbered or keyed
906 extraction. Accept either.
912 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
915 JH/02 The smtp transport option "multi_domain" is now expanded.
917 JH/03 The smtp transport now requests PRDR by default, if the server offers
920 JH/04 Certificate name checking on server certificates, when exim is a client,
921 is now done by default. The transport option tls_verify_cert_hostnames
922 can be used to disable this per-host. The build option
923 EXPERIMENTAL_CERTNAMES is withdrawn.
925 JH/05 The value of the tls_verify_certificates smtp transport and main options
926 default to the word "system" to access the system default CA bundle.
927 For GnuTLS, only version 3.0.20 or later.
929 JH/06 Verification of the server certificate for a TLS connection is now tried
930 (but not required) by default. The verification status is now logged by
931 default, for both outbound TLS and client-certificate supplying inbound
934 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
937 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
938 Status Notification (bounce) messages are now MIME format per RFC 3464.
939 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
940 under the control of the dsn_advertise_hosts option, and routers may
941 have a dsn_lasthop option.
943 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
944 default, modifiable by a malware= option. The list separator for
945 the options can now be changed in the usual way. Bug 68.
947 JH/10 The smtp_receive_timeout main option is now expanded before use.
949 JH/11 The incoming_interface log option now also enables logging of the
950 local interface on delivery outgoing connections.
952 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
953 if the interface and destination host and port all match.
955 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
958 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
959 Patch from Andrew Lewis.
961 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
962 now supports optional time-restrictions, weighting, and priority
963 modifiers per server. Patch originally by <rommer@active.by>.
965 JH/16 The spamd_address main option now supports a mixed list of local
966 and remote servers. Remote servers can be IPv6 addresses, and
967 specify a port-range.
969 JH/17 Bug 68: The spamd_address main option now supports an optional
970 timeout value per server.
972 JH/18 Bug 1581: Router and transport options headers_add/remove can
973 now have the list separator specified.
975 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
978 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
981 JH/21 Support for the A6 type of dns record is withdrawn.
983 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
984 rather than the verbs used.
986 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
987 from 255 to 1024 chars.
989 JH/24 Verification callouts now attempt to use TLS by default.
991 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
992 are generic router options now. The defaults didn't change.
994 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
995 Original patch from Alexander Shikoff, worked over by JH.
997 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
998 files if ZCAT_COMMAND is not executable.
1000 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1002 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1004 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1005 Normally benign, it bites when the pair was led to by a CNAME;
1006 modern usage is to not canonicalize the domain to a CNAME target
1007 (and we were inconsistent anyway for A-only vs AAAA+A).
1009 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1011 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1012 when evaluating $sender_host_dnssec.
1014 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1015 $sender_helo_dnssec variable.
1017 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1019 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1021 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1023 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1024 documented as working, but never had. Support all but $spam_report.
1026 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1027 added for tls authenticator.
1029 HS/03 Add perl_taintmode main config option
1035 TL/01 When running the test suite, the README says that variables such as
1036 no_msglog_check are global and can be placed anywhere in a specific
1037 test's script, however it was observed that placement needed to be near
1038 the beginning for it to behave that way. Changed the runtest perl
1039 script to read through the entire script once to detect and set these
1040 variables, reset to the beginning of the script, and then run through
1041 the script parsing/test process like normal.
1043 TL/02 The BSD's have an arc4random API. One of the functions to induce
1044 adding randomness was arc4random_stir(), but it has been removed in
1045 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1046 function when detected.
1048 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1049 cause callback expansion.
1051 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1052 syntax errors in an expansion can be treated as a string instead of
1053 logging or causing an error, due to the internal use of bool_lax
1054 instead of bool when processing it.
1056 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1057 server certificates when making smtp deliveries.
1059 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1061 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1063 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1065 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1066 Merged patch from Sebastian Wiedenroth.
1068 JH/05 Fix results-pipe from transport process. Several recipients, combined
1069 with certificate use, exposed issues where response data items split
1070 over buffer boundaries were not parsed properly. This eventually
1071 resulted in duplicates being sent. This issue only became common enough
1072 to notice due to the introduction of connection certificate information,
1073 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1075 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1076 size buffer was used, resulting in syntax errors when an expansion
1079 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1080 version 3.3.6 or later.
1082 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1083 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1084 both become "event_action", the variables become $event_name, $event_data
1085 and $event_defer_errno. There is a new variable $verify_mode, usable in
1086 routers, transports and related events. The tls:cert event is now also
1087 raised for inbound connections, if the main configuration event_action
1090 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1091 early OCSP support, but no stapling (appears to be less than 1.0.0).
1093 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1094 server certificate names available under the smtp transport option
1095 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1098 JH/10 Time-related extraction expansions from certificates now use the main
1099 option "timezone" setting for output formatting, and are consistent
1100 between OpenSSL and GnuTLS compilations. Bug 1541.
1102 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1103 encoded parameter in the incoming message. Bug 1558.
1105 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1106 include certificate info, eximon was claiming there were spoolfile
1109 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1111 JH/14 Log delivery-related information more consistently, using the sequence
1112 "H=<name> [<ip>]" wherever possible.
1114 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1115 are problematic for Debian distribution, omit them from the release
1118 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1120 JH/16 Fix string representation of time values on 64bit time_t architectures.
1123 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1124 output list separator was used.
1129 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1130 checkers that were complaining about end of non-void function with no
1133 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1134 This was a regression introduced in 4.83 by another bugfix.
1136 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1138 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1139 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1145 TF/01 Correctly close the server side of TLS when forking for delivery.
1147 When a message was received over SMTP with TLS, Exim failed to clear up
1148 the incoming connection properly after forking off the child process to
1149 deliver the message. In some situations the subsequent outgoing
1150 delivery connection happened to have the same fd number as the incoming
1151 connection previously had. Exim would try to use TLS and fail, logging
1152 a "Bad file descriptor" error.
1154 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1155 utilities have not been installed.
1157 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1158 temporary space as the ACL may create new global variables.
1160 TL/01 LDAP support uses per connection or global context settings, depending
1161 upon the detected version of the libraries at build time.
1163 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1164 to extract and use the src ip:port in logging and expansions as if it
1165 were a direct connection from the outside internet. PPv2 support was
1166 updated based on HAProxy spec change in May 2014.
1168 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1170 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1171 Properly escape header and check for NULL return.
1173 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1176 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1178 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1179 characters in header names, implemented as a verify condition.
1180 Contributed by Michael Fischer v. Mollard.
1182 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1183 results permerror and temperror. Previous values are deprecated but
1184 still accepted. In a future release, err_perm and err_temp will be
1185 completely removed, which will be a backward incompatibility if the
1186 ACL tests for either of these two old results. Patch contributed by
1187 user bes-internal on the mailing list.
1189 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1191 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1192 selectors, in both main and reject logs.
1194 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1197 JH/07 Add malware type "sock" for talking to simple daemon.
1199 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1201 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1202 routers/transports under cutthrough routing.
1204 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1205 numbers. Touch up "bool" conditional to keep the same definition.
1207 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1209 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1211 JH/12 Expand items in router/transport headers_add or headers_remove lists
1212 individually rather than the list as a whole. Bug 1452.
1214 Required for reasonable handling of multiple headers_ options when
1215 they may be empty; requires that headers_remove items with embedded
1216 colons must have them doubled (or the list-separator changed).
1218 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1219 view the policy declared in the DMARC record. Currently, $dmarc_status
1220 is a combined value of both the record presence and the result of the
1223 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1225 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1226 dnslookup router and the smtp transport (applying to the forward
1229 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1230 of ldap servers used for a specific lookup. Patch provided by Heiko
1233 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1234 New variable $lookup_dnssec_authenticated for observability.
1236 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1237 Patch submitted by Lars Timman.
1239 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1241 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1242 Requires trusted mode and valid format message id, aborts otherwise.
1243 Patch contributed by Heiko Schlichting.
1245 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1246 certextract with support for various fields. Bug 1358.
1248 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1249 is requested by default, modifiable by smtp transport option
1252 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1253 operate on certificate variables to give certificate fingerprints
1254 Also new ${sha256:cert_variable}.
1256 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1258 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1259 Christian Aistleitner.
1261 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1263 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1264 file. Patch from Wolfgang Breyha.
1266 JH/25 Expand the coverage of the delivery $host and $host_address to
1267 client authenticators run in verify callout. Bug 1476.
1269 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1270 align with daemon_smtp_ports. Bug 72.
1272 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1273 support and error reporting did not work properly.
1275 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1276 and is readable. Patch from Andrew Colin Kissa.
1278 TL/14 Enhance documentation of ${run expansion and how it parses the
1279 commandline after expansion, particularly in the case when an
1280 unquoted variable expansion results in an empty value.
1282 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1284 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1285 by renaming away from it. Addresses GH issue 15, reported by
1288 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1290 TL/15 SECURITY: prevent double expansion in math comparison functions
1291 (can expand unsanitized data). Not remotely exploitable.
1298 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1300 PP/02 Make -n do something, by making it not do something.
1301 When combined with -bP, the name of an option is not output.
1303 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1306 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1307 $sender_host_name and config options to manage this, and basic check
1310 PP/05 DSCP support for outbound connections and control modifier for inbound.
1312 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1313 (Only plugin which currently uses this is kerberos4, which nobody should
1314 be using, but we should make it available and other future plugins might
1315 conceivably use it, even though it would break NAT; stuff *should* be
1316 using channel bindings instead).
1318 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1319 name; added for Sendmail compatibility; requires admin caller.
1320 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1321 just ignore it); requires trusted caller.
1322 Also parse but ignore: -Ac -Am -X<logfile>
1325 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1327 TL/02 Add +smtp_confirmation as a default logging option.
1329 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1330 Patch by Magnus Holmgren from 2007-02-20.
1332 TL/04 Bugzilla 1281 - Spec typo.
1333 Bugzilla 1283 - Spec typo.
1334 Bugzilla 1290 - Spec grammar fixes.
1336 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1338 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1340 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1341 mailing list by by Dmitry Isaikin.
1343 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1345 JH/02 Support "G" suffix to numbers in ${if comparisons.
1347 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1349 NM/01 Bugzilla 1197 - Spec typo
1350 Bugzilla 1196 - Spec examples corrections
1352 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1354 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1355 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1358 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1359 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1361 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1362 "acl {{name}{arg}...}", and optional args on acl condition
1365 JH/05 Permit multiple router/transport headers_add/remove lines.
1367 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1369 JH/07 Avoid using a waiting database for a single-message-only transport.
1370 Performance patch from Paul Fisher. Bugzilla 1262.
1372 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1375 JH/09 Add $headers_added variable, with content from use of ACL modifier
1376 add_header (but not yet added to the message). Bugzilla 199.
1378 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1379 Pulled from Bugzilla 817 by Wolfgang Breyha.
1381 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1383 (nb: this is the same fix as in Exim 4.80.1)
1385 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1388 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1391 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1393 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1394 advertise SMTP AUTH mechanism to us, instead of a generic
1395 protocol violation error. Also, make Exim more robust to bad
1396 data from the Dovecot auth socket.
1398 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1400 When a queue runner is handling a message, Exim first routes the
1401 recipient addresses, during which it prunes them based on the retry
1402 hints database. After that it attempts to deliver the message to
1403 any remaining recipients. It then updates the hints database using
1406 So if a recipient address works intermittently, it can get repeatedly
1407 deferred at routing time. The retry hints record remains fresh so the
1408 address never reaches the final cutoff time.
1410 This is a fairly common occurrence when a user is bumping up against
1411 their storage quota. Exim had some logic in its local delivery code
1412 to deal with this. However it did not apply to per-recipient defers
1413 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1415 This change adds a proper retry rule check during routing so that the
1416 final cutoff time is checked against the message's age. We only do
1417 this check if there is an address retry record and there is not a
1418 domain retry record; this implies that previous attempts to handle
1419 the address had the retry_use_local_parts option turned on. We use
1420 this as an approximation for the destination being like a local
1421 delivery, as in LMTP.
1423 I suspect this new check makes the old local delivery cutoff check
1424 redundant, but I have not verified this so I left the code in place.
1426 TF/02 Correct gecos expansion when From: is a prefix of the username.
1428 Test 0254 submits a message to Exim with the header
1432 When I ran the test suite under the user fanf2, Exim expanded
1433 the header to contain my full name, whereas it should have added
1434 a Resent-Sender: header. It erroneously treats any prefix of the
1435 username as equal to the username.
1437 This change corrects that bug.
1439 GF/01 DCC debug and logging tidyup
1440 Error conditions log to paniclog rather than rejectlog.
1441 Debug lines prefixed by "DCC: " to remove any ambiguity.
1443 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1445 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1446 Bug spotted by Jeremy Harris; was flawed since initial commit.
1447 Would have resulted in OCSP responses post-SNI triggering an Exim
1448 NULL dereference and crash.
1450 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1452 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1453 Bug detection, analysis and fix by Samuel Thibault.
1454 Bugzilla 1331, Debian bug #698092.
1456 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1458 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1459 Server implementation by Todd Lyons, client by JH.
1460 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1461 config variable "prdr_enable" controls whether the server
1462 advertises the facility. If the client requests PRDR a new
1463 acl_data_smtp_prdr ACL is called once for each recipient, after
1464 the body content is received and before the acl_smtp_data ACL.
1465 The client is controlled by both of: a hosts_try_prdr option
1466 on the smtp transport, and the server advertisement.
1467 Default client logging of deliveries and rejections involving
1468 PRDR are flagged with the string "PRDR".
1470 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1471 fclose(). Diagnosis by Todd Lyons.
1473 PP/17 Update configure.default to handle IPv6 localhost better.
1474 Patch by Alain Williams (plus minor tweaks).
1477 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1478 This is now consistent with GnuTLS, and is now documented: the
1479 previous undocumented portable approach to treating the option as
1480 unset was to force an expansion failure. That still works, and
1481 an empty string is now equivalent.
1483 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1484 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1485 not performing validation itself.
1487 PP/20 Added force_command boolean option to pipe transport.
1488 Patch from Nick Koston, of cPanel Inc.
1490 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1493 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1495 PP/21 Fix eximon continuous updating with timestamped log-files.
1496 Broken in a format-string cleanup in 4.80, missed when I repaired the
1497 other false fix of the same issue.
1498 Report and fix from Heiko Schlichting.
1501 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1502 Report from Prashanth Katuri.
1504 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1505 It's SecureTransport, so affects any MacOS clients which use the
1506 system-integrated TLS libraries, including email clients.
1508 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1509 using a MIME ACL for non-SMTP local injection.
1510 Report and assistance in diagnosis by Warren Baker.
1512 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1514 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1516 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1517 last id that failed so it may be referenced in subsequent ACL's.
1519 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1522 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1523 ldap library initialization, allowing self-signed CA's to be
1524 used. Also properly sets require_cert option later in code by
1525 using NULL (global ldap config) instead of ldap handle (per
1526 session). Bug diagnosis and testing by alxgomz.
1528 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1529 the src/util/ subdirectory.
1531 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1532 renamed to Transport Post Delivery Action by Jeremy Harris, as
1535 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1536 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1537 redis_servers = needs to be configured which will be used by the redis
1538 lookup. Patch from Warren Baker, of The Packet Hub.
1540 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1542 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1543 hostname or reverse DNS when processing a host list. Used suggestions
1544 from multiple comments on this bug.
1546 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1548 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1549 Missed a few lines, added it to make the runtest require no keyboard
1552 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1553 contains upper case chars. Make router use caseful_local_part.
1555 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1556 support when GnuTLS has been built with p11-kit.
1562 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1564 This, or similar/improved, will also be change PP/11 of 4.82.
1570 PP/01 Handle short writes when writing local log-files.
1571 In practice, only affects FreeBSD (8 onwards).
1572 Bugzilla 1053, with thanks to Dmitry Isaikin.
1574 NM/01 Bugzilla 949 - Documentation tweak
1576 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1579 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1581 PP/02 Implemented gsasl authenticator.
1583 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1585 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1586 `pkg-config foo` for cflags/libs.
1588 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1589 with rest of GSASL and with heimdal_gssapi.
1591 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1592 `pkg-config foo` for cflags/libs for the TLS implementation.
1594 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1595 properties get this fed in as external SSF. A number of robustness
1596 and debugging improvements to the cyrus_sasl authenticator.
1598 PP/08 cyrus_sasl server now expands the server_realm option.
1600 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1601 Patch by Jeremy Harris.
1603 PP/10 Added dbmjz lookup type.
1605 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1607 PP/12 MAIL args handles TAB as well as SP, for better interop with
1608 non-compliant senders.
1609 Analysis and variant patch by Todd Lyons.
1611 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1612 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1613 Patch from Dirk Mueller <dmueller@suse.com>
1615 PP/13 tls_peerdn now print-escaped for spool files.
1616 Observed some $tls_peerdn in wild which contained \n, which resulted
1617 in spool file corruption.
1619 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1620 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1621 or write after TLS renegotiation, which otherwise led to messages
1624 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1625 as a tracking header (ie: a signed header comes before the signature).
1626 Patch from Wolfgang Breyha.
1628 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1629 comma-sep list; embedded commas doubled.
1631 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1633 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1635 Report and patch from Dmitry Banschikov.
1637 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1638 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1639 failures. We appear to now support TLS1.1+ with Exim.
1641 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1642 lets Exim select keys and certificates based upon TLS SNI from client.
1643 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1644 before an outbound SMTP session. New log_selector, +tls_sni.
1646 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1647 NULL dereference. Report and patch from Alun Jones.
1649 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1650 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1651 Not seeing resolver debug output on NetBSD, but suspect this is a
1652 resolver implementation change.
1654 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1655 Left warnings. Added "eximon gdb" invocation mode.
1657 PP/21 Defaulting "accept_8bitmime" to true, not false.
1659 PP/22 Added -bw for inetd wait mode support.
1661 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1662 locate the relevant includes and libraries. Made this the default.
1664 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1665 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1667 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1668 This may cause build issues on older platforms.
1670 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1671 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1672 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1673 Added SNI support via GnuTLS too.
1674 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1676 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1678 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1679 Applied second patch from Janne, implementing suggestion to default
1680 multiple-strings-in-record handling to match SPF spec.
1682 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1684 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1685 read-only, out of scope).
1686 Patch from Wolfgang Breyha, report from Stuart Northfield.
1688 PP/29 Fix three issues highlighted by clang analyser static analysis.
1689 Only crash-plausible issue would require the Cambridge-specific
1690 iplookup router and a misconfiguration.
1691 Report from Marcin Mirosław.
1693 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1695 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1696 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1697 As part of this, removing so much warning spew let me fix some minor
1698 real issues in debug logging.
1700 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1701 assignment on my part. Fixed.
1703 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1704 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1705 Janne Snabb (who went above and beyond: thank you).
1707 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1708 string otherwise requires a connection and a bunch more work and it's
1709 relatively easy to get wrong. Should also expose TLS library linkage
1712 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1713 64-bit ${eval} (JH/03).
1715 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1716 GNU libc to support some of the 64-bit stuff, should not lead to
1717 conflicts. Defined before os.h is pulled in, so if a given platform
1718 needs to override this, it can.
1720 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1721 protection layer was required, which is not implemented.
1722 Bugzilla 1254, patch from Wolfgang Breyha.
1724 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1725 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1726 tls_dhparam take prime identifiers. Also unbreak combination of
1727 OpenSSL+DH_params+TLSSNI.
1729 PP/39 Disable SSLv2 by default in OpenSSL support.
1735 PP/01 Solaris build fix for Oracle's LDAP libraries.
1736 Bugzilla 1109, patch from Stephen Usher.
1738 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1740 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1743 TF/02 Fix a couple more cases where we did not log the error message
1744 when unlink() failed. See also change 4.74-TF/03.
1746 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1747 lock up or crash if it happened to be inside a call to libc when it
1748 got a SIGUSR1 from exiwhat.
1750 The SIGUSR1 handler appends the current process status to the process
1751 log which is later printed by exiwhat. It used to use the general
1752 purpose logging code to do this, but several functions it calls are
1753 not safe for signals.
1755 The new output code in the SIGUSR1 handler is specific to the process
1756 log, and simple enough that it's easy to inspect for signal safety.
1757 Removing some special cases also simplifies the general logging code.
1758 Removing the spurious timestamps from the process log simplifies
1761 TF/04 Improved ratelimit ACL condition.
1763 The /noupdate option has been deprecated in favour of /readonly which
1764 has clearer semantics. The /leaky, /strict, and /readonly update modes
1765 are mutually exclusive. The update mode is no longer included in the
1766 database key; it just determines when the database is updated. (This
1767 means that when you upgrade Exim will forget old rate measurements.)
1769 Exim now checks that the per_* options are used with an update mode that
1770 makes sense for the current ACL. For example, when Exim is processing a
1771 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1772 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1773 must specify per_mail/readonly. If you omit the update mode it defaults to
1774 /leaky where that makes sense (as before) or /readonly where required.
1776 The /noupdate option is now undocumented but still supported for
1777 backwards compatibility. It is equivalent to /readonly except that in
1778 ACLs where /readonly is required you may specify /leaky/noupdate or
1779 /strict/noupdate which are treated the same as /readonly.
1781 A useful new feature is the /count= option. This is a generalization
1782 of the per_byte option, so that you can measure the throughput of other
1783 aggregate values. For example, the per_byte option is now equivalent
1784 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1786 The per_rcpt option has been generalized using the /count= mechanism
1787 (though it's more complicated than the per_byte equivalence). When it is
1788 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1789 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1790 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1791 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1792 Note that using per_rcpt with a non-readonly update mode in more than
1793 one ACL will cause the recipients to be double-counted. (The per_mail
1794 and per_byte options don't have this problem.)
1796 The handling of very low rates has changed slightly. If the computed rate
1797 is less than the event's count (usually one) then this event is the first
1798 after a long gap. In this case the rate is set to the same as this event's
1799 count, so that the first message of a spam run is counted properly.
1801 The major new feature is a mechanism for counting the rate of unique
1802 events. The new per_addr option counts the number of different
1803 recipients that someone has sent messages to in the last time period. It
1804 behaves like per_rcpt if all the recipient addresses are different, but
1805 duplicate recipient addresses do not increase the measured rate. Like
1806 the /count= option this is a general mechanism, so the per_addr option
1807 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1808 example, measure the rate that a client uses different sender addresses
1809 with the options per_mail/unique=$sender_address. There are further
1810 details in the main documentation.
1812 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1814 TF/06 Removed a few PCRE remnants.
1816 TF/07 Automatically extract Exim's version number from tags in the git
1817 repository when doing development or release builds.
1819 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1820 Bugzilla 879. Patch from Paul Fisher.
1822 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1823 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1826 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1828 PP/05 Variable $av_failed, true if the AV scanner deferred.
1829 Bugzilla 1078. Patch from John Horne.
1831 PP/06 Stop make process more reliably on build failure.
1832 Bugzilla 1087. Patch from Heiko Schlittermann.
1834 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1835 Bugzilla 1089. Patch from Heiko Schlittermann.
1837 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1838 Bugzilla 1131. Patch from Holger Weiß.
1840 PP/09 Handle IPv6 addresses with SPF.
1841 Bugzilla 860. Patch from Wolfgang Breyha.
1843 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1845 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1848 PP/11 match_* no longer expand right-hand-side by default.
1849 New compile-time build option, EXPAND_LISTMATCH_RHS.
1850 New expansion conditions, "inlist", "inlisti".
1852 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1854 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1856 PP/14 fix log_write() format string regression from TF/03.
1857 Bugzilla 1152. Patch from Dmitry Isaikin.
1863 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1865 PP/02 Harmonised TLS library version reporting; only show if debugging.
1866 Layout now matches that introduced for other libraries in 4.74 PP/03.
1868 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1870 PP/04 New "dns_use_edns0" global option.
1872 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1875 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1876 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1878 TK/01 Updated PolarSSL code to 0.14.2.
1879 Bugzilla 1097. Patch from Andreas Metzler.
1881 PP/07 Catch divide-by-zero in ${eval:...}.
1882 Fixes bugzilla 1102.
1884 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1887 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1888 format-string attack -- SECURITY: remote arbitrary code execution.
1890 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1891 time unintentionally subject to list matching rules, letting the header
1892 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1893 arbitrary string expansion). This allowed for information disclosure.
1895 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1896 INT_MIN/-1 -- value coerced to INT_MAX.
1902 NM/01 Workaround for PCRE version dependency in version reporting
1905 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1906 This fixes portability to compilers other than gcc, notably
1907 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1909 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1910 makefiles for portability to HP-UX and POSIX correctness.
1912 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1913 Also via indented variable definition in the Makefile.
1914 (Debugging by Oliver Heesakkers).
1916 PP/02 Restore caching of spamd results with expanded spamd_address.
1917 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1919 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1920 Improves build reliability. Fix from: Frank Elsner
1922 NM/02 Fix wide character breakage in the rfc2047 coding
1923 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1925 NM/03 Allow underscore in dnslist lookups
1926 Fixes bug 1026. Patch from Graeme Fowler
1928 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1929 Code patches from Adam Ciarcinski of NetBSD.
1931 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1934 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1935 is logged, to avoid truncation. Patch from John Horne.
1937 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1938 Patch from Jakob Hirsch.
1940 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1941 SQL string expansion failure details.
1942 Patch from Andrey Oktyabrski.
1944 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1945 Patch from Simon Arlott.
1947 PP/09 New lookups functionality failed to compile on old gcc which rejects
1948 extern declarations in function scope.
1949 Patch from Oliver Fleischmann
1951 PP/10 Use sig_atomic_t for flags set from signal handlers.
1952 Check getgroups() return and improve debugging.
1953 Fixed developed for diagnosis in bug 927 (which turned out to be
1956 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1957 Patch from Mark Zealey.
1959 PP/12 Bugzilla 1056: Improved spamd server selection.
1960 Patch from Mark Zealey.
1962 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1963 Based on patch from Heiko Schlittermann.
1965 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1966 Patch from Uwe Doering, sign-off by Michael Haardt.
1968 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1969 variable declaration deep within a block. Bug and patch from
1972 PP/15 lookups-Makefile IRIX compatibility coercion.
1974 PP/16 Make DISABLE_DKIM build knob functional.
1976 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1977 Patch by Simon Arlott
1979 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1980 variable argument macros. Our copy now differs from upstream.
1986 TF/01 Failure to get a lock on a hints database can have serious
1987 consequences so log it to the panic log.
1989 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1990 controlled using the smtp_confirmation log selector.
1992 TF/03 Include the error message when we fail to unlink a spool file.
1994 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1995 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1996 for maintaining out-of-tree patches for some time.
1998 PP/01 Bugzilla 139: Documentation and portability issues.
1999 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2000 Handle per-OS dynamic-module compilation flags.
2002 PP/02 Let /dev/null have normal permissions.
2003 The 4.73 fixes were a little too stringent and complained about the
2004 permissions on /dev/null. Exempt it from some checks.
2005 Reported by Andreas M. Kirchwitz.
2007 PP/03 Report version information for many libraries, including
2008 Exim version information for dynamically loaded libraries. Created
2009 version.h, now support a version extension string for distributors
2010 who patch heavily. Dynamic module ABI change.
2012 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2013 privilege escalation vulnerability whereby the Exim run-time user
2014 can cause root to append content of the attacker's choosing to
2017 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2020 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2021 If dropping privileges for untrusted macros, we disabled normal logging
2022 on the basis that it would fail; for the Exim run-time user, this is not
2023 the case, and it resulted in successful deliveries going unlogged.
2024 Fixed. Reported by Andreas Metzler.
2030 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2031 only prepend for the Resent-* case. Fixes regression introduced in
2032 Exim 4.70 by NM/22 for Bugzilla 607.
2034 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2035 increasing numbers of administrators be bitten by this.
2037 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2039 PP/03 Bugzilla 994: added openssl_options main configuration option.
2041 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2043 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2045 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2047 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2048 without a peer certificate, leading to a segfault because of an
2049 assumption that peers always have certificates. Be a little more
2050 paranoid. Problem reported by Martin Tscholak.
2052 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2053 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2054 NB: ClamAV planning to remove STREAM in "middle of 2010".
2055 CL also introduces -bmalware, various -d+acl logging additions and
2056 more caution in buffer sizes.
2058 PP/09 Implemented reverse_ip expansion operator.
2060 PP/10 Bugzilla 937: provide a "debug" ACL control.
2062 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2064 PP/12 Bugzilla 973: Implement --version.
2066 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2068 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2070 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2072 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2073 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2074 ignore trailing whitespace.
2076 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2078 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2081 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2082 Notification from Dr Andrew Aitchison.
2084 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2085 ExtendedDetectionInfo response format.
2086 Notification from John Horne.
2088 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2091 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2092 XSL and documented dependency on system catalogs, with examples of how
2095 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2098 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2099 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2100 configuration file which is writeable by the Exim user or group.
2102 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2103 of configuration files to cover files specified with the -C option if
2104 they are going to be used with root privileges, not just the default
2107 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2108 option (effectively making it always true).
2110 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2111 files to be used while preserving root privileges.
2113 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2114 that rogue child processes cannot use them.
2116 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2117 run-time user, instead of root.
2119 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2120 Exim run-time user without dropping privileges.
2122 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2123 result string, instead of calling string_vformat() twice with the same
2126 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2127 for other users. Others should always drop root privileges if they use
2128 -C on the command line, even for a whitelisted configure file.
2130 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2132 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2138 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2139 $data_path, and $header_path variables; fixed documentation bugs and
2142 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2143 exipick to access non-standard spools, including the "frozen" queue
2146 NM/01 Bugzilla 965: Support mysql stored procedures.
2147 Patch from Alain Williams
2149 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2151 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2152 Patch from Andreas Metzler
2154 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2155 Patch from Kirill Miazine
2157 NM/05 Bugzilla 671: Added umask to procmail example.
2159 JJ/03 installed exipick 20100323.0, fixing doc bug
2161 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2162 directory. Notification and patch from Dan Rosenberg.
2164 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2166 TK/02 Improve log output when DKIM signing operation fails.
2168 MH/01 Treat the transport option dkim_domain as a colon separated
2169 list, not as a single string, and sign the message with each element,
2170 omitting multiple occurences of the same signer.
2172 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2173 Bugzilla 985, 986. Patch by Simon Arlott
2175 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2176 Patch by Simon Arlott
2178 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2179 MBX locking. Notification from Dan Rosenberg.
2185 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2187 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2189 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2191 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2193 NM/04 Bugzilla 915: Flag broken perl installation during build.
2199 TK/01 Added patch by Johannes Berg that expands the main option
2200 "spamd_address" if it starts with a dollar sign.
2202 TK/02 Write list of recipients to X-Envelope-Sender header when building
2203 the mbox-format spool file for content scanning (suggested by Jakob
2206 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2207 (http://www.dcc-servers.net/) support via dccifd. Activated by
2208 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2210 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2211 by Mark Daniel Reidel <mr@df.eu>.
2213 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2214 When building exim an external PCRE library is now needed -
2215 PCRE is a system library on the majority of modern systems.
2216 See entry on PCRE_LIBS in EDITME file.
2218 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2219 conversation. Added nologin parameter to request.
2220 Patch contributed by Kirill Miazine.
2222 TF/01 Do not log submission mode rewrites if they do not change the address.
2224 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2226 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2227 log files in place. Contributed by Roberto Lima.
2229 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2231 TF/03 Bugzilla 615: When checking the local_parts router precondition
2232 after a local_part_suffix or local_part_prefix option, Exim now
2233 does not use the address's named list lookup cache, since this
2234 contains cached lookups for the whole local part.
2236 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2237 Robert Millan. Documentation is in experimental-spec.txt.
2239 TF/04 Bugzilla 668: Fix parallel build (make -j).
2241 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2243 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2244 Patch provided by Jan Srzednicki.
2246 TF/05 Leading white space used to be stripped from $spam_report which
2247 wrecked the formatting. Now it is preserved.
2249 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2250 that they are available at delivery time.
2252 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2254 TF/08 TLS error reporting now respects the incoming_interface and
2255 incoming_port log selectors.
2257 TF/09 Produce a more useful error message if an SMTP transport's hosts
2258 setting expands to an empty string.
2260 NM/06 Bugzilla 744: EXPN did not work under TLS.
2261 Patch provided by Phil Pennock.
2263 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2264 Patch provided by Richard Godbee.
2266 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2267 acl_smtp_notquit, added index entry.
2269 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2270 Patch provided by Eugene Bujak.
2272 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2273 accept(). Patch provided by Maxim Dounin.
2275 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2276 Patch provided by Phil Pennock.
2278 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2280 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2281 Patch provided by Brad "anomie" Jorsch.
2283 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2284 Patch provided by Dean Brooks.
2286 TK/05 Add native DKIM support (does not depend on external libraries).
2288 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2289 Patch provided by Graeme Fowler.
2291 NM/16 Bugzilla 851: Documentation example syntax fix.
2293 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2295 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2298 NM/19 Bugzilla 745: TLS version reporting.
2299 Patch provided by Phil Pennock.
2301 NM/20 Bugzilla 167: bool: condition support.
2302 Patch provided by Phil Pennock.
2304 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2305 clients. Patch provided by Phil Pennock.
2307 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2308 Patch provided by Brad "anomie" Jorsch.
2310 NM/23 Bugzilla 687: Fix misparses in eximstats.
2311 Patch provided by Heiko Schlittermann.
2313 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2314 Patch provided by Heiko Schlittermann.
2316 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2317 plus update to original patch.
2319 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2321 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2322 Patch provided by David Brownlee.
2324 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2326 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2328 NM/30 Bugzilla 888: TLS documentation bugfixes.
2330 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2332 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2333 Unlike the original bugzilla I have changed all shell scripts in src tree.
2335 NM/33 Bugzilla 898: Transport filter timeout fix.
2336 Patch by Todd Rinaldo.
2338 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2339 Patch by Serge Demonchaux.
2341 NM/35 Bugzilla 39: Base64 decode bug fixes.
2342 Patch by Jakob Hirsch.
2344 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2346 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2348 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2350 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2356 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2357 ALT-N's libdkim that I have put here:
2358 http://duncanthrax.net/exim-experimental/
2360 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2361 were called 'true' and it seems that C99 defines that as a reserved
2362 keyword to be used with 'bool' variable types. That means you could
2363 not include C99-style headers which use bools without triggering
2364 build errors in sieve.c.
2366 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2367 as mailq or other aliases. Changed the --help handling significantly
2368 to do whats expected. exim_usage() emits usage/help information.
2370 SC/01 Added the -bylocaldomain option to eximstats.
2372 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2374 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2376 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2382 PH/01 Another patch from the Sieve maintainer.
2384 PH/02 When an IPv6 address is converted to a string for single-key lookup
2385 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2386 dots are used instead of colons so that keys in lsearch files need not
2387 contain colons. This was done some time before quoting was made available
2388 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2389 (notated using the quote facility) so as to distinguish them from IPv4
2390 keys. This meant that lookups for IP addresses in host lists did not work
2391 for iplsearch lookups.
2393 This has been fixed by arranging for IPv6 addresses to be expressed with
2394 colons if the lookup type is iplsearch. This is not incompatible, because
2395 previously such lookups could never work.
2397 The situation is now rather anomalous, since one *can* have colons in
2398 ordinary lsearch keys. However, making the change in all cases is
2399 incompatible and would probably break a number of configurations.
2401 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2404 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2405 conversion specification without a maximum field width, thereby enabling
2406 a rogue spamd server to cause a buffer overflow. While nobody in their
2407 right mind would setup Exim to query an untrusted spamd server, an
2408 attacker that gains access to a server running spamd could potentially
2409 exploit this vulnerability to run arbitrary code as the Exim user.
2411 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2412 $primary_hostname instead of what libspf2 thinks the hosts name is.
2414 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2415 a directory entry by the name of the lookup key. Previously, if a
2416 symlink pointed to a non-existing file or a file in a directory that
2417 Exim lacked permissions to read, a lookup for a key matching that
2418 symlink would fail. Now it is enough that a matching directory entry
2419 exists, symlink or not. (Bugzilla 503.)
2421 PH/03 The body_linecount and body_zerocount variables are now exported in the
2424 PH/04 Added the $dnslist_matched variable.
2426 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2427 This means they are set thereafter only if the connection becomes
2430 PH/06 Added the client_condition to authenticators so that some can be skipped
2431 by clients under certain conditions.
2433 PH/07 The error message for a badly-placed control=no_multiline_responses left
2434 "_responses" off the end of the name.
2436 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2438 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2439 (without spaces) instead of just copying the configuration text.
2441 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2443 PH/11 Added $max_received_linelength.
2445 PH/12 Added +ignore_defer and +include_defer to host lists.
2447 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2448 way in which PCRE > 7.0 is built.
2450 PH/14 Implemented queue_only_load_latch.
2452 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2453 MAIL command. The effect was to mangle the value on 64-bit systems.
2455 PH/16 Another patch from the Sieve maintainer.
2457 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2459 PH/18 If a system quota error occurred while trying to create the file for
2460 a maildir delivery, the message "Mailbox is full" was not appended to the
2461 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2462 only to a quota excession during the actual writing of the file.
2464 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2465 characters?) which causes problems in log lines. The DN values are now
2466 passed through string_printing() before being added to log lines.
2468 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2469 and InterBase are left for another time.)
2471 PH/21 Added message_body_newlines option.
2473 PH/22 Guard against possible overflow in moan_check_errorcopy().
2475 PH/23 POSIX allows open() to be a macro; guard against that.
2477 PH/24 If the recipient of an error message contained an @ in the local part
2478 (suitably quoted, of course), incorrect values were put in $domain and
2479 $local_part during the evaluation of errors_copy.
2485 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2486 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2489 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2490 issue a MAIL command.
2492 PH/02 In an ACL statement such as
2494 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2496 if a client was not listed at all, or was listed with a value other than
2497 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2498 the condition was not true (as it should be), so access was not denied.
2499 The bug was that the ! inversion was incorrectly passed on to the second
2500 item. This has been fixed.
2502 PH/03 Added additional dnslists conditions == and =& which are different from
2503 = and & when the dns lookup returns more than one IP address.
2505 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2506 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2508 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2509 FSYNC, which compiles an option called disable_fsync that allows for
2510 bypassing fsync(). The documentation is heavily laced with warnings.
2512 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2514 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2515 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2516 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2517 including adding "make clean"; (3) Added -fPIC when compiling the test
2518 dynamically loaded module, to get rid of a warning.
2520 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2521 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2522 = 0s. The bug is otherwise harmless.
2524 PH/07 There was a bug in the dovecot authenticator such that the value of
2525 $auth1 could be overwritten, and so not correctly preserved, after a
2526 successful authentication. This usually meant that the value preserved by
2527 the server_setid option was incorrect.
2529 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2531 PH/09 Installed PCRE release 7.0.
2533 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2534 run for batched SMTP input. It is now run at the start of every message
2535 in the batch. While fixing this I discovered that the process information
2536 (output by running exiwhat) was not always getting set for -bs and -bS
2537 input. This is fixed, and it now also says "batched" for BSMTP.
2539 PH/11 Added control=no_pipelining.
2541 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2542 patch, slightly modified), and move the expansion of helo_data till after
2543 the connection is made in the smtp transport (so it can use these
2546 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2548 PH/14 Added log_selector = +pid.
2550 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2552 PH/16 Add ${if forany and ${if forall.
2554 PH/17 Added dsn_from option to vary the From: line in DSNs.
2556 PH/18 Flush SMTP output before performing a callout, unless control =
2557 no_callout_flush is set.
2559 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2560 was true (the default) a successful delivery failed to delete the retry
2561 item, thus causing premature timeout of the address. The bug is now
2564 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2566 PH/21 Long custom messages for fakedefer and fakereject are now split up
2567 into multiline responses in the same way that messages for "deny" and
2568 other ACL rejections are.
2570 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2571 with slight modification.
2573 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2574 draft, changing the syntax and factoring some duplicate code".
2576 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2577 for deliveries of the second and subsequent messages over the same SMTP
2580 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2581 ${reduce, with only minor "tidies".
2583 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2585 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2586 expansion side effects.
2588 PH/27 When a message times out after an over-quota error from an Exim-imposed
2589 quota, the bounce message says "mailbox is full". This message was not
2590 being given when it was a system quota that was exceeded. It now should
2593 MH/03 Made $recipients available in local_scan(). local_scan() already has
2594 better access to the recipient list through recipients_list[], but
2595 $recipients can be useful in postmaster-provided expansion strings.
2597 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2598 in the case of a MAIL command with additional options following the
2599 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2600 were accidentally chopped off.
2602 PH/29 SMTP synchronization checks are implemented when a command is read -
2603 there is a check that no more input is waiting when there shouldn't be
2604 any. However, for some commands, a delay in an ACL can mean that it is
2605 some time before the response is written. In this time, more input might
2606 arrive, invalidly. So now there are extra checks after an ACL has run for
2607 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2608 pipelining has not been advertised.
2610 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2612 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2613 returned in the SMTP response when smtp_return_error_details was false.
2614 This has been fixed.
2616 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2617 instead of the C I/O that was originally supplied, because problems were
2618 reported on Solaris.
2620 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2621 Exim which did not show up earlier: it was assuming that a call to
2622 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2623 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2624 was a macro that became an assignment, so it seemed to work. This has
2625 changed to a proper function call with a void return, hence the compile
2626 error. Exim's code has been fixed.
2628 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2631 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2633 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2635 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2636 because it turns out to be too big or there is a timeout) there is no
2637 "Completed" line in the log. When some messages of this type were
2638 selected by exigrep, they were listed as "not completed". Others were
2639 picked up by some special patterns. I have improved the selection
2640 criteria to be more general.
2642 PH/38 The host_find_failed option in the manualroute router can now be set
2643 to "ignore", to completely ignore a host whose IP address cannot be
2644 found. If all hosts are ignored, the behaviour is controlled by the new
2645 host_all_ignored option.
2647 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2648 homing or because of multiple MX records with /mx) generated more than
2649 one IP address, and the following item turned out to be the local host,
2650 all the secondary addresses of the first item were incorrectly removed
2651 from the list, along with the local host and any following hosts (which
2652 is what is supposed to happen).
2654 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2655 whoever called Exim into the -H file. In the case of the daemon it was
2656 behaving confusingly. When first started, it used values for whoever
2657 started the daemon, but after a SIGHUP it used the Exim user (because it
2658 calls itself on a restart). I have changed the code so that it now always
2661 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2662 message are rejected with the same error (e.g. no authentication or bad
2663 sender address), and a DATA command is nevertheless sent (as can happen
2664 with PIPELINING or a stupid MUA), the error message that was given to the
2665 RCPT commands is included in the rejection of the DATA command. This is
2666 intended to be helpful for MUAs that show only the final error to their
2669 PH/42 Another patch from the Sieve maintainer.
2671 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2672 Eximstats - Fixed some broken HTML links and added missing column headers
2674 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2675 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2677 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2679 PH/43 Yet another patch from the Sieve maintainer.
2681 PH/44 I found a way to check for a TCP/IP connection going away before sending
2682 the response to the final '.' that terminates a message, but only in the
2683 case where the client has not sent further data following the '.'
2684 (unfortunately, this is allowed). However, in many cases there won't be
2685 any further data because there won't be any more messages to send. A call
2686 to select() can be used: if it shows that the input is "ready", there is
2687 either input waiting, or the socket has been closed. An attempt to read
2688 the next input character can distinguish the two cases. Previously, Exim
2689 would have sent an OK response which the client would never have see.
2690 This could lead to message repetition. This fix should cure that, at
2691 least in a lot of common cases.
2693 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2694 advertised in response to EHLO.
2700 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2701 fixed by 4.65/MH/01 (is this a record?) are fixed:
2703 (i) An empty string was always treated as zero by the numeric comparison
2704 operators. This behaviour has been restored.
2706 (ii) It is documented that the numeric comparison operators always treat
2707 their arguments as decimal numbers. This was broken in that numbers
2708 starting with 0 were being interpreted as octal.
2710 While fixing these problems I realized that there was another issue that
2711 hadn't been noticed. Values of message_size_limit (both the global option
2712 and the transport option) were treated as octal if they started with 0.
2713 The documentation was vague. These values are now always treated as
2714 decimal, and I will make that clear in the documentation.
2720 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2721 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2724 MH/01 Don't check that the operands of numeric comparison operators are
2725 integers when their expansion is in "skipping" mode (fixes bug
2726 introduced by 4.64-PH/07).
2728 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2729 child addresses, Exim now panics and dies. Previously, because the count
2730 is held in a short int, deliveries were likely to be lost. As such a
2731 large number of recipients for a single message is ridiculous
2732 (performance will be very, very poor), I have chosen to impose a limit
2733 rather than extend the field.
2739 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2740 leftover -K file (the existence of which was triggered by #402).
2741 While we were at it, introduced process PID as part of the -K
2742 filename. This should rule out race conditions when creating
2745 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2746 processing considerably. Previous code took too long for large mails,
2747 triggering a timeout which in turn triggers #401.
2749 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2750 in the DK code in transports.c. sendfile() is not really portable,
2751 hence the _LINUX specificness.
2753 TF/01 In the add_headers option to the mail command in an Exim filter,
2754 there was a bug that Exim would claim a syntax error in any
2755 header after the first one which had an odd number of characters
2758 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2759 callout verification, Exim cached a "reject" for the entire domain. This
2760 is correct for most verifications, but it is not correct for a recipient
2761 verification with use_sender or use_postmaster set, because in that case
2762 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2763 case of MAIL FROM:<> rejection from other early rejections (e.g.
2764 rejection of HELO). When verifying a recipient using a non-null MAIL
2765 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2766 Whatever the result of the callout, the value of the domain cache is
2767 left unchanged (for any other kind of callout, getting as far as trying
2768 RCPT means that the domain itself is ok).
2770 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2773 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2774 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2775 session, the daemon ignores ECONNECT errors and logs others; it now
2776 ignores EPIPE as well.
2778 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2779 (quoted-printable decoding).
2781 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2782 later the small subsequent patch to fix an introduced bug.
2784 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2786 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2788 PH/08 An error is now given if message_size_limit is specified negative.
2790 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2791 to be given (somewhat) arbitrary names.
2793 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2796 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2797 miscellaneous code fixes
2799 PH/10 Added the log_reject_target ACL modifier to specify where to log
2802 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2803 hostname. This is wrong, because it relates to the incoming message (and
2804 probably the interface on which it is arriving) and not to the outgoing
2805 callout (which could be using a different interface). This has been
2806 changed to use the value of the helo_data option from the smtp transport
2807 instead - this is what is used when a message is actually being sent. If
2808 there is no remote transport (possible with a router that sets up host
2809 addresses), $smtp_active_hostname is used.
2811 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2812 tweaks were necessary in order to get it to work (see also 21 below):
2813 (a) The code assumed that strncpy() returns a negative number on buffer
2814 overflow, which isn't the case. Replaced with Exim's string_format()
2816 (b) There were several signed/unsigned issues. I just did the minimum
2817 hacking in of casts. There is scope for a larger refactoring.
2818 (c) The code used strcasecmp() which is not a standard C function.
2819 Replaced with Exim's strcmpic() function.
2820 (d) The code set only $1; it now sets $auth1 as well.
2821 (e) A simple test gave the error "authentication client didn't specify
2822 service in request". It would seem that Dovecot has changed its
2823 interface. Fortunately there's a specification; I followed it and
2824 changed what the client sends and it appears to be working now.
2826 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2829 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2830 address A is aliased to B and C, where B exists and C does not. Without
2831 -v the output is "A verified" because verification stops after a
2832 successful redirection if more than one address is generated. However,
2833 with -v the child addresses are also verified. Exim was outputting "A
2834 failed to verify" and then showing the successful verification for C,
2835 with its parentage. It now outputs "B failed to verify", showing B's
2836 parentage before showing the successful verification of C.
2838 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2839 look up a TXT record in a specific list after matching in a combined
2842 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2843 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2844 they consult the DNS. I had assumed they would set it the way they
2845 wanted; and indeed my experiments on Linux seem to show that in some
2846 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2847 To be on the safe side, however, I have now made the interface to
2848 host_find_byname() similar to host_find_bydns(), with an argument
2849 containing the DNS resolver options. The host_find_byname() function now
2850 sets these options at its start, just as host_find_bydns() does. The smtp
2851 transport options dns_qualify_single and dns_search_parents are passed to
2852 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2853 of host_find_byname() use the default settings of RES_DEFNAMES
2854 (qualify_single) but not RES_DNSRCH (search_parents).
2856 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2857 spool_read_header() do less string testing, by means of a preliminary
2858 switch on the second character of optional "-foo" lines. (This is
2859 overdue, caused by the large number of possibilities that now exist.
2860 Originally there were few.) While I was there, I also converted the
2861 str(n)cmp tests so they don't re-test the leading "-" and the first
2862 character, in the hope this might squeeze out yet more improvement.
2864 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2865 flag allowing group syntax was set by the header_syntax check but not
2866 turned off, possible causing trouble later; (2) The flag was not being
2867 set at all for the header_verify test, causing "group"-style headers to
2868 be rejected. I have now set it in this case, and also caused header_
2869 verify to ignore an empty address taken from a group. While doing this, I
2870 came across some other cases where the code for allowing group syntax
2871 while scanning a header line wasn't quite right (mostly, not resetting
2872 the flag correctly in the right place). These bugs could have caused
2873 trouble for malformed header lines. I hope it is now all correct.
2875 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2876 with the "reply" argument non-NULL. The code, however (which originally
2877 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2878 but it didn't always do it. This confused somebody who was copying the
2879 code for some other use. I have removed all the tests.
2881 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2882 feature that was used to support insecure browsers during the U.S. crypto
2883 embargo. It requires special client support, and Exim is probably the
2884 only MTA that supported it -- and would never use it because real RSA is
2885 always available. This code has been removed, because it had the bad
2886 effect of slowing Exim down by computing (never used) parameters for the
2887 RSA_EXPORT functionality.
2889 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2890 authenticator to fail if there's a tab character in the incoming data
2891 (there should never be unless someone is messing about, as it's supposed
2892 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2893 if the connection is using TLS or if the remote IP is the same as the
2894 local IP, and the "valid-client-cert option" if a client certificate has
2897 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2898 authenticators. This can be used for authorization after authentication
2899 succeeds. (In the case of plaintext, it servers for both authentication
2902 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2903 if any retry times were supplied.
2905 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2906 connection, where there is no client IP address to check. In this
2907 situation, the verify now always succeeds.
2909 PH/25 Applied John Jetmore's -Mset patch.
2911 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2913 PH/27 In a string expansion for a processed (not raw) header when multiple
2914 headers of the same name were present, leading whitespace was being
2915 removed from all of them, but trailing whitespace was being removed only
2916 from the last one. Now trailing whitespace is removed from each header
2917 before concatenation. Completely empty headers in a concatenation (as
2918 before) are ignored.
2920 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2921 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2923 PH/29 [Removed. This was a change that I later backed out, and forgot to
2924 correct the ChangeLog entry (that I had efficiently created) before
2925 committing the later change.]
2927 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2928 address errors (4xx response to RCPT) over the same connection as other
2929 messages routed to the same hosts. Such deliveries are always "forced",
2930 so retry times are not inspected. This resulted in far too many retries
2931 for the affected addresses. The effect occurred only when there were more
2932 hosts than the hosts_max_try setting in the smtp transport when it had
2933 the 4xx errors. Those hosts that it had tried were not added to the list
2934 of hosts for which the message was waiting, so if all were tried, there
2935 was no problem. Two fixes have been applied:
2937 (i) If there are any address or message errors in an SMTP delivery, none
2938 of the hosts (tried or untried) are now added to the list of hosts
2939 for which the message is waiting, so the message should not be a
2940 candidate for sending over the same connection that was used for a
2941 successful delivery of some other message. This seems entirely
2942 reasonable: after all the message is NOT "waiting for some host".
2943 This is so "obvious" that I'm not sure why it wasn't done
2944 previously. Hope I haven't missed anything, but it can't do any
2945 harm, as the worst effect is to miss an optimization.
2947 (ii) If, despite (i), such a delivery is accidentally attempted, the
2948 routing retry time is respected, so at least it doesn't keep
2949 hammering the server.
2951 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2952 in ${readsocket because some servers need this prod.
2954 PH/32 Added some extra debug output when updating a wait-xxx database.
2956 PH/33 The hint "could be header name not terminated by colon", which has been
2957 given for certain expansion errors for a long time, was not being given
2958 for the ${if def:h_colon_omitted{... case.
2960 PH/34 The spec says: "With one important exception, whenever a domain list is
2961 being scanned, $domain contains the subject domain." There was at least
2962 one case where this was not true.
2964 PH/35 The error "getsockname() failed: connection reset by peer" was being
2965 written to the panic log as well as the main log, but it isn't really
2966 panic-worthy as it just means the connection died rather early on. I have
2967 removed the panic log writing for the ECONNRESET error when getsockname()
2970 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2971 runs only) independently of the message's sender address. This meant
2972 that, if the 4xx error was in fact related to the sender, a different
2973 message to the same recipient with a different sender could confuse
2974 things. In particular, this can happen when sending to a greylisting
2975 server, but other circumstances could also provoke similar problems.
2976 I have changed the default so that the retry time for these errors is now
2977 based a combination of the sender and recipient addresses. This change
2978 can be overridden by setting address_retry_include_sender=false in the
2981 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2982 remote server are returned as part of bounce messages. This was not
2983 happening for LMTP over a pipe (the lmtp transport), but now it is the
2984 same for both kinds of LMTP.
2986 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2987 in header lines that were in fact CNAMEs. This is no longer the case.
2989 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2990 and queue runs started by the daemon processed all messages. This has
2991 been fixed so that -R and -S can now usefully be given with -q<time>.
2993 PH/40 Import PCRE release 6.7 (fixes some bugs).
2995 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2997 PH/42 Give an error if -q is specified more than once.
2999 PH/43 Renamed the variables $interface_address and $interface_port as
3000 $received_ip_address and $received_port, to make it clear that these
3001 values apply to message reception, and not to the outgoing interface when
3002 a message is delivered. (The old names remain recognized, of course.)
3004 PH/44 There was no timeout on the connect() call when using a Unix domain
3005 socket in the ${readsocket expansion. There now is.
3007 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3008 be meaningful with "accept".
3010 SC/01 Eximstats V1.43
3011 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3013 SC/02 Eximstats V1.44
3014 Use a glob alias rather than an array ref in the generated
3015 parser. This improves both readability and performance.
3017 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3018 Collect SpamAssassin and rejection statistics.
3019 Don't display local sender or destination tables unless
3020 there is data to show.
3021 Added average volumes into the top table text output.
3023 SC/04 Eximstats V1.46
3024 Collect data on the number of addresses (recipients)
3025 as well as the number of messages.
3027 SC/05 Eximstats V1.47
3028 Added 'Message too big' to the list of mail rejection
3029 reasons (thanks to Marco Gaiarin).
3031 SC/06 Eximstats V1.48
3032 Mainlog lines which have GMT offsets and are too short to
3033 have a flag are now skipped.
3035 SC/07 Eximstats V1.49 (Alain Williams)
3036 Added the -emptyok flag.
3038 SC/08 Eximstats V1.50
3039 Fixes for obtaining the IP address from reject messages.
3041 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3042 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3043 whitespace changes from 4.64-PH/27
3045 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3048 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3049 are found, allow negative numbers in numeric criteria)
3051 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3053 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3054 to match changes made in 4.64-PH/43
3056 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3058 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3059 "domain" is actually an IP address, and force a failure. This locks out
3060 those revolvers/nameservers that support "A-for-A" lookups, in
3061 contravention of the specifications.
3063 PH/48 When a host name was looked up from an IP address, and the subsequent
3064 forward lookup of the name timed out, the host name was left in
3065 $sender_host_name, contrary to the specification.
3067 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3068 restricted to single-key lookups, Exim was not diagnosing an error if
3069 * or *@ was used with a query-style lookup.
3071 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3073 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3074 long ago, but noone interested enough thought of it. Let's just say that
3075 the "1.1" means that there are some new functions that weren't there at
3076 some point in the past.
3078 PH/51 Error processing for expansion failure of helo_data from an smtp
3079 transport during callout processing was broken.
3081 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3082 tested/used via the -bh/-bhc/-bs options.
3084 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3085 bug, fixed in subsequent PCRE releases).
3087 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3088 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3090 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3096 SC/01 Use a glob alias rather than an array ref in eximstats generated
3097 parser. This improves both readability and performance.
3099 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3100 Don't display local sender or destination tables in eximstats unless
3101 there is data to show.
3102 Added average volumes into the eximstats top table text output.
3104 SC/03 Collect data on the number of addresses (recipients) as well
3105 as the number of messages in eximstats.
3107 TF/01 Correct an error in the documentation for the redirect router. Exim
3108 does (usually) call initgroups() when daemonizing.
3110 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3111 with consistent privilege compared to when running as a daemon.
3113 TF/03 Note in the spec that $authenticated_id is not set for local
3114 submissions from trusted users.
3116 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3117 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3119 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3120 by adding some example configuration directives to the default
3121 configuration file. A little bit of work is required to uncomment the
3122 directives and define how usernames and passwords are checked, but
3123 there is now a framework to start from.
3125 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3126 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3127 without this. I don't know how relevant this is to other LDAP libraries.
3129 PH/02 Add the verb name to the "unknown ACL verb" error.
3131 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3133 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3135 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3136 directory not expanded when it should be if an expanded home directory
3137 was set for the address (which is overridden by the transport).
3139 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3142 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3143 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3144 because it is too late at that time, and has no effect.
3146 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3147 security issue with \' (bugzilla #107). I could not use the
3148 PQescapeStringConn() function, because it needs a PGconn value as one of
3151 PH/08 When testing addresses using -bt, indicate those final addresses that
3152 are duplicates that would not cause an additional delivery. At least one
3153 person was confused, thinking that -bt output corresponded to deliveries.
3154 (Suppressing duplicates isn't a good idea as you lose the information
3155 about possibly different redirections that led to the duplicates.)
3157 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3158 systems where poll() doesn't work, in particular OS X.
3160 PH/10 Added more information to debugging output for retry time not reached.
3162 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3163 operations in malware.c.
3165 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3168 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3169 syslog_duplication set false, log lines that would normally be written
3170 both the the main log and to the reject log were not written to syslog at
3173 PH/14 In the default configuration, change the use of "message" in ACL warn
3174 statements to "add_header".
3176 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3177 not followed by a command (e.g. "seen endif").
3179 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3180 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3183 PH/17 Added extra conditions to the default value of delay_warning_condition
3187 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3188 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3189 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3192 The Auto-Submitted: and various List- headers are standardised, whereas I
3193 don't think Precedence: ever was.
3195 PH/18 Refactored debugging code in route_finduser() to show more information,
3196 in particular, the error code if getpwnam() issues one.
3198 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3199 This is apparently needed in addition to the PH/07 change above to avoid
3200 any possible encoding problems.
3202 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3203 but not after initializing Perl.
3205 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3206 output them only if debugging. By default they are written stderr,
3207 apparently, which is not desirable.
3209 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3212 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3215 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3217 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3218 authenticated or an ident call has been made. Suppress the default
3219 values for $authenticated_id and $authenticated_sender (but permit -oMai
3220 and -oMas) when testing with -bh.
3222 PH/24 Re-jigged the order of the tests in the default configuration so that the
3223 tests for valid domains and recipients precede the DNS black list and CSA
3224 tests, on the grounds that those ones are more expensive.
3226 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3227 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3228 This bug exists in every version of Exim that I still have, right back to
3231 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3232 However, an attempt to turn on case-sensitivity in a regex key by
3233 including (?-i) didn't work because the subject string was already
3234 lowercased, and the effects were non-intuitive. It turns out that a
3235 one-line patch can be used to allow (?-i) to work as expected.
3241 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3242 other effects) broke the use of negated acl sub-conditions.
3244 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3247 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3248 "Deny" causes Exim to reject the incoming connection with a 554 error.
3249 Unfortunately, if there is a major crisis, such as a disk failure,
3250 tcp-wrappers gives "deny", whereas what one would like would be some
3251 kind of temporary error. A kludge has been added to help with this.
3252 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3253 554 error is used if errno is still zero or contains ENOENT (which occurs
3254 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3257 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3259 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3260 errors. Otherwise a message that provokes a temporary error (when other
3261 messages do not) can cause a whole host to time out.
3263 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3264 addresses were routed directly to files or pipes from a redirect router.
3265 File deliveries just didn't batch; pipe deliveries might have suffered
3268 PH/06 A failure to get a lock for a hints database would erroneously always say
3269 "Failed to get write lock", even when it was really a read lock.
3271 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3272 of 0600. This has been changed to use the value of the lockfile_mode
3273 option (which defaults to 0600).
3275 PH/08 Applied small patch from the Sieve maintainer.
3277 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3278 folder from quota calculations, a direct delivery into this folder messed
3279 up the contents of the maildirsize file. This was because the regex was
3280 used only to exclude .Trash (or whatever) when the size of the mailbox
3281 was calculated. There was no check that a delivery was happening into an
3282 excluded directory. This bug has been fixed by ignoring all quota
3283 processing for deliveries into excluded directories.
3285 PH/10 Added the maildirfolder_create_regex option to appendfile.
3291 PH/01 The code for finding all the local interface addresses on a FreeBSD
3292 system running IPv6 was broken. This may well have applied to all BSD
3293 systems, as well as to others that have similar system calls. The broken
3294 code found IPv4 interfaces correctly, but gave incorrect values for the
3295 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3296 that it would not match correctly against @[] and not recognize the IPv6
3299 PH/02 The ipliteral router was not recognizing addresses of the form user@
3300 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3302 PH/03 Added disable_ipv6.
3304 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3305 decoded form, because it is most often used to construct To: headers
3306 lines in autoreplies, and the decoded form may well be syntactically
3307 invalid. However, $reply_address has leading white space removed, and all
3308 newlines turned into spaces so that the autoreply transport does not
3311 PH/05 If group was specified without a user on a router, and no group or user
3312 was specified on a transport, the group from the router was ignored.
3314 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3315 for visible compile-time settings that can be used to change these
3316 numbers, for those that want even more. Backwards compatibility with old
3317 spool files has been maintained. However, going back to a previous Exim
3318 release will lost any variables that are in spool files.
3320 PH/07 Two small changes when running in the test harness: increase delay when
3321 passing a TCP/IP connection to a new process, in case the original
3322 process has to generate a bounce, and remove special handling of
3323 127.0.0.2 (sic), which is no longer necessary.
3325 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3326 be the same on different OS.
3328 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3331 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3332 whether --show-vars was specified or not
3334 JJ/02 exipick: Added support for new ACL variable spool format introduced
3337 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3338 syntactically invalid From: or Reply-to: line, and a filter used this to
3339 generate an autoreply, and therefore failed to obtain an address for the
3340 autoreply, Exim could try to deliver to a non-existent relative file
3341 name, causing unrelated and misleading errors. What now happens is that
3342 it logs this as a hard delivery error, but does not attempt to create a
3345 PH/11 The exinext utility has a -C option for testing purposes, but although
3346 the given file was scanned by exinext itself; it wasn't being passed on
3347 when Exim was called.
3349 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3350 an end-of-file indication when reading a command response.
3352 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3353 compiled. In many other places in Exim, IPv6 addresses are always
3354 recognized, so I have changed this. It also means that IPv4 domain
3355 literals of the form [IPV4:n.n.n.n] are now always recognized.
3357 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3358 used if the router is not running as root, for example, when verifying at
3359 ACL time, or when using -bh. The debugging output from this situation was
3360 non-existent - all you got was a failure to exec. I have made two
3363 (a) Failures to set uid/gid, the current directory, or a process leader
3364 in a subprocess such as that created by queryprogram now generate
3365 suitable debugging output when -d is set.
3367 (b) The queryprogram router detects when it is not running as root,
3368 outputs suitable debugging information if -d is set, and then runs
3369 the subprocess without attempting to change uid/gid.
3371 PH/15 Minor change to Makefile for building test_host (undocumented testing
3374 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3375 additional section of a DNS packet that returns MX or SRV records.
3376 Instead, it always explicitly searches for A/AAAA records. This avoids
3377 major problems that occur when a DNS server includes only records of one
3378 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3379 fixed another bug: if SRV records were looked up and the corresponding
3380 address records were *not* found in the additional section, the port
3381 values from the SRV records were lost.
3383 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3384 using the correct key (the original address) when searching the retry
3385 rules in order to find which one to use for generating the retry hint.
3387 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3388 adding the default one. Similarly, if it contains a Reply-To: header, the
3389 errors_reply_to option, if set, is not used.
3391 PH/19 When calculating a retry time, Exim used to measure the "time since
3392 failure" by looking at the "first failed" field in the retry record. Now
3393 it does not use this if it is later than than the arrival time of the
3394 message. Instead it uses the arrival time. This makes for better
3395 behaviour in cases where some deliveries succeed, thus re-setting the
3396 "first failed" field. An example is a quota failure for a huge message
3397 when small messages continue to be delivered. Without this change, the
3398 "time since failure" will always be short, possible causing more frequent
3399 delivery attempts for the huge message than are intended.
3400 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3402 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3403 $1, $2, $3) because the numerical variables can be reset during some
3404 expansion items (e.g. "match"), thereby losing the authentication data.
3406 PH/21 Make -bV show the size of off_t variables so that the test suite can
3407 decide whether to run tests for quotas > 2G.
3409 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3410 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3411 filecount value is greater than 2G or if a quota value is greater than 2G
3412 on a system where the size of off_t is not greater than 4, a panic error
3415 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3416 never match. The debug and -bh output now contains an explicit error
3417 message indicating a malformed IPv4 address or mask.
3419 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3420 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3421 PH/23 above applies.
3423 PH/25 Do not write to syslog when running in the test harness. The only
3424 occasion when this arises is a failure to open the main or panic logs
3425 (for which there is an explicit test).
3427 PH/26 Added the /no_tell option to "control=freeze".
3429 PH/27 If a host name lookup failed very early in a connection, for example, if
3430 the IP address matched host_lookup and the reverse lookup yielded a name
3431 that did not have a forward lookup, an error message of the form "no IP
3432 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3433 could be logged. Now it outputs the IP address instead of "NULL".
3435 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3436 allows the sender and the authenticated sender to be set when
3437 submitting a message from within Exim. Since child_open_exim() is
3438 documented for local_scan(), the new function should be too.
3440 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3441 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3442 results in an empty string is now treated as unset.
3444 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3446 PH/31 Added sender_verify_fail logging option.
3448 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3449 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3450 tidied the source and removed it altogether.
3452 PH/33 When a queue run was abandoned because the load average was too high, a
3453 log line was always written; now it is written only if the queue_run log
3454 selector is set. In addition, the log line for abandonment now contains
3455 information about the queue run such as the pid. This is always present
3456 in "start" and "stop" lines but was omitted from the "abandon" line.
3458 PH/34 Omit spaces between a header name and the colon in the error message that
3459 is given when verify = headers_syntax fails (if there are lots of them,
3460 the message gets confusing).
3462 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3463 names, as there are now some PTR records that contain slashes. This check
3464 is only to protect against broken name servers that fall over on strange
3465 characters, so the fact that it applies to all lookups doesn't matter.
3467 PH/36 Now that the new test suite is complete, we can remove some of the
3468 special code in Exim that was needed for the old test suite. For example,
3469 sorting DNS records because real resolvers return them in an arbitrary
3470 order. The new test suite's fake resolver always returns records in the
3473 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3474 bounces) except when queue_only is set, to avoid logging races between
3475 the different processes.
3477 PH/38 Panic-die if .include specifies a non-absolute path.
3479 PH/39 A tweak to the "H" retry rule from its user.
3481 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3482 a label. They prevented compilation on older perls.
3484 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3485 a warning to be raised on newish perls.
3487 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3488 on queue. Changes to match documented behaviour of showing count of
3489 messages matching specified criteria.
3491 PH/40 Changed the default ident timeout from 30s to 5s.
3493 PH/41 Added support for the use of login_cap features, on those BSD systems
3494 that have them, for controlling the resources used by pipe deliveries.
3496 PH/42 The content-scanning code uses fopen() to create files in which to put
3497 message data. Previously it was not paying any attention to the mode of
3498 the files. Exim runs with umask(0) because the rest of the code creates
3499 files with open(), and sets the required mode explicitly. Thus, these
3500 files were ending up world-writeable. This was not a big issue, because,
3501 being within the spool directory, they were not world-accessible. I have
3502 created a function called modefopen, which takes an additional mode
3503 argument. It sets umask(777), creates the file, chmods it to the required
3504 mode, then resets the umask. All the relevant calls to fopen() in the
3505 content scanning code have been changed to use this function.
3507 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3508 to 24 hours. This avoids potential overflow problems when processing G
3509 and H retry rules. I suspect nobody ever tinkers with this value.
3511 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3513 PH/45 When the plaintext authenticator is running as a client, the server's
3514 challenges are checked to ensure they are valid base64 strings. By
3515 default, the authentication attempt is cancelled if an invalid string is
3516 received. Setting client_ignore_invalid_base64 true ignores these errors.
3517 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3518 they are received. Thus, the responses can be made to depend on the
3519 challenges. If an invalid string is ignored, an empty string is placed in
3522 PH/46 Messages that are created by the autoreply transport now contains a
3523 References: header, in accordance with RFCs 2822 and 3834.
3525 PH/47 Added authenticated_sender_force to the smtp transport.
3527 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3529 PH/49 Installed latest patch from the Sieve maintainer.
3531 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3532 was also set, the appendfile transport was unnecessarily scanning a
3533 directory of message files (e.g. for maildir delivery) to find the count
3534 of files (along with the size), even though it did not need this
3535 information. It now does the scan only if it needs to find either the
3536 size of the count of files.
3538 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3540 PH/52 Two bugs concerned with error handling when the smtp transport is
3543 (i) Exim was not creating retry information for temporary errors given
3544 for individual recipients after the DATA command when the smtp transport
3545 was used in LMTP mode. This meant that they could be retried too
3546 frequently, and not timed out correctly.
3548 (ii) Exim was setting the flag that allows error details to be returned
3549 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3550 recipients that were returned after the DATA command.
3552 PH/53 This is related to PH/52, but is more general: for any failing address,
3553 when detailed error information was permitted to be returned to the
3554 sender, but the error was temporary, then after the final timeout, only
3555 "retry timeout exceeded" was returned. Now it returns the full error as
3556 well as "retry timeout exceeded".
3558 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3559 do this, and (what is worse) MTAs that accept it.
3561 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3562 will now be deprecated.
3564 PH/56 New os.c-cygwin from the Cygwin maintainer.
3566 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3567 formats (previously only available in exim formats via -bpr, -bpru,
3568 and -bpra. Now also available in native and exiqgrep formats)
3570 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3571 with very large, slow to parse queues
3573 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3575 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3577 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3578 responses to authentication challenges, though it was showing the
3579 challenges; (ii) I've removed the CR characters from the debug output for
3582 PH/58 Allow for the insertion of a newline as well as a space when a string
3583 is turned into more than one encoded-word during RFC 2047 encoding. The
3584 Sieve code now uses this.
3586 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3587 data_4xx, lost_connection, tls_required.
3589 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3590 message was being sent as an SMTP response.
3592 PH/61 Add -l and -k options to exicyclog.
3594 PH/62 When verifying, if an address was redirected to one new address, so that
3595 verification continued, and the new address failed or deferred after
3596 having set something in $address_data, the value of $address_data was not
3597 passed back to the ACL. This was different to the case when no
3598 redirection occurred. The value is now passed back in both cases.
3600 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3601 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3602 use login_cap.h, so on its own it isn't the distinguishing feature. The
3603 new name refers directly to the setclassresources() function.
3605 PH/65 Added configuration files for NetBSD3.
3607 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3609 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3610 is preferred over IPv4.
3612 PH/68 The bounce_return_message and bounce_return_body options were not being
3613 honoured for bounces generated during the reception of non-SMTP messages.
3614 In particular, this applied to messages rejected by the ACL. This bug has
3615 been fixed. However, if bounce_return_message is true and bounce_return_
3616 body is false, the headers that are returned for a non-SMTP message
3617 include only those that have been read before the error was detected.
3618 (In the case of an ACL rejection, they have all been read.)
3620 PH/69 The HTML version of the specification is now built in a directory called
3621 spec_html instead of spec.html, because the latter looks like a path with
3622 a MIME-type, and this confuses some software.
3624 PH/70 Catch two compiler warnings in sieve.c.
3626 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3627 function verify_get_ident() calls ip_connect() to connect a socket, but
3628 if the "connect()" function timed out, ip_connect() used to close the
3629 socket. However, verify_get_ident() also closes the socket later, and in
3630 between Exim writes to the log, which may get opened at this point. When
3631 the socket was closed in ip_connect(), the log could get the same file
3632 descriptor number as the socket. This naturally causes chaos. The fix is
3633 not to close the socket in ip_connect(); the socket should be closed by
3634 the function that creates it. There was only one place in the code where
3635 this was missing, in the iplookup router, which I don't think anybody now
3636 uses, but I've fixed it anyway.
3638 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3639 well as to direct DNS lookups. Otherwise the handling of names in host
3640 lists is inconsistent and therefore confusing.
3646 PH/01 Two changes to the default runtime configuration:
3648 (1) Move the checks for relay_from_hosts and authenticated clients from
3649 after to before the (commented out) DNS black list checks.
3651 (2) Add control=submission to the relay_from_hosts and authenticated
3652 clients checks, on the grounds that messages accepted by these
3653 statements are most likely to be submissions.
3655 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3657 (1) Generate an error if the third argument for the ${prvs expansion is
3660 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3663 (3) Reset the variables that are obtained from the first argument of
3664 ${prvscheck and used in the second argument before leaving the code,
3665 because their memory is reclaimed, so using them afterwards may do
3668 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3669 one (it's much easier than Tom thought :-).
3671 (5) Because of (4), we can now allow for the use of $prvscheck_result
3672 inside the third argument.
3674 PH/03 For some reason, the default setting of PATH when running a command from
3675 a pipe transport was just "/usr/bin". I have changed it to
3678 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3679 anything to be listed in the output from -bV.
3681 PH/05 When a filter generated an autoreply, the entire To: header line was
3682 quoted in the delivery log line, like this:
3684 => >A.N.Other <ano@some.domain> <original@ddress> ...
3686 This has been changed so that it extracts the operative address. There
3687 may be more than one such address. If so, they are comma-separated, like
3690 => >ano@some.domain,ona@other.domain <original@ddress> ...
3692 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3693 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3694 not being looked up in the rDNS to get a host name, Exim was showing the
3695 IP address twice in Received: lines, even though the IP addresses were
3696 identical. For example:
3698 Received: from [1.2.3.4] (helo=[1.2.3.4])
3700 However, if the real host name was known, it was omitting the HELO data
3701 if it matched the actual IP address. This has been tidied up so that it
3702 doesn't show the same IP address twice.
3704 PH/07 When both +timestamp and +memory debugging was on, the value given by
3705 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3706 called by the debug printing, thereby overwriting the timestamp buffer.
3707 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3709 PH/08 When the original message was included in an autoreply transport, it
3710 always said "this is a copy of the message, including all the headers",
3711 even if body_only or headers_only was set. It now gives an appropriate
3714 PH/09 Applied a patch from the Sieve maintainer which:
3716 o fixes some comments
3717 o adds the (disabled) notify extension core
3718 o adds some debug output for the result of if/elsif tests
3719 o points to the current vacation draft in the documentation
3720 and documents the missing references header update
3724 o fixes a bug in processing the envelope test (when testing
3725 multiple envelope elements, the last element determined the
3728 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3729 Electronic Mail") by including:
3731 Auto-submitted: auto-generated
3733 in the messages that it generates (bounce messages and others, such as
3734 warnings). In the case of bounce messages for non-SMTP messages, there was
3735 also a typo: it was using "Auto_submitted" (underscore instead of
3736 hyphen). Since every message generated by Exim is necessarily in response
3737 to another message, thes have all been changed to:
3739 Auto-Submitted: auto-replied
3741 in accordance with these statements in the RFC:
3743 The auto-replied keyword:
3745 - SHOULD be used on messages sent in direct response to another
3746 message by an automatic process,
3748 - MUST NOT be used on manually-generated messages,
3750 - MAY be used on Delivery Status Notifications (DSNs) and Message
3751 Disposition Notifications (MDNs),
3753 - MUST NOT be used on messages generated by automatic or periodic
3754 processes, except for messages which are automatic responses to
3757 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3758 to the default Received: header definition.
3760 PH/12 Added log selector acl_warn_skipped (default on).
3762 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3763 variables because (a) they are in the wrong storage pool and (b) even if
3764 they were copied, it wouldn't work properly because of the caching.
3766 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3767 checking when decoding. Apparently there are clients that generate
3768 overlong encoded strings. Why am I not surprised?
3770 PH/15 If the first argument of "${if match_address" was not empty, but did not
3771 contain an "@" character, Exim crashed. Now it writes a panic log message
3772 and treats the condition as false.
3774 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3776 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3777 extension "envelope-auth". The code is finished and in agreement with
3778 other implementations, but there is no documentation so far and in fact,
3779 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3780 not changing the active code.
3782 Print executed "if" and "elsif" statements when debugging is used. This
3783 helps a great deal to understand what a filter does.
3785 Document more things not specified clearly in RFC3028. I had all this
3786 sorted out, when out of a sudden new issues came to my mind. Oops."
3788 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3791 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3792 canonical form (as documented). However, after a host name lookup from
3793 the IP address, check_host() was doing a simple string comparison with
3794 addresses acquired from the DNS when checking that the found name did
3795 have the original IP as one of its addresses. Since any found IPv6
3796 addresses are likely to be in abbreviated form, the comparison could
3797 fail. Luckily, there already exists a function for doing the comparison
3798 by converting both addresses to binary, so now that is used instead of
3799 the text comparison.
3801 PH/20 There was another similar case to PH/19, when a complete host name was
3802 given in a host list; looking up its IP address could give an abbreviated
3803 form, whereas the current host's name might or might not be abbreviated.
3804 The same fix has been applied.
3810 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3811 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3814 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3815 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3817 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3819 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3820 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3821 Also, for "Auto-Submitted", treat anything other than "no" as
3822 non-personal, in accordance with RFC 3834. (Previously it treated
3823 anything starting "auto-" as non-personal.)
3825 TF/01 The control=submission/name=... option had a problem with syntax
3826 errors if the name included a slash character. The /name= option
3827 now slurps the rest of the string, so it can include any characters
3828 but it must come last in the list of options (after /sender_retain
3831 PH/05 Some modifications to the interface to the fake nameserver for the new
3839 TK/01 Added the "success_on_redirect" address verification option. See
3840 NewStuff for rationale and an example.
3842 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3844 PH/02 Patch to exigrep to allow it to work on syslog lines.
3846 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3847 fread() to skip over the body file's header line, because in Cygwin the
3848 header line is locked and is inaccessible.
3850 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3851 co-exist for some time) to make it clear that it is the Exim ID that is
3852 referenced, not the Message-ID: header line.
3854 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3855 string_format() function, because snprintf() does not exist on all
3858 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3859 ${stat: expansion item.
3861 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3862 protocol synchronization error", to keep the pedants happy.
3864 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3865 well as for IRIX systems, when gcc is being used. See the host.c source
3868 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3870 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3873 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3874 transport and to the smtp transport in LMTP mode.
3876 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3878 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3879 run a filter in a subprocess. This could lead to confusion in subsequent
3880 lookups in the parent process. There should also be a search_tidyup() at
3881 the end of the subprocess.
3883 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3884 only if the host matched helo_try_verify_hosts, which caused the
3885 verification to occur when the EHLO/HELO command was issued. The ACL just
3886 tested the remembered result. Now, if a previous verification attempt has
3887 not happened, "verify = helo" does it there and then.
3889 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3891 TK/03 Fix log output including CR from clamd.
3893 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3894 exist provoked a memory error which could cause a segfault.
3896 PH/15 Installed PCRE 6.2
3898 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3900 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3901 of the problem. Specifically, suggested +O2 rather than +O1 for the
3904 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3906 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3907 record, the port was not correctly logged when the outgoing_port log
3908 selector was set (it logged the transort's default port).
3910 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3911 fallback_hosts, and "hosts" in the smtp transport.
3913 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3914 host errors such as "Connection refused".
3916 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3917 authentication with radiusclient 0.4.9:
3919 - Error returned from rc_read_config was caught wrongly
3920 - Username/password not passed on to radius server due to wrong length.
3922 The presumption is that some radiusclient API changes for 4.51/PH/17
3923 were not taken care of correctly. The code is still untested by me (my
3924 Linux distribution still has 0.3.2 of radiusclient), but it was
3925 contributed by a Radius user.
3927 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3928 expanding the "port" option of the smtp transport.
3930 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3931 while reading a MIME header. Thanks to Tom Hughes for a patch.
3933 PH/24 Include config.h inside local_scan.h so that configuration settings are
3936 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3937 that in an ACL for RCPT (for example), you can examine exactly what was
3940 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3941 commands, but it was not correctly comparing the address with the actual
3942 client host address. Thus, it would show the EHLO address in Received:
3943 header lines when this was not necessary.
3945 PH/27 Added the % operator to ${eval:}.
3947 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3948 it should be ignoring failures (because with -C, for example, it has lost
3949 privilege). It wasn't ignoring creation failures other than "already
3952 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3955 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3956 because an input line was too long, either on its own, or by virtue of
3957 too many continuations, the temporary file was not being removed, and the
3958 return code was incorrect.
3960 PH/31 Missing "BOOL" in function definition in filtertest.c.
3962 PH/32 Applied Sieve patches from the maintainer.
3964 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3966 PH/33 Added "verify = not_blind".
3968 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3969 Local/Makefile (with some defaults set). These are used in built scripts
3970 such as exicyclog, but they have never been used in the exim_install
3971 script (though there are many overriding facilities there). I have
3972 arranged that the exim_install script now takes note of these two
3975 PH/35 Installed configuration files for Dragonfly.
3977 PH/36 When a locally submitted message by a trusted user did not contain a
3978 From: header, and the sender address was obtained from -f or from an SMTP
3979 MAIL command, and the trusted user did not use -F to supply a sender
3980 name, $originator_name was incorrectly used when constructing a From:
3981 header. Furthermore, $originator_name was used for submission mode
3982 messages from external hosts without From: headers in a similar way,
3983 which is clearly wrong.
3985 PH/37 Added control=suppress_local_fixups.
3987 PH/38 When log_selector = +received_sender was set, and the addition of the
3988 sender made the log line's construction buffer exactly full, or one byte
3989 less than full, an overflow happened when the terminating "\n" was
3992 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3993 when the result of a list match is failure because a DNS lookup failed.
3995 PH/40 RM_COMMAND is now used in the building process.
3997 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3998 the "build-* directories that it finds.
4000 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4001 address in a domain literal was a prefix of an interface address.
4003 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4004 when verifying a sender address, unless rewrite_headers is false.
4006 PH/44 Wrote a long comment about why errors_to addresses are verified as
4007 recipients, not senders.
4009 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4010 the ratelimit ACL was added.
4012 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4014 PH/46 Added extra information about PostgreSQL errors to the error string.
4016 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4017 avoiding the need to install special zones in a real server. This is
4018 backwards compatible; if it can't find the fake resolver, it drops back.
4019 Thus, both old and new test suites can be run.
4021 TF/02 Added util/ratelimit.pl
4023 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4024 clock is set back in time.
4026 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4027 Candler <B.Candler@pobox.com>.
4029 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4030 OK for addresses that are the result of redirections.
4032 PH/48 A number of further additions for the benefit of the new test suite,
4033 including a fake gethostbyname() that interfaces to the fake DNS resolver
4036 TF/06 The fix for widen_domains has also been applied to qualify_single and
4037 search_parents which are the other dnslookup options that can cause
4040 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4043 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4045 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4046 service" if it encountered an MX record with an empty target hostname.
4047 The message is now "an MX or SRV record indicated no SMTP service".
4049 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4050 if the DNS of the sending site is misconfigured. This is quite a
4051 common situation. This change restores the behaviour of treating a
4052 helo verification defer as a failure.
4054 PH/51 If self=fail was set on a router, the bounce message did not include the
4055 actual error message.
4061 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4063 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4064 command itself ended in error, the underlying message about the transport
4065 filter timeout was being overwritten with the pipe command error. Now the
4066 underlying error message should be appended to the second error message.
4068 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4070 PH/02 Reduce the amount of output that "make" produces by default. Full output
4071 can still be requested.
4073 PH/03 The warning log line about a condition test deferring for a "warn" verb
4074 was being output only once per connection, rather than after each
4075 occurrence (because it was using the same function as for successful
4076 "warn" verbs). This seems wrong, so I have changed it.
4078 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4079 it should not have, which might have caused a crash in the right
4080 circumstances, but probably never did.
4082 PH/04 Installed a modified version of Tony Finch's patch to make submission
4083 mode fix the return path as well as the Sender: header line, and to
4084 add a /name= option so that you can make the user's friendly name appear
4087 TF/03 Added the control = fakedefer ACL modifier.
4089 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4090 Mark Lowes for thorough testing.
4092 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4094 TK/03 Merged latest SRS patch from Miles Wilton.
4096 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4097 with the definition in sysexits.h (which is #included earlier).
4098 Fortunately, Exim does not actually use EX_OK. The code used to try to
4099 preserve the sysexits.h value, by assuming that macro definitions were
4100 scanned for macro replacements. I have been disabused of this notion,
4101 so now the code just undefines EX_OK before #including unistd.h.
4103 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4104 in the smtp transport. When a block could not be written in a single
4105 write() function, the timeout was being re-applied to each part-write.
4106 This seems wrong - if the receiver was accepting one byte at a time it
4107 would take for ever. The timeout is now adjusted when this happens. It
4108 doesn't have to be particularly precise.
4110 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4111 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4113 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4114 without a domain if the check to <postmaster@domain> fails.
4116 SC/01 Eximstats: added -xls and the ability to specify output files
4117 (patch written by Frank Heydlauf).
4119 SC/02 Eximstats: use FileHandles for outputting results.
4121 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4123 SC/04 Eximstats: fixed display of large numbers with -nvr option
4125 SC/05 Eximstats: fixed merging of reports with empty tables.
4127 SC/06 Eximstats: added the -include_original_destination flag
4129 SC/07 Eximstats: removed tabs and trailing whitespace.
4131 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4133 TK/06 MBOX spool code: Add real "From " MBOX separator line
4134 so the .eml file is really in mbox format (even though
4135 most programs do not really care). Patch from Alex Miller.
4137 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4138 The latter is generated from $received_to and is only set if the
4139 message has one envelope recipient. SA can use these headers,
4140 obviously out-of-the-box. Patch from Alex Miller.
4142 PH/08 The ${def test on a variable was returning false if the variable's
4143 value was "0", contrary to what the specification has always said!
4144 The result should be true unless the variable is empty.
4146 PH/09 The syntax error of a character other than { following "${if
4147 def:variable_name" (after optional whitespace) was not being diagnosed.
4148 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4149 accidental colon was present, for example, could give incorrect results.
4151 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4152 result is used (not including appendfile, where other changes are about
4155 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4156 This involved changing a lot of size variables from int to off_t. It
4157 should work with maildirs and everything.
4159 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4160 spamd dying while we are connected to it.
4162 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4165 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4166 function for BDB 4.3.
4168 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4170 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4171 the sources for PCRE have been re-arranged and more files are now
4174 PH/15 The code I had for printing potentially long long variables in PH/11
4175 above was not the best (it lost precision). The length of off_t variables
4176 is now inspected at build time, and an appropriate printing format (%ld
4177 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4178 to be "long long int" or "long int". This is needed for the internal
4179 formatting function string_vformat().
4181 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4182 the configuration file to be ":syslog", then the script "guesses" where
4183 the logs files are, rather than using the compiled in default. In our
4184 case the guess is not the same as the compiled default, so the script
4185 suddenly stopped working when I started to use syslog. The patch checks
4186 to see if log_file_path is "". If so, it attempts to read it from exim
4187 with no configuration file to get the compiled in version, before it
4188 falls back to the previous guessing code."
4190 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4191 implementing BATV in an Exim configuration. See NewStuff for the gory
4194 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4195 Makefile that are specific to HP-UX.
4197 PH/18 If the "use_postmaster" option was set for a recipient callout together
4198 with the "random" option, the postmaster address was used as the MAIL
4199 FROM address for the random test, but not for the subsequent recipient
4200 test. It is now used for both.
4202 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4203 patch removes a few documentation additions to RFC 3028, because the
4204 latest draft now contains them. It adds the new en;ascii-case comparator
4205 and a new error check for 8bit text in MIME parts. Comparator and
4206 require names are now matched exactly. I enabled the subaddress
4207 extension, but it is not well tested yet (read: it works for me)."
4209 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4210 rework some of the code of TK/09 above to avoid the hardwired use of
4211 "%lld" and "long long". Replaced the call to snprintf() with a call to
4214 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4215 records point to non-existent hosts", "retry timeout exceeded", and
4216 "retry time not reached for any host after a long failure period".
4218 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4219 experimental DomainKeys support:
4221 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4222 (2) On an error such as an illegally used "control", the wrong name for
4223 the control was given.
4225 These problems did NOT occur unless DomainKeys support was compiled.
4227 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4229 PH/24 Added ${if match_ip condition.
4231 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4232 databases so that it will be absolutely obvious if a crash occurs in the
4233 DB library. This is a regular occurrence (often caused by mis-matched
4236 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4237 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4238 picked up on a user's system that detects such things. There doesn't seem
4239 to be a gcc warning option for this - only an attribute that has to be
4240 put on the function's prototype. It seems that in Fedora Core 4 they have
4241 set this on a number of new functions. No doubt there will be more in due
4244 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4245 specify a transport. However, if an address that was verified by such a
4246 router was the subject of a callout, Exim crashed because it tried to
4247 read the rcpt_include_affixes from the non-existent transport. Now it
4248 just assumes that the setting of that option is false. This bug was
4249 introduced by 4.51/PH/31.
4251 PH/28 Changed -d+all to exclude +memory, because that information is very
4252 rarely of interest, but it makes the output a lot bigger. People tend to
4253 do -d+all out of habit.
4255 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4256 code in os-type was giving problems when libc.so lives in lib64, like on
4259 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4260 aren't the modern standard, and it seems that some systems' include files
4261 don't always have them. Exim was already checking for some of the newer
4262 ones like T_AAAA, and defining it itself. I've added checks for all the
4263 record types that Exim uses.
4265 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4266 not automatically generating a new one, as it is supposed to. This
4267 prevented TLS from working. If the file did exist, but contained invalid
4268 data, a new version was generated, as expected. It was only the case of a
4269 non-existent file that was broken.
4271 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4272 with a change in libdomainkeys > 0.64.
4274 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4275 from DNS. If the selector record carries the flag, it now has
4276 precedence over the domain-wide flag.
4278 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4280 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4281 the use of a port name that isn't defined in /etc/services) occurred, the
4282 message was deferred as in a normal delivery, and thus remained on the
4283 spool, instead of being failed because of the mua_wrapper setting. This
4284 is now fixed, and I tidied up some of the mua_wrapper messages at the
4287 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4288 the messages in a hash of arrays rather than using individual hashes.
4289 This is a bit cleaner and results in dramatic memory savings, albeit
4290 at a slight CPU cost.
4292 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4293 as requested by Marc Sherman.
4295 SC/10 Eximstats: added histograms for user specified patterns as requested
4298 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4300 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4301 fopen() in the content-scanning modules that did not already have it.
4307 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4308 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4310 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4312 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4314 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4315 to test Sieve filters that use "vacation".
4317 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4318 that changes the way the GnuTLS parameters are stored in the cache file.
4319 The new format can be generated externally. For backward compatibility,
4320 if the data in the cache doesn't make sense, Exim assumes it has read an
4321 old-format file, and it generates new data and writes a new file. This
4322 means that you can't go back to an older release without removing the
4325 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4326 work if there are any delivery delays because "one_time" forces the
4327 parent to be marked "delivered", so its unseen clone is never tried
4328 again. For this reason, Exim now forbids the simultaneous setting of
4331 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4332 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4333 entry if you want to know the details. The fix, however, neglected to
4334 consider the case when local delivery batching is involved. The test for
4335 "previously delivered" was not happening when checking to see if an
4336 address could be batched with a previous (undelivered) one; under
4337 certain circumstances this could lead to multiple deliveries to the same
4340 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4341 in its include files, and this causes problems building Exim.
4343 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4344 header_syntax) but Exim was just ignoring anything given after a slash.
4345 In particular, this caused confusion with an attempt to use "verify =
4346 reverse_host_lookup/defer_ok". An error is now given when options are
4347 supplied for verify items that do not have them. (Maybe reverse_host_
4348 lookup should have a defer_ok option, but that's a different point.)
4350 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4351 defined by RFC 821) to 2048, because there were problems with some AUTH
4352 commands, and RFC 1869 says the size should be increased for extended
4353 SMTP commands that take arguments.
4355 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4358 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4359 "unknown" error; now it says that the functionality isn't in the binary.
4361 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4362 an address' error message when a string expansion fails (syntax or
4363 whatever). Otherwise the password may appear in the log. Following change
4364 PH/42 below, there is no longer a chance of it appearing in a bounce
4367 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4369 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4370 of its addresses was ever tried. (Bugzilla bug #2.)
4372 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4373 the result incorrectly in the debug output. (It correctly added a newline
4374 to what was transported.)
4376 TF/01 Added $received_time.
4378 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4379 commented out examples of how to interface to a virus scanner and to
4380 SpamAssassin. Also added commented examples of av_scanner and
4381 spamd_address settings.
4383 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4384 and controls are allowed in which ACLs. There were a couple of minor
4385 errors. Some of the entries in the conditions table (which is a table of
4386 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4387 negation of where the condition IS allowed.
4389 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4391 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4392 header file does not have a version number, so I've had to invent a new
4393 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4394 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4395 radiusclient), but it was contributed by a Radius user.
4397 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4398 files or directories, for OpenSSL.
4400 PH/19 When an Exim process that is running as root has to create an Exim log
4401 file, it does so in a subprocess that runs as exim:exim so as to get the
4402 ownership right at creation (otherwise, other Exim processes might see
4403 the file with the wrong ownership). There was no test for failure of this
4404 fork() call, which would lead to the process getting stuck as it waited
4405 for a non-existent subprocess. Forks do occasionally fail when resources
4406 run out. I reviewed all the other calls to fork(); they all seem to check
4409 PH/20 When checking for unexpected SMTP input at connect time (before writing
4410 the banner), Exim was not dealing correctly with a non-positive return
4411 from the read() function. If the client had disconnected by this time,
4412 the result was a log entry for a synchronization error with an empty
4413 string after "input=" when read() returned zero. If read() returned -1
4414 (an event I could not check), uninitialized data bytes were printed.
4415 There were reports of junk text (parts of files, etc) appearing after
4418 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4420 PH/22 Added support for macro redefinition, and (re)definition in between
4421 driver and ACL definitions.
4423 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4424 forgetting to use the resulting value; it was using the unexpanded value.
4426 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4427 hadn't been configured. The fix is from Juergen Kreileder, who
4428 understands it better than I do:
4430 "Here's what I see happening with three configured cyrus_sasl
4431 authenticators configured (plain, login, cram-md5):
4433 On startup auth_cyrus_sasl_init() gets called for each of these.
4434 This means three calls to sasl_listmech() without a specified mech_list.
4435 => SASL tests which mechs of all available mechs actually work
4436 => three warnings about OTP not working
4437 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4439 With the patch, sasl_listmech() also gets called three times. But now
4440 SASL's mech_list option is set to the server_mech specified in the the
4441 authenticator. Or in other words, the answer from sasl_listmech()
4442 gets limited to just the mech you're testing for (which is different
4444 => the return list contains just 'plain' or 'login', 'cram-md5' or
4445 nothing depending on the value of ob->server_mech.
4447 I've just tested the patch: Authentication still works fine,
4448 unavailable mechs specified in the exim configuration are still
4449 caught, and the auth.log warnings about OTP are gone."
4451 PH/25 When debugging is enabled, the contents of the command line are added
4452 to the debugging output, even when log_selector=+arguments is not
4455 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4456 answer is "GNU", and only if the return is "GNU/something" is the answer
4459 PH/27 $acl_verify_message is now set immediately after the failure of a
4460 verification in an ACL, and so is available in subsequent modifiers. In
4461 particular, the message can be preserved by coding like this:
4463 warn !verify = sender
4464 set acl_m0 = $acl_verify_message
4466 Previously, $acl_verify_message was set only while expanding "message"
4467 and "log_message" when a very denied access.
4469 PH/28 Modified OS/os.c-Linux with
4471 -#ifndef OS_LOAD_AVERAGE
4472 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4474 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4475 nomenclature these days.)
4477 PH/29 Installed patch from the Sieve maintainer that adds the options
4478 sieve_useraddress and sieve_subaddress to the redirect router.
4480 PH/30 In these circumstances:
4481 . Two addresses routed to the same list of hosts;
4482 . First host does not offer TLS;
4483 . First host accepts first address;
4484 . First host gives temporary error to second address;
4485 . Second host offers TLS and a TLS session is established;
4486 . Second host accepts second address.
4487 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4488 and peerdn, if requested) that were in fact used only for the second
4491 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4492 attention to any local part prefix or suffix that was matched by the
4493 router that accepted the address. It now behaves in the same way as it
4494 does for delivery: the affixes are removed from the local part unless
4495 rcpt_include_affixes is set on the transport.
4497 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4498 timeout during the DATA phase of an incoming message.
4500 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4501 applied a patch sanctioned by the Sieve maintainer.
4503 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4504 the uid or gid is negative. A case of a negative gid caused this to be
4505 noticed. The fix allows for either to be negative.
4507 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4508 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4511 PH/36 The change PH/12 above was broken. Fixed it.
4513 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4514 the grounds that routing the same address twice would always produce the
4515 same answer. This might have been true once, but it is certainly no
4516 longer true now. Routing a child address may depend on the previous
4517 routing that produced that child. Some complicated redirection strategies
4518 went wrong when messages had multiple recipients, and made Exim's
4519 behaviour dependent on the order in which the addresses were given.
4521 I have moved the duplicate checking until after the routing is complete.
4522 Exim scans the addresses that are assigned to local and remote
4523 transports, and removes any duplicates. This means that more work will be
4524 done, as duplicates will always all be routed, but duplicates are
4525 presumably rare, so I don't expect this is of any significance.
4527 For deliveries to pipes, files, and autoreplies, the duplicate checking
4528 still happens during the routing process, since they are not going to be
4531 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4532 It corrects a timeout issue with spamd. This is Ian's comment: "The
4533 background is that sometimes spamd either never reads data from a
4534 connection it has accepted, or it never writes response data. The exiscan
4535 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4536 blindly assumes that writes won't block so it may never time out."
4538 PH/39 Allow G after quota size as well as K and M.
4540 PH/40 The value set for $authenticated_id in an authenticator may not contain
4541 binary zeroes or newlines because the value is written to log lines and
4542 to spool files. There was no check on this. Now the value is run through
4543 the string_printing() function so that such characters are converted to
4544 printable escape sequences.
4546 PH/41 $message_linecount is a new variable that contains the total number of
4547 lines in the message. Compare $body_linecount, which is the count for the
4550 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4551 bounce and delay warning messages, except in certain special cases, which
4554 (a) An SMTP error message from a remote host;
4555 (b) A message specified in a :fail: redirection;
4556 (c) A message specified in a "fail" command in a system filter;
4557 (d) A message specified in a FAIL return from the queryprogram router;
4558 (e) A message specified by the cannot_route_message router option.
4560 In these cases only, Exim does include the error details in bounce and
4561 warning messages. There are also a few cases where bland messages such
4562 as "unrouteable address" or "local delivery error" are given.
4564 PH/43 $value is now also set for the "else" part of a ${run expansion.
4566 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4567 being worked on, but at least Exim now implements the latest version to
4570 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4571 process to complete was treated as a delivery failure, a timeout while
4572 writing the message to the pipe was logged, but erroneously treated as a
4573 successful delivery. Such timeouts include transport filter timeouts. For
4574 consistency with the overall process timeout, these timeouts are now
4575 treated as errors, giving rise to delivery failures by default. However,
4576 there is now a new Boolean option for the pipe transport called
4577 timeout_defer, which, if set TRUE, converts the failures into defers for
4578 both kinds of timeout. A transport filter timeout is now identified in
4581 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4582 systems where "make" and "gmake" are different, calling "gmake" at top
4583 level broke things. I've arranged for the value of $(MAKE) to be passed
4584 from the Makefile to this script so that it can call the same version of
4588 A note about Exim versions 4.44 and 4.50
4589 ----------------------------------------
4591 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4592 changes of various kinds. As a consequence, a big documentation update was
4593 needed. This delayed the release for rather longer than seemed good, especially
4594 in the light of a couple of (minor) security issues. Therefore, the changes
4595 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4596 release. So 4.44 and 4.50 are in effect two different branches that both start
4599 I have left the 4.50 change log unchanged; it contains all the changes since
4600 4.43. The change log for 4.44 is below; many of its items are identical to
4601 those for 4.50. This seems to be the most sensible way to preserve the
4602 historical information.
4608 1. Minor wording change to the doc/README.SIEVE file.
4610 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4611 computation of the current number of files was incorrect.
4613 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4614 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4617 4. Give more explanation in the error message when the command for a transport
4618 filter fails to execute.
4620 5. There are several places where Exim runs a non-Exim command in a
4621 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4622 was being done only for the command run by the queryprogram router. It is
4623 now done for all such subprocesses. The other cases are: ${run, transport
4624 filters, and the commands run by the lmtp and pipe transports.
4626 6. Added CONFIGURE_GROUP build-time option.
4628 7. Some older OS have a limit of 256 on the maximum number of file
4629 descriptors. Exim was using setrlimit() to set 1000 as a large value
4630 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4631 systems. I've change it so that if it can't get 1000, it tries for 256.
4633 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4634 was an oversight, and furthermore, ever since the addition of extra
4635 controls (e.g. 4.43/32), the checks on when to allow different forms of
4636 "control" were broken. There should now be diagnostics for all cases when a
4637 control that does not make sense is encountered.
4639 9. Added the /retain_sender option to "control=submission".
4641 10. $recipients is now available in the predata ACL (oversight).
4643 11. Tidy the search cache before the fork to do a delivery from a message
4644 received from the command line. Otherwise the child will trigger a lookup
4645 failure and thereby defer the delivery if it tries to use (for example) a
4646 cached ldap connection that the parent has called unbind on.
4648 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4649 of $address_data from the recipient verification was clobbered by the
4650 sender verification.
4652 13. The value of address_data from a sender verification is now available in
4653 $sender_address_data in subsequent conditions in the ACL statement.
4655 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4657 15. Added a new option "connect=<time>" to callout options, to set a different
4660 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4661 was its contents. (It was OK if the option was not defined at all.)
4663 17. A "Completed" log line is now written for messages that are removed from
4664 the spool by the -Mrm option.
4666 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4667 information about exactly what failed.
4669 19. Added -dd to debug only the daemon process.
4671 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4672 handles timeouts, both on the server side and network timeouts. Renamed the
4673 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4675 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4676 It is now set to "smtps".
4678 22. $host_address is now set to the target address during the checking of
4679 ignore_target_hosts.
4681 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4682 being passed; this would have caused $sender_host_name to have been used if
4683 matching the list had actually called for a host name (not very likely,
4684 since this list is usually IP addresses). A host name is now passed as
4687 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4688 code that specifies a non-restarting handler (typically sigaction() in
4689 modern systems) in an attempt to fix a rare and obscure crash bug.
4691 25. Narrowed the window for a race in the daemon that could cause it to ignore
4692 SIGCHLD signals. This is not a major problem, because they are used only to
4693 wake it up if nothing else does.
4695 26. A malformed maildirsize file could cause Exim to calculate negative values
4696 for the mailbox size or file count. Odd effects could occur as a result.
4697 The maildirsize information is now recalculated if the size or filecount
4700 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4701 support for a long time. Removed HAVE_SYS_VFS_H.
4703 28. Installed the latest version of exipick from John Jetmore.
4705 29. In an address list, if the pattern was not a regular expression, an empty
4706 subject address (from a bounce message) matched only if the pattern was an
4707 empty string. Non-empty patterns were not even tested. This was the wrong
4708 because it is perfectly reasonable to use an empty address as part of a
4709 database query. An empty address is now tested by patterns that are
4710 lookups. However, all the other forms of pattern expect the subject to
4711 contain a local part and a domain, and therefore, for them, an empty
4712 address still always fails if the pattern is not itself empty.
4714 30. Exim went into a mad DNS loop when attempting to do a callout where the
4715 host was specified on an smtp transport, and looking it up yielded more
4716 than one IP address.
4718 31. Re-factored the code for checking spool and log partition space into a
4719 function that finds that data and another that does the check. The former
4720 is then used to implement four new variables: $spool_space, $log_space,
4721 $spool_inodes, and $log_inodes.
4723 32. The RFC2047 encoding function was originally intended for short strings
4724 such as real names; it was not keeping to the 75-character limit for
4725 encoded words that the RFC imposes. It now respects the limit, and
4726 generates multiple encoded words if necessary. To be on the safe side, I
4727 have increased the buffer size for the ${rfc2047: expansion operator from
4730 33. It is now permitted to omit both strings after an "if" condition; if the
4731 condition is true, the result is "true". As before, when the second string
4732 is omitted, a false condition yields an empty string. This makes it less
4733 cumbersome to write custom ACL and router conditions.
4735 34. Failure to deliver a bounce message always caused it to be frozen, even if
4736 there was an errors_to setting on the router. The errors_to setting is now
4739 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4740 canonical form (fully expanded) before being placed in
4741 $sender_host_address.
4743 36. The table in the code that translates DNS record types into text (T_A to
4744 "A" for instance) was missing entries for NS and CNAME. It is just possible
4745 that this could have caused confusion if both these types were looked up
4746 for the same domain, because the text type is used as part of Exim's
4747 per-process caching. But the chance of anyone hitting this buglet seems
4750 37. The dnsdb lookup has been extended in a number of ways.
4752 (1) There is a new type, "zns", which walks up the domain tree until it
4753 finds some nameserver records. It should be used with care.
4755 (2) There is a new type, "mxh", which is like "mx" except that it returns
4756 just the host names, not the priorities.
4758 (3) It is now possible to give a list of domains (or IP addresses) to be
4759 looked up. The behaviour when one of the lookups defers can be
4760 controlled by a keyword.
4762 (4) It is now possible to specify the separator character for use when
4763 multiple records are returned.
4765 38. The dnslists ACL condition has been extended: it is now possible to supply
4766 a list of IP addresses and/or domains to be looked up in a particular DNS
4769 39. Added log_selector=+queue_time_overall.
4771 40. When running the queue in the test harness, wait just a tad after forking a
4772 delivery process, to get repeatability of debugging output.
4774 41. Include certificate and key file names in error message when GnuTLS fails
4775 to set them up, because the GnuTLS error message doesn't include the name
4776 of the failing file when there is a problem reading it.
4778 42. Allow both -bf and -bF in the same test run.
4780 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4782 44. The "Exiscan patch" is now merged into the mainline Exim source.
4784 45. Sometimes the final signoff response after QUIT could fail to get
4785 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4786 < 0 before doing a fflush(). This bug looks as though it goes back to the
4787 introduction of TLS in release 3.20, but "sometimes" must have been rare
4788 because the tests only now provoked it.
4790 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4791 (this can affect the format of dates).
4793 47. exim_tidydb, when checking for the continued existence of a message for
4794 which it has found a message-specific retry record, was not finding
4795 messages that were in split spool directories. Consequently, it was
4796 deleting retry records that should have stayed in existence.
4798 48. Steve fixed some bugs in eximstats.
4800 49. The SPA authentication driver was not abandoning authentication and moving
4801 on to the next authenticator when an expansion was forced to fail,
4802 contradicting the general specification for all authenticators. Instead it
4803 was generating a temporary error. It now behaves as specified.
4805 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4806 (the order specifies the preference for clients). The order is now AES256,
4807 AES128, 3DES, ARCFOUR128.
4809 51. Small patch to Sieve code - explicitly set From: when generating an
4812 52. Exim crashed if a remote delivery caused a very long error message to be
4813 recorded - for instance if somebody sent an entire SpamAssassin report back
4814 as a large number of 550 error lines. This bug was coincidentally fixed by
4815 increasing the size of one of Exim's internal buffers (big_buffer) that
4816 happened as part of the Exiscan merge. However, to be on the safe side, I
4817 have made the code more robust (and fixed the comments that describe what
4820 53. Now that there can be additional text after "Completed" in log lines (if
4821 the queue_time_overall log selector is set), a one-byte patch to exigrep
4822 was needed to allow it to recognize "Completed" as not the last thing in
4825 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4826 patch that reportedly fixes this has been added. I am not expert enough to
4827 create a test for it. This is what the patch creator wrote:
4829 "I found a little strange behaviour of ldap code when working with
4830 Windows 2003 AD Domain, where users was placed in more than one
4831 Organization Units. When I tried to give exim partial DN, the exit code
4832 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4833 But simultaneously result of request was absolutely normal ldap result,
4834 so I produce this patch..."
4836 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4837 REFERENCE, so I have modified the code to exclude the patch when that macro
4840 55. Some experimental protocols are using DNS PTR records for new purposes. The
4841 keys for these records are domain names, not reversed IP addresses. The
4842 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4843 leaves it alone. Component reversal etc. now happens only for IP addresses.
4846 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4848 57. Double the size of the debug message buffer (to 2048) so that more of very
4849 long debug lines gets shown.
4851 58. The exicyclog utility now does better if the number of log files to keep
4852 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4854 59. Two changes related to the smtp_active_hostname option:
4856 (1) $smtp_active_hostname is now available as a variable.
4857 (2) The default for smtp_banner uses $smtp_active_hostname instead
4858 of $primary_hostname.
4860 60. The host_aton() function is supposed to be passed a string that is known
4861 to be a valid IP address. However, in the case of IPv6 addresses, it was
4862 not checking this. This is a hostage to fortune. Exim now panics and dies
4863 if the condition is not met. A case was found where this could be provoked
4864 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4865 components; fortuitously, this particular loophole had already been fixed
4866 by change 4.50/55 above.
4868 If there are any other similar loopholes, the new check in host_aton()
4869 itself should stop them being exploited. The report I received stated that
4870 data on the command line could provoke the exploit when Exim was running as
4871 exim, but did not say which command line option was involved. All I could
4872 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4873 running as the user.
4876 61. There was a buffer overflow vulnerability in the SPA authentication code
4877 (which came originally from the Samba project). I have added a test to the
4878 spa_base64_to_bits() function which I hope fixes it.
4881 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4882 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4884 63. The daemon start-up calls getloadavg() while still root for those OS that
4885 need the first call to be done as root, but it missed one case: when
4886 deliver_queue_load_max is set with deliver_drop_privilege. This is
4887 necessary for the benefit of the queue runner, because there is no re-exec
4888 when deliver_drop_privilege is set.
4890 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4891 This has been fixed.
4893 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4894 was in use, was not putting the data itself into the right store pool;
4895 consequently, it could be overwritten for a subsequent message in the same
4896 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4899 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4901 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4902 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4903 the function were treating the return as a boolean value, which happened to
4904 work because 0=false and not-0=true, but is not correct code.
4906 68. The host_aton() function was not handling scoped IPv6 addresses (those
4907 with, for example, "%eth0" on the end) correctly.
4909 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4910 negated items (that is, ~something) in unsigned ints. Some compilers
4911 apparently mutter when there is no cast.
4913 70. If an address verification called from an ACL failed, and did not produce a
4914 user-specific message (i.e. there was only a "system" message), nothing was
4915 put in $acl_verify_message. In this situation, it now puts the system
4918 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4919 session; change 4.31/43 added the unwanted input to the log line - except
4920 that it did not do this in the start of session case. It now does.
4922 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4923 This is wrong and can cause the other end to generate a synchronization
4924 error if it is another Exim or anything else that does the synchronization
4925 check. A QUIT command is no longer sent after a timeout.
4927 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4928 during host lookups.
4930 74. The defer_ok option of callout verification was not working if it was used
4931 when verifying addresses in header lines, that is, for this case:
4933 verify = header_sender/callout=defer_ok
4935 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4936 those file descriptors could be used for SMTP connections. If anything
4937 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4938 could be sent to the SMTP client, causing chaos. The daemon now opens
4939 stdin, stdout, and stderr to /dev/null when it puts itself into the
4942 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4943 log by default. The user can override this with suitable Perl magic.
4945 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4946 the log message when discard triggers, was not working for the DATA ACL or
4947 for the non-SMTP ACL.
4949 78. Error message wording change in sieve.c.
4951 79. If smtp_accept_max_per_host was set, the number of connections could be
4952 restricted to fewer than expected, because the daemon was trying to set up
4953 a new connection before checking whether the processes handling previous
4954 connections had finished. The check for completed processes is now done
4955 earlier. On busy systems, this bug wouldn't be noticed because something
4956 else would have woken the daemon, and it would have reaped the completed
4959 80. If a message was submitted locally by a user whose login name contained one
4960 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4961 It caused a spool format error. I have fixed the spool reading code. A
4962 related problem was that the "from" clause in the Received: line became
4963 illegal because of the space(s). It is now covered by ${quote_local_part.
4965 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4968 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4970 83. Patch from Sieve maintainer to fix unterminated string problem in
4971 "vacation" handling.
4973 84. Some minor changes to the Linux configuration files to help with other
4974 OS variants using glibc.
4976 85. One more patch for Sieve to update vacation handling to latest spec.
4979 ----------------------------------------------------
4980 See the note above about the 4.44 and 4.50 releases.
4981 ----------------------------------------------------
4987 1. Change 4.43/35 introduced a bug that caused file counts to be
4988 incorrectly computed when quota_filecount was set in an appendfile
4991 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4992 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4995 3. Give more explanation in the error message when the command for a transport
4996 filter fails to execute.
4998 4. There are several places where Exim runs a non-Exim command in a
4999 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5000 was being done only for the command run by the queryprogram router. It is
5001 now done for all such subprocesses. The other cases are: ${run, transport
5002 filters, and the commands run by the lmtp and pipe transports.
5004 5. Some older OS have a limit of 256 on the maximum number of file
5005 descriptors. Exim was using setrlimit() to set 1000 as a large value
5006 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5007 systems. I've change it so that if it can't get 1000, it tries for 256.
5009 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5010 was an oversight, and furthermore, ever since the addition of extra
5011 controls (e.g. 4.43/32), the checks on when to allow different forms of
5012 "control" were broken. There should now be diagnostics for all cases when a
5013 control that does not make sense is encountered.
5015 7. $recipients is now available in the predata ACL (oversight).
5017 8. Tidy the search cache before the fork to do a delivery from a message
5018 received from the command line. Otherwise the child will trigger a lookup
5019 failure and thereby defer the delivery if it tries to use (for example) a
5020 cached ldap connection that the parent has called unbind on.
5022 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5023 of $address_data from the recipient verification was clobbered by the
5024 sender verification.
5026 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5027 was its contents. (It was OK if the option was not defined at all.)
5029 11. A "Completed" log line is now written for messages that are removed from
5030 the spool by the -Mrm option.
5032 12. $host_address is now set to the target address during the checking of
5033 ignore_target_hosts.
5035 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5036 being passed; this would have caused $sender_host_name to have been used if
5037 matching the list had actually called for a host name (not very likely,
5038 since this list is usually IP addresses). A host name is now passed as
5041 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5042 code that specifies a non-restarting handler (typically sigaction() in
5043 modern systems) in an attempt to fix a rare and obscure crash bug.
5045 15. Narrowed the window for a race in the daemon that could cause it to ignore
5046 SIGCHLD signals. This is not a major problem, because they are used only to
5047 wake it up if nothing else does.
5049 16. A malformed maildirsize file could cause Exim to calculate negative values
5050 for the mailbox size or file count. Odd effects could occur as a result.
5051 The maildirsize information is now recalculated if the size or filecount
5054 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5055 support for a long time. Removed HAVE_SYS_VFS_H.
5057 18. Updated exipick to current release from John Jetmore.
5059 19. Allow an empty sender to be matched against a lookup in an address list.
5060 Previously the only cases considered were a regular expression, or an
5063 20. Exim went into a mad DNS lookup loop when doing a callout where the
5064 host was specified on the transport, if the DNS lookup yielded more than
5067 21. The RFC2047 encoding function was originally intended for short strings
5068 such as real names; it was not keeping to the 75-character limit for
5069 encoded words that the RFC imposes. It now respects the limit, and
5070 generates multiple encoded words if necessary. To be on the safe side, I
5071 have increased the buffer size for the ${rfc2047: expansion operator from
5074 22. Failure to deliver a bounce message always caused it to be frozen, even if
5075 there was an errors_to setting on the router. The errors_to setting is now
5078 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5079 canonical form (fully expanded) before being placed in
5080 $sender_host_address.
5082 24. Updated eximstats to version 1.33
5084 25. Include certificate and key file names in error message when GnuTLS fails
5085 to set them up, because the GnuTLS error message doesn't include the name
5086 of the failing file when there is a problem reading it.
5088 26. Expand error message when OpenSSL has problems setting up cert/key files.
5091 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5092 (this can affect the format of dates).
5094 28. exim_tidydb, when checking for the continued existence of a message for
5095 which it has found a message-specific retry record, was not finding
5096 messages that were in split spool directories. Consequently, it was
5097 deleting retry records that should have stayed in existence.
5099 29. eximstats updated to version 1.35
5100 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5101 1.35 - bugfix such that pie charts by volume are generated correctly
5103 30. The SPA authentication driver was not abandoning authentication and moving
5104 on to the next authenticator when an expansion was forced to fail,
5105 contradicting the general specification for all authenticators. Instead it
5106 was generating a temporary error. It now behaves as specified.
5108 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5109 (the order specifies the preference for clients). The order is now AES256,
5110 AES128, 3DES, ARCFOUR128.
5112 31. Small patch to Sieve code - explicitly set From: when generating an
5115 32. Exim crashed if a remote delivery caused a very long error message to be
5116 recorded - for instance if somebody sent an entire SpamAssassin report back
5117 as a large number of 550 error lines. This bug was coincidentally fixed by
5118 increasing the size of one of Exim's internal buffers (big_buffer) that
5119 happened as part of the Exiscan merge. However, to be on the safe side, I
5120 have made the code more robust (and fixed the comments that describe what
5123 33. Some experimental protocols are using DNS PTR records for new purposes. The
5124 keys for these records are domain names, not reversed IP addresses. The
5125 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5126 leaves it alone. Component reversal etc. now happens only for IP addresses.
5129 34. The host_aton() function is supposed to be passed a string that is known
5130 to be a valid IP address. However, in the case of IPv6 addresses, it was
5131 not checking this. This is a hostage to fortune. Exim now panics and dies
5132 if the condition is not met. A case was found where this could be provoked
5133 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5134 components; fortuitously, this particular loophole had already been fixed
5135 by change 4.50/55 or 4.44/33 above.
5137 If there are any other similar loopholes, the new check in host_aton()
5138 itself should stop them being exploited. The report I received stated that
5139 data on the command line could provoke the exploit when Exim was running as
5140 exim, but did not say which command line option was involved. All I could
5141 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5142 running as the user.
5145 35. There was a buffer overflow vulnerability in the SPA authentication code
5146 (which came originally from the Samba project). I have added a test to the
5147 spa_base64_to_bits() function which I hope fixes it.
5150 36. The daemon start-up calls getloadavg() while still root for those OS that
5151 need the first call to be done as root, but it missed one case: when
5152 deliver_queue_load_max is set with deliver_drop_privilege. This is
5153 necessary for the benefit of the queue runner, because there is no re-exec
5154 when deliver_drop_privilege is set.
5156 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5157 was in use, was not putting the data itself into the right store pool;
5158 consequently, it could be overwritten for a subsequent message in the same
5159 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5162 38. Sometimes the final signoff response after QUIT could fail to get
5163 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5164 < 0 before doing a fflush(). This bug looks as though it goes back to the
5165 introduction of TLS in release 3.20, but "sometimes" must have been rare
5166 because the tests only now provoked it.
5172 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5173 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5174 Now it returns an integer. A number of calls were still expecting a T/F
5175 return. Fortuitously, in all cases, the tests worked in OK situations,
5176 which is the norm. However, things would have gone wrong on any write
5177 failures on the smtp file descriptor. This function is used when sending
5178 messages over SMTP and also when doing verify callouts.
5180 2. When Exim is called to do synchronous delivery of a locally submitted
5181 message (the -odf or -odi options), it no longer closes stderr before doing
5184 3. Implemented the mua_wrapper option.
5186 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5188 5. Implemented the functions header_remove(), header_testname(),
5189 header_add_at_position(), and receive_remove_recipient(), and exported them
5192 6. If an ACL "warn" statement specified the addition of headers, Exim already
5193 inserted X-ACL-Warn: at the start if there was no header name. However, it
5194 was not making this test for the second and subsequent header lines if
5195 there were newlines in the string. This meant that an invalid header could
5196 be inserted if Exim was badly configured.
5198 7. Allow an ACL "warn" statement to add header lines at the start or after all
5199 the Received: headers, as well as at the end.
5201 8. Added the rcpt_4xx retry error code.
5203 9. Added postmaster_mailfrom=xxx to callout verification option.
5205 10. Added mailfrom=xxxx to the callout verification option, for verify=
5208 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5209 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5211 12. Inserted some casts to stop certain compilers warning when using pointer
5212 differences as field lengths or precisions in printf-type calls (mostly
5213 affecting debugging statements).
5215 13. Added optional readline() support for -be (dynamically loaded).
5217 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5218 same clock tick as a message's arrival, so that its received time was the
5219 same as the "first fail" time on the retry record, and that message
5220 remained on the queue past the ultimate address timeout, every queue runner
5221 would try a delivery (because it was past the ultimate address timeout) but
5222 after another failure, the ultimate address timeout, which should have then
5223 bounced the address, did not kick in. This was a "< instead of <=" error;
5224 in most cases the first failure would have been in the next clock tick
5225 after the received time, and all would be well.
5227 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5228 being recognized when the domain list was tested by the match_domain
5229 condition in an expansion string.
5231 16. Added the ${str2b64: operator.
5233 17. Exim was always calling setrlimit() to set a large limit for the number of
5234 processes, without checking whether the existing limit was already
5235 adequate. (It did check for the limit on file descriptors.) Furthermore,
5236 errors from getrlimit() and setrlimit() were being ignored. Now they are
5237 logged to the main and panic logs, but Exim does carry on, to try to do its
5238 job under whatever limits there are.
5240 18. Imported PCRE 5.0.
5242 19. Trivial typo in log message " temporarily refused connection" (the leading
5245 20. If the log selector return_path_on_delivery was set and an address was
5246 redirected to /dev/null, the delivery process crashed because it assumed
5247 that a return path would always be set for a "successful" delivery. In this
5248 case, the whole delivery is bypassed as an optimization, and therefore no
5251 21. Internal re-arrangement: the function for sending a challenge and reading
5252 a response while authentication was assuming a zero-terminated challenge
5253 string. It's now changed to take a pointer and a length, to allow for
5254 binary data in such strings.
5256 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5258 23. Exim was not respecting finduser_retries when seeking the login of the
5259 uid under which it was called; it was always trying 10 times. (The default
5260 setting of finduser_retries is zero.) Also, it was sleeping after the final
5261 failure, which is pointless.
5263 24. Implemented tls_on_connect_ports.
5265 25. Implemented acl_smtp_predata.
5267 26. If the domain in control=submission is set empty, Exim assumes that the
5268 authenticated id is a complete email address when it generates From: or
5269 Sender: header lines.
5271 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5272 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5273 chown and chgrp in /bin and hostname in /usr/bin.
5275 28. Exim was keeping the "process log" file open after each use, just as it
5276 does for the main log. This opens the possibility of it remaining open for
5277 long periods when the USR1 signal hits a daemon. Occasional processlog
5278 errors were reported, that could have been caused by this. Anyway, it seems
5279 much more sensible not to leave this file open at all, so that is what now
5282 29. The long-running daemon process does not normally write to the log once it
5283 has entered its main loop, and it closes the log before doing so. This is
5284 so that log files can straightforwardly be renamed and moved. However,
5285 there are a couple of unusual error situations where the daemon does write
5286 log entries, and I had neglected to close the log afterwards.
5288 30. The text of an SMTP error response that was received during a remote
5289 delivery was being truncated at 512 bytes. This is too short for some of
5290 the long messages that one sometimes sees. I've increased the limit to
5293 31. It is now possible to make retry rules that apply only when a message has a
5294 specific sender, in particular, an empty sender.
5296 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5297 it possible to be selective about when SMTP synchronization is enforced.
5299 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5301 32. Implemented hosts_connection_nolog.
5303 33. Added an ACL for QUIT.
5305 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5308 35. Added mailbox_size and mailbox_filecount to appendfile.
5310 36. Added control = no_multiline_responses to ACLs.
5312 37. There was a bug in the logic of the code that waits for the clock to tick
5313 in the case where the clock went backwards by a substantial amount such
5314 that the microsecond fraction of "now" was more than the microsecond
5315 fraction of "then" (but the whole seconds number was less).
5317 38. Added support for the libradius Radius client library this is found on
5318 FreeBSD (previously only the radiusclient library was supported).
5324 1. When certain lookups returned multiple values in the form name=value, the
5325 quoting of the values was not always being done properly. Specifically:
5326 (a) If the value started with a double quote, but contained no whitespace,
5328 (b) If the value contained whitespace other than a space character (i.e.
5329 tabs or newlines or carriage returns) it was not quoted.
5330 This fix has been applied to the mysql and pgsql lookups by writing a
5331 separate quoting function and calling it from the lookup code. The fix
5332 should probably also be applied to nisplus, ibase and oracle lookups, but
5333 since I cannot test any of those, I have not disturbed their existing code.
5335 2. A hit in the callout cache for a specific address caused a log line with no
5336 reason for rejecting RCPT. Now it says "Previous (cached) callout
5337 verification failure".
5339 3. There was an off-by-one bug in the queryprogram router. An over-long
5340 return line was truncated at 256 instead of 255 characters, thereby
5341 overflowing its buffer with the terminating zero. As well as fixing this, I
5342 have increased the buffer size to 1024 (and made a note to document this).
5344 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5345 when Exim is waiting for an SMTP response from a remote server, Exim
5346 restarts its select() call on the socket, thereby resetting its timeout.
5347 This is not a problem when such interrupts are rare. Somebody set up a cron
5348 job to run exiwhat every 2 minutes, which is less than the normal select()
5349 timeout (5 or 10 minutes). This meant that the select() timeout never
5350 kicked in because it was always reset. I have fixed this by comparing the
5351 time when an interrupt arrives with the time at the start of the first call
5352 to select(). If more time than the timeout has elapsed, the interrupt is
5353 treated as a timeout.
5355 5. Some internal re-factoring in preparation for the addition of Sieve
5356 extensions (by MH). In particular, the "personal" test is moved to a
5357 separate function, and given an option for scanning Cc: and Bcc: (which is
5358 not set for Exim filters).
5360 6. When Exim created an email address using the login of the caller as the
5361 local part (e.g. when creating a From: or Sender: header line), it was not
5362 quoting the local part when it contained special characters such as @.
5364 7. Installed new OpenBSD configuration files.
5366 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5367 try to make them clearer.
5369 9. Callout options, other than the timeout value, were being ignored when
5370 verifying sender addresses in header lines. For example, when using
5372 verify = header_sender/callout=no_cache
5374 the cache was (incorrectly) being used.
5376 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5377 only the Cygwin environment.
5379 11. When return_path_on_delivery was set as a log selector, if different remote
5380 addresses in the same message used different return paths and parallel
5381 remote delivery occurred, the wrong values would sometimes be logged.
5382 (Whenever a remote delivery process finished, the return path value from
5383 the most recently started remote delivery process was logged.)
5385 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5386 lines when AUTH and/or TLS are in use. This is the "received protocol"
5387 field. Exim used to use "asmtp" for authenticated SMTP, without any
5388 indication (in the protocol name) for TLS use. Now it follows the RFC and
5389 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5390 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5391 names appear in log lines as well as in Received: header lines.
5393 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5394 extensions, and comparison tests, and to fix some bugs.
5396 14. Changes to the "personal" filter test:
5398 (1) The test was buggy in that it was just doing the equivalent of
5399 "contains" tests on header lines. For example, if a user's address was
5400 anne@some.where, the "personal" test would incorrectly be true for
5402 To: susanne@some.where
5404 This test is now done by extracting each address from the header in turn,
5405 and checking the entire address. Other tests that are part of "personal"
5406 are now done using regular expressions (for example, to check local parts
5407 of addresses in From: header lines).
5409 (2) The list of non-personal local parts in From: addresses has been
5410 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5411 taken from the Sieve specification recommendations.
5413 (3) If the message contains any header line starting with "List-" it is
5414 treated as non-personal.
5416 (4) The test for "circular" in the Subject: header line has been removed
5417 because it now seems ill-conceived.
5419 15. Minor typos in src/EDITME comments corrected.
5421 16. Installed latest exipick from John Jetmore.
5423 17. If headers_add on a router specified a text string that was too long for
5424 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5425 of string_sprintf() is now avoided.
5427 18. $message_body_size was not set (it was always zero) when running the DATA
5428 ACL and the local_scan() function.
5430 19. For the "mail" command in an Exim filter, no default was being set for
5431 the once_repeat time, causing a random time value to be used if "once" was
5432 specified. (If the value happened to be <= 0, no repeat happened.) The
5433 default is now 0s, meaning "never repeat". The "vacation" command was OK
5434 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5435 (I found it when inspecting the code).
5437 20. There is now an overall timeout for performing a callout verification. It
5438 defaults to 4 times the callout timeout, which applies to individual SMTP
5439 commands during the callout. The overall timeout applies when there is more
5440 than one host that can be tried. The timeout is checked before trying the
5441 next host. This prevents very long delays if there are a large number of
5442 hosts and all are timing out (e.g. when the network connections are timing
5443 out). The value of the overall timeout can be changed by specifying an
5444 additional sub-option for "callout", called "maxwait". For example:
5446 verify = sender/callout=5s,maxwait=20s
5448 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5449 to the end before writing, but this should make it even safer).
5451 22. Exim was forgetting that it had advertised PIPELINING for the second and
5452 subsequent messages on an SMTP connection. It was also not resetting its
5453 memory on STARTTLS and an internal HELO.
5455 23. When Exim logs an SMTP synchronization error within a session, it now
5456 records whether PIPELINING has been advertised or not.
5458 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5459 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5460 rather than long int.
5462 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5464 26. Added the never_mail option to autoreply.
5470 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5471 crash if the getsockname() call failed; this can happen if a connection is
5472 closed very soon after it is established. The problem was simply in the
5473 order in which certain operations were done, causing Exim to try to write
5474 to the SMTP stream before it had set up the file descriptor. The bug has
5475 been fixed by making things happen in the correct order.
5481 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5482 before the connection was closed, thus losing the rejection response.
5484 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5485 some early Solaris releases, but causes trouble in current releases where
5486 socklen_t is defined.
5488 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5491 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5494 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5495 that releasing the top of it at the end releases what was used for sub-
5496 expansions (unless the block got too big). However, discard this block if
5497 the first thing is a variable or header, so that we can use its block when
5498 it is dynamic (useful for very large $message_headers, for example).
5500 6. Lookups now cache *every* query, not just the most recent. A new, separate
5501 store pool is used for this. It can be recovered when all lookup caches are
5502 flushed. Lookups now release memory at the end of their result strings.
5503 This has involved some general refactoring of the lookup sources.
5505 7. Some code has been added to the store_xxx() functions to reduce the amount
5506 of flapping under certain conditions.
5508 8. log_incoming_interface used to affect only the <= reception log lines. Now
5509 it causes the local interface and port to be added to several more SMTP log
5510 lines, for example "SMTP connection from", and rejection lines.
5512 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5514 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5516 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5517 could be overwritten at the end of the current message (or the start of a
5518 new message if it was set in a HELO ACL). The value is now preserved for
5519 the duration of the SMTP connection.
5521 12. If a transport had a headers_rewrite setting, and a matching header line
5522 contained an unqualified address, that address was qualified, even if it
5523 did not match any rewriting rules. The underlying bug was that the values
5524 of the flags that permit the existence of unqualified sender and recipient
5525 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5526 non-local messages, and by -bnq for local messages) were not being
5527 preserved with the message after it was received.
5529 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5530 "next input=" as part of the text comprising the host identity instead of
5531 the correct text. The code was using the same buffer for two different
5532 strings. However, depending on which order the printing function evaluated
5533 its arguments, the bug did not always show up. Under Linux, for example, my
5534 test suite worked just fine.
5536 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5537 doesn't work with some older versions of Perl. It has been changed to "my",
5538 which in any case is probably the better facility to use.
5540 15. A really picky compiler found some instances of statements for creating
5541 error messages that either had too many or two few arguments for the format
5544 16. The size of the buffer for calls to the DNS resolver has been increased
5545 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5546 for addresses that have a lot of PTR records. This alleviates a problem; it
5547 does not fully solve it.
5549 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5550 buffer now truncates the list and logs the incident, which is the same
5551 action as happens when Exim is looking up a host name and its aliases.
5552 Previously in this situation something unpredictable would happen;
5553 sometimes it was "internal error: store_reset failed".
5555 18. If a server dropped the connection unexpectedly when an Exim client was
5556 using GnuTLS and trying to read a response, the client delivery process
5557 crashed while trying to generate an error log message.
5559 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5560 string, for example:
5562 warn message = H1: something\nH2: something
5564 the text was added as a single header line from Exim's point of view
5565 though it ended up OK in the delivered message. However, searching for the
5566 second and subsequent header lines using $h_h2: did not work. This has been
5567 fixed. Similarly, if a system filter added multiple headers in this way,
5568 the routers could not see them.
5570 20. Expanded the error message when iplsearch is called with an invalid key to
5571 suggest using net-iplsearch in a host list.
5573 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5574 ACLs are no longer actually imposed (and a message to that effect is
5577 22. If a "gecos" field in a passwd entry contained escaped characters, in
5578 particular, if it contained a \" sequence, Exim got it wrong when building
5579 a From: or a Sender: header from that name. A second bug also caused
5580 incorrect handling when an unquoted " was present following a character
5581 that needed quoting.
5583 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5584 was not being matched caselessly.
5586 24. Arranged for all hyphens in the exim.8 source to be escaped with
5589 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5590 properly. Recipient callout cache records were still being keyed to include
5591 the sender, even when use_sender was set false. This led to far more
5592 callouts that were necessary. The sender is no longer included in the key
5593 when use_sender is false.
5595 26. Added "control = submission" modifier to ACLs.
5597 27. Added the ${base62d: operator to decode base 62 numbers.
5599 28. dnsdb lookups can now access SRV records.
5601 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5602 the configuration file.
5604 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5605 (-v) mode. This makes the output for a verbose queue run more intelligible.
5607 31. Added a use_postmaster feature to recipient callouts.
5609 32. Added the $body_zerocount variable, containing the number of binary zero
5610 bytes in the message body.
5612 33. The time of last modification of the "new" subdirectory is now used as the
5613 "mailbox time last read" when there is a quota error for a maildir
5616 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5618 35. Added +ignore_unknown as a special item in host lists.
5620 36. Code for decoding IPv6 addresses in host lists is now included, even if
5621 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5622 address was recognized as an IP address, but was then not correctly decoded
5623 into binary, causing unexpected and incorrect effects when compared with
5630 1. Very minor rewording of debugging text in manualroute to say "list of
5631 hosts" instead of "hostlist".
5633 2. If verify=header_syntax was set, and a header line with an unqualified
5634 address (no domain) and a large number of spaces between the end of the
5635 name and the colon was received, the reception process suffered a buffer
5636 overflow, and (when I tested it) crashed. This was caused by some obsolete
5637 code that should have been removed. The fix is to remove it!
5639 3. When running in the test harness, delay a bit after writing a bounce
5640 message to get a bit more predictability in the log output.
5642 4. Added a call to search_tidyup() just before forking a reception process. In
5643 theory, someone could use a lookup in the expansion of smtp_accept_max_
5644 per_host which, without the tidyup, could leave open a database connection.
5646 5. Added the variables $recipient_data and $sender_data which get set from a
5647 lookup success in an ACL "recipients" or "senders" condition, or a router
5648 "senders" option, similar to $domain_data and $local_part_data.
5650 6. Moved the writing of debug_print from before to after the "senders" test
5653 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5654 problems for message scanning, either using a data ACL, or using
5655 local_scan() because the Received: header was not generated till after they
5656 were called (in order to set the time as the time of reception completion).
5657 I have revised the way this works. The header is now generated after the
5658 body is received, but before the ACL or local_scan() are called. After they
5659 are run, the timestamp in the header is updated.
5665 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5666 before starting a queue runner without re-exec. This happened only when
5667 deliver_drop_privilege was set or when the Exim user was set to root. The
5668 effect of the bug was that timeouts during subsequent deliveries caused
5669 crashes instead of being properly handled. The handler is now left at its
5670 default (and expected) setting.
5672 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5673 message, again when deliver_drop_privilege is set or Exim is run as root.
5674 The bug described in (1) was not present in this case, but the tidying up
5675 of the other signals was missing. I have made the two cases consistent.
5677 3. The ignore_target_hosts setting on a manualroute router was being ignored
5678 for hosts that were looked up using the /MX notation.
5680 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5683 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5684 operated on the sender address. After changing the $sender_address to <>
5685 for the sender address verify, Exim was re-instated it as the original
5686 (before rewriting) address, but remembering that it had rewritten it, so it
5687 wasn't rewriting it again. This bug also had the effect of breaking the
5688 sender address verification caching when the sender address was rewritten.
5690 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5691 This has been changed so that if the ip literal address matches
5692 ignore_target_hosts, the router declines.
5694 7. Added expansion conditions match_domain, match_address, and match_local_
5695 part (NOT match_host).
5697 8. The placeholder for the Received: header didn't have a length field set.
5699 9. Added code to Exim itself and to exim_lock to test for a specific race
5700 condition that could lead to file corruption when using MBX delivery. The
5701 issue is with the lockfile that is created in /tmp. If this file is removed
5702 after a process has opened it but before that process has acquired a lock,
5703 there is the potential for a second process to recreate the file and also
5704 acquire a lock. This could lead to two Exim processes writing to the file
5705 at the same time. The added code performs the same test as UW imapd; it
5706 checks after acquiring the lock that its file descriptor still refers to
5707 the same named file.
5709 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5710 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5713 11. Added the smtp_active_hostname option. If used, this will typically be made
5714 to depend on the incoming interface address. Because $interface_address is
5715 not set up until the daemon has forked a reception process, error responses
5716 that can happen earlier (such as "too many connections") no longer contain
5719 12. If an expansion in a condition on a "warn" statement fails because a lookup
5720 defers, the "warn" statement is abandoned, and the next ACL statement is
5721 processed. Previously this caused the whole ACL to be aborted.
5723 13. Added the iplsearch lookup type.
5725 14. Added ident_timeout as a log selector.
5727 15. Added tls_certificate_verified as a log selector.
5729 16. Added a global option tls_require_ciphers (compare the smtp transport
5730 option of the same name). This controls incoming TLS connections.
5732 17. I finally figured out how to make tls_require_ciphers do a similar thing
5733 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5734 before starting the TLS session.
5736 18. Tabs are now shown as \t in -bP output.
5738 19. If the log selector return_path_on_delivery was set, Exim crashed when
5739 bouncing a message because it had too many Received: header lines.
5741 20. If two routers both had headers_remove settings, and the first one included
5742 a superfluous trailing colon, the final name in the first list and the
5743 first name in the second list were incorrectly joined into one item (with a
5744 colon in the middle).
5750 1. Added -C and -D options to the exinext utility, mainly to make it easier
5751 to include in the automated testing, but these could be helpful when
5752 multiple configurations are in use.
5754 2. The exinext utility was not formatting the output nicely when there was
5755 an alternate port involved in the retry record key, nor when there was a
5756 message id as well (for retries that were specific to a specific message
5757 and a specific host). It was also confused by IPv6 addresses, because of
5758 the additional colons they contain. I have fixed the IPv4 problem, and
5759 patched it up to do a reasonable job for IPv6.
5761 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5762 delivery, the log line now contains "pipelined" if PIPELINING was used.
5764 4. An SMTP transport process used to panic and die if the bind() call to set
5765 an explicit outgoing interface failed. This has been changed; it is now
5766 treated in the same way as a connect() failure.
5768 5. A reference to $sender_host_name in the part of a conditional expansion
5769 that was being skipped was still causing a DNS lookup. This no longer
5772 6. The def: expansion condition was not recognizing references to header lines
5773 that used bh_ and bheader_.
5775 7. Added the _cache feature to named lists.
5777 8. The code for checking quota_filecount in the appendfile transport was
5778 allowing one more file than it should have been.
5780 9. For compatibility with Sendmail, the command line option
5788 and sets the incoming protocol and host name (for trusted callers). The
5789 host name and its colon can be omitted when only the protocol is to be set.
5790 Note the Exim already has two private options, -pd and -ps, that refer to
5791 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5792 "s", but I don't think that's a major issue.
5794 10. A number of refactoring changes to the code, none of which should affect
5797 (a) The number of logging options was getting close to filling up the
5798 32-bit word that was used as a bit map. I have split them into two classes:
5799 those that are passed in the argument to log_write(), and those that are
5800 only ever tested independently outside of that function. These are now in
5801 separate 32-bit words, so there is plenty of room for expansion again.
5802 There is no change in the user interface or the logging behaviour.
5804 (b) When building, for example, log lines, the code previously used a
5805 macro that called string_cat() twice, in order to add two strings. This is
5806 not really sufficiently general. Furthermore, there was one instance where
5807 it was actually wrong because one of the argument was used twice, and in
5808 one call a function was used. (As it happened, calling the function twice
5809 did not affect the overall behaviour.) The macro has been replaced by a
5810 function that can join an arbitrary number of extra strings onto a growing
5813 (c) The code for expansion conditions now uses a table and a binary chop
5814 instead of a serial search (which was left over from when there were very
5815 few conditions). Also, it now recognizes conditions like "pam" even when
5816 the relevant support is not compiled in: a suitably worded error message is
5817 given if an attempt is made to use such a condition.
5819 11. Added ${time_interval:xxxxx}.
5821 12. A bug was causing one of the ddress fields not to be passed back correctly
5822 from remote delivery subprocesses. The field in question was not being
5823 subsequently used, so this caused to problems in practice.
5825 13. Added new log selectors queue_time and deliver_time.
5827 14. Might have fixed a bug in maildirsizefile handling that threw up
5828 "unexpected character" debug warnings, and recalculated the data
5829 unnecessarily. In any case, I expanded the warning message to give more
5832 15. Added the message "Restricted characters in address" to the statements in
5833 the default ACL that block characters like @ and % in local parts.
5835 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5836 Three changes have been made:
5838 (a) There was a serious bug; a negative response to MAIL caused the whole
5839 recipient domain to be cached as invalid, thereby blocking all messages
5840 to all local parts at the same domain, from all senders. This bug has
5841 been fixed. The domain is no longer cached after a negative response to
5842 MAIL if the sender used is not empty.
5844 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5847 (c) A new callout option, "use_sender" has been added for people who want
5848 the modified behaviour.
5854 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5857 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5858 indeed breaks things for older releases.
5860 3. Added additional logging to the case where there is a problem reading data
5861 from a filter that is running in a subprocess using a pipe, in order to
5862 try to track down a specific problem.
5864 4. Testing facility fudge: when running in the test harness and attempting
5865 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5866 getting "No route to host". Convert this to a timeout.
5868 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5871 6. Some OS don't have socklen_t but use size_t instead. This affects the
5872 fifth argument of getsockopt() amongst other things. This is now
5873 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5874 can be set for individual OS. I have set it for SunOS5, OSF1, and
5875 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5876 some earlier ones do not.
5878 7. Change 4.30/15 was not doing the test caselessly.
5880 8. The standard form for an IPv6 address literal was being rejected by address
5881 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5882 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5883 this, as well as the form without the "IPv6" on the front (but only when
5884 address literals are enabled, of course).
5886 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5888 10. Exim crashed if a message with an empty sender address specified by -f
5889 encountered a router with an errors_to setting. This could be provoked only
5890 by a command such as
5894 where an empty string was supplied; "<>" did not hit this bug.
5896 11. Installed PCRE release 4.5.
5898 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5899 remained set. It is now erased.
5901 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5902 times from message ids (which are base 36 rather than the normal 62).
5904 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5905 were being counted as actual protocol errors, and logged if the log
5906 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5907 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5908 and DATA following a set of rejected RCPTs do not count as protocol errors.
5909 In other words, Exim assumes they were pipelined, though this may not
5910 actually be the case. Of course, in all cases the client gets an
5911 appropriate error code.
5913 15. If a lookup fails in an ACL condition, a message about the failure may
5914 be available; it is used if testing the ACL cannot continue, because most
5915 such messages specify what the cause of the deferral is. However, some
5916 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5917 that caused an old message to be retained and used if a later statement
5918 caused a defer, replacing the real cause of the deferral.
5920 16. If an IP address had so many PTR records that the DNS lookup buffer
5921 was not large enough to hold them, Exim could crash while trying to process
5922 the truncated data. It now detects and logs this case.
5924 17. Further to 4.21/58, another change has been made: if (and only if) the
5925 first line of a message (the first header line) ends with CRLF, a bare LF
5926 in a subsequent header line has a space inserted after it, so as not to
5927 terminate the header.
5929 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5930 unnecessarily, used atoi() instead of strtol(), and didn't check the
5931 termination when getting file sizes from file names by regex.
5933 19. Completely re-implemented the support for maildirsize files, in the light
5934 of a number of problems with the previous contributed implementation
5935 (4.30/29). In particular:
5937 . If the quota is zero, the maildirsize file is maintained, but no quota is
5940 . If the maildir directory does not exist, it is created before any attempt
5941 to write a maildirsize file.
5943 . The quota value in the file is just a cache; if the quota is changed in
5944 the transport, the new value overrides.
5946 . A regular expression is available for excluding directories from the
5949 20. The autoreply transport checks the characters in options that define the
5950 message's headers; it allows continued headers, but it was checking with
5951 isspace() after an embedded newline instead of explicitly looking for a
5954 21. If all the "regular" hosts to which an address was routed had passed their
5955 expiry times, and had not reached their retry times, the address was
5956 bounced, even if fallback hosts were defined. Now Exim should go on to try
5959 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5960 equivalent code in the SMTP transport. Some hosts send humungous responses
5961 to HELO/EHLO, more than 1024 it seems.
5963 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5964 gives compiler warnings in some environments. I've now done it "properly",
5967 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5968 (because of problems with the built-in one) was declared to return uschar *
5969 instead of char *, causing compiler failure.
5971 25. Fixed a file descriptor leak when processing alias/forward files.
5973 26. Fixed a minor format string issue in dbfn.c.
5975 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5977 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5978 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5981 29. When a sender address is verified, it is cached, to save repeating the test
5982 when there is more than one recipient in a message. However, when the
5983 verification involves a callout, it is possible for different callout
5984 options to be set for different recipients. It is too complicated to keep
5985 track of this in the cache, so now Exim always runs a verification when a
5986 callout is required, relying on the callout cache for the optimization.
5987 The overhead is duplication of the address routing, but this should not be
5990 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5991 to be verified with callout=postmaster, and the main callout worked but the
5992 postmaster check failed, the verification correctly failed. However, if a
5993 subsequent RCPT command asked for sender verification *without* the
5994 postmaster check, incorrect caching caused this verification also to fail,
5997 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5998 it was not caching the DNS options (qualify_single, search_parents) that
5999 were used when the lookup failed. A subsequent lookup with different
6000 options therefore always gave the same answer, though there were cases
6001 where it should not have. (Example: a "domains = !$mx_any" option on a
6002 dnslookup router: the "domains" option is always processed without any
6003 widening, but the router might have qualify_single set.) Now Exim uses the
6004 cached value only when the same options are set.
6006 32. Added John Jetmore's "exipick" utility to the distribution.
6008 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6009 than a timeout (e.g. a certificate is required, and is not provided), an
6010 Exim server now closes the connection immediately. Previously it waited for
6011 the client to close - but if the client is SSL, it seems that they each
6012 wait for each other, leading to a delay before one of them times out.
6014 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6015 maintained 0.8.x compatibility because I don't think many are using it, and
6016 it is clearly obsolete.
6018 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6021 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6022 client certificate was expired. A simple patch fixes this, though I don't
6023 understand the full logic of why the verify callback is called multiple
6026 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6027 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6028 which causes problems with some clients (such as the Certicom SSL Plus
6029 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6030 disables the coutermeasure allowing Eudora to connect."
6032 38. Exim was not checking that a write() to a log file succeeded. This could
6033 lead to Bad Things if a log got too big, in particular if it hit a file
6034 size limit. Exim now panics and dies if it cannot write to a log file, just
6035 as it does if it cannot open a log file.
6037 39. Modified OS/Makefile-Linux so that it now contains
6039 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6041 The two -D definitions ensure that Exim is compiled with large file
6042 support, which makes it possible to handle log files that are bigger than
6045 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6046 instance) a domain was checked against a named list that involved a lookup,
6047 causing $domain_data to be set, then another domain was checked against the
6048 same list, then the first domain was re-checked, the value of $domain_data
6049 after the final check could be wrong. In particular, if the second check
6050 failed, it could be set empty. This bug probably also applied to
6053 41. The strip_trailing_dot option was not being applied to the address given
6054 with the -f command-line option.
6056 42. The code for reading a message's header from the spool was incrementing
6057 $received_count, but never initializing it. This meant that the value was
6058 incorrect (doubled) while delivering a message in the same process in which
6059 it was received. In the most common configuration of Exim, this never
6060 happens - a fresh exec is done - but it can happen when
6061 deliver_drop_privilege is set.
6063 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6064 it now includes up to 150 characters of the unexpected data in the log
6067 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6068 and building data strings. The size of both of these buffers was 10 000
6069 bytes - far larger than anybody would *ever* want, thought I. Needless to
6070 say, somebody hit the limit. I have increased the maximum line length to
6071 20 000 and the maximum data length of concatenated lines to 100 000. I have
6072 also fixed two bugs, because there was no checking on these buffers. Tsk,
6073 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6074 buffer is too small.
6076 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6077 lsearch lookups. Now it does.
6079 46. When parsing a route_list item in a manualroute router, a fixed-length
6080 buffer was used for the list of hosts. I made this 1024 bytes long,
6081 thinking that nobody would ever have a list of hosts that long. Wrong.
6082 Somebody had a whole pile of complicated expansion conditions, and the
6083 string was silently truncated, leading to an expansion error. It turns out
6084 that it is easier to change to an unlimited length (owing to other changes
6085 that have happened since this code was originally written) than to build
6086 structure for giving a limitation error. The length of the item that
6087 expands into the list of hosts is now unlimited.
6089 47. The lsearch lookup could not handle data where the length of text line was
6090 more than 4095 characters. Such lines were truncated, leading to shortened
6091 data being returned. It should now handle lines of any length.
6093 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6094 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6097 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6098 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6099 no longer happens in comment lines. A list of replacements is now placed
6100 at the head of all of the source files, except those whose only change is
6101 to replace PERL_COMMAND in the very first #! line.
6103 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6104 messages on the queue, with a bottom-up merge sort, using code contributed
6105 by Michael Haardt. This should make operations like -bp somewhat faster on
6106 large queues. It won't affect queue runners, except when queue_run_in_order
6109 51. Installed eximstats 1.31 in the distribution.
6111 52. Added support for SRV lookups to the dnslookup router.
6113 53. If an ACL referred to $message_body or $message_body_end, the value was not
6114 reset for any messages that followed in the same SMTP session.
6116 54. The store-handling optimization for building very long strings was not
6117 differentiating between the different store pools. I don't think this
6118 actually made any difference in practice, but I've tidied it.
6120 55. While running the routers to verify a sender address, $sender_address
6121 was still set to the sender address. This is wrong, because when routing to
6122 send a bounce to the sender, it would be empty. Therefore, I have changed
6123 it so that, while verifying a sender address, $sender_address is set to <>.
6124 (There is no change to what happens when verifying a recipient address.)
6126 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6127 target A or AAAA records (if not already returned) without resetting the
6128 qualify_single or search_parents options of the DNS resolver. These are
6129 inappropriate in this case because the targets of MX and SRV records must
6130 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6131 target that, when qualified, matched something in the local domain. These
6132 two options are now turned off when doing these lookups.
6134 57. It seems that at least some releases of Reiserfs (which does not have the
6135 concept of a fixed number of inodes) returns zero and not -1 for the
6136 number of available inodes. This interacted badly with check_spool_inodes,
6137 which assumed that -1 was the "no such thing" setting. What I have done is
6138 to check that the total number of inodes is greater than zero before doing
6139 the test of how many are available.
6141 58. When a "warn" ACL statement has a log_message modifier, the message is
6142 remembered, and not repeated. This is to avoid a lot of repetition when a
6143 message has many recipients that cause the same warning to be written.
6144 However, Exim was preserving the list of already written lines for an
6145 entire SMTP session, which doesn't seem right. The memory is now reset if a
6146 new message is started.
6148 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6149 debugging output unless log_rewrite was also set.
6151 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6152 of (int)(handle) when we know that handle contains (void *)(-1).
6154 61. The Exim daemon panic-logs an error return when it closes the incoming
6155 connection. However "connection reset by peer" seems to be common, and
6156 isn't really an error worthy of noting specially, so that particular error
6159 62. When Exim is trying to find all the local interfaces, it used to panic and
6160 die if the ioctl to get the interface flags failed. However, it seems that
6161 on at least one OS (Solaris 9) it is possible to have an interface that is
6162 included in the list of interfaces, but for which you get a failure error
6163 for this call. This happens when the interface is not "plumbed" into a
6164 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6165 failure of the "get flags" call assumes that the interface is down.
6167 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6168 makes life easier for people who are doing arithmetic on fields extracted
6169 from dates, where you often get leading zeros that should not be
6170 interpreted as octal.
6172 64. Added qualify_domain to the redirect router, to override the global
6175 65. If a pathologically long header line contained very many addresses (the
6176 report of this problem mentioned 10 000) and each of them was rewritten,
6177 Exim could use up a very large amount of memory. (It kept on making new
6178 copies of the header line as it rewrote, and never released the old ones.)
6179 At the expense of a bit more processing, the header rewriting function has
6180 been changed so that it no longer eats memory in this way.
6182 66. The generation of the Received: header has been moved from the time that a
6183 message starts to be received, to the time that it finishes. The timestamp
6184 in the Received: header should now be very close to that of the <= log
6185 line. There are two side-effects of this change:
6187 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6188 logged header lines no longer include the local Received: line, because
6189 it has not yet been created. The same applies to a copy of the message
6190 that is returned to a non-SMTP sender when a message is rejected.
6192 (b) When a filter file is tested using -bf, no additional Received: header
6193 is added to the test message. After some thought, I decided that this
6196 This change does not affect the value of $received_for. It is still set
6197 after address rewriting, but before local_scan() is called.
6199 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6201 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6202 gave an unhelpful panic error message, and a defer error. I have managed to
6203 change this behaviour so that it now rejects any supplied certificate,
6204 which seems right, as the list of acceptable certificates is empty.
6206 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6207 gave an unhelpful defer error. I have not managed to make this reject any
6208 supplied certificates, but the error message it gives is "no certificate
6209 supplied", which is not helpful.
6211 70. exigrep's output now also includes lines that are not associated with any
6212 message, but which match the given pattern. Implemented by a patch from
6213 Martin Sluka, which also tidied up the Perl a bit.
6215 71. Recipient callout verification, like sender verification, was using <> in
6216 the MAIL FROM command. This isn't really the right thing, since the actual
6217 sender may affect whether the remote host accepts the recipient or not. I
6218 have changed it to use the actual sender in the callout; this means that
6219 the cache record is now keyed on a recipient/sender pair, not just the
6220 recipient address. There doesn't seem to be a real danger of callout loops,
6221 since a callout by the remote host to check the sender would use <>.
6222 [SEE ABOVE: changed after hitting problems.]
6224 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6225 temporary errors. However, in the case of such a code being given after
6226 the end of a data transmission (i.e. after ".") Exim was failing to write
6227 a retry record for the message. (Yes, there was some broken host that was
6228 actually sending 8xx at this point.)
6230 73. An unknown lookup type in a host list could cause Exim to panic-die when
6231 the list was checked. (An example that provoked this was putting <; in the
6232 middle of a list instead of at the start.) If this happened during a DATA
6233 ACL check, a -D file could be left lying around. This kind of configuration
6234 error no longer causes Exim to die; instead it causes a defer error. The
6235 incident is still logged to the main and panic logs.
6237 74. Buglet left over from Exim 3 conversion. The message "too many messages
6238 in one connection" was written to the rejectlog but not the mainlog, except
6239 when address rewriting (yes!) was being logged.
6241 75. Added write_rejectlog option.
6243 76. When a system filter was run not as root (that is, when system_filter_user
6244 was set), the values of the $n variables were not being returned to the
6245 main process; thus, they were not subsequently available in the $sn
6248 77. Added +return_path_on_delivery log selector.
6250 78. A connection timeout was being treated differently from recipients deferred
6251 when testing hosts_max_try with a message that was older than the host's
6252 retry timeout. (The host should not be counted, thus allowing all hosts to
6253 be tried at least once before bouncing.) This may have been the cause of an
6254 occasionally reported bug whereby a message would remain on the queue
6255 longer than the retry timeout, but would be bounced if a delivery was
6256 forced. I say "may" because I never totally pinned down the problem;
6257 setting up timeout/retry tests is difficult. See also the next item.
6259 79. The ultimate address timeout was not being applied to errors that involved
6260 a combination of host plus message (for example, a timeout on a MAIL
6261 command). When an address resolved to a number of possible hosts, and they
6262 were not all tried for each delivery (e.g. because of hosts_max_try), a
6263 message could remain on the queue longer than the retry timeout.
6265 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6268 81. Fixed an obscure SMTP outgoing bug which required at least the following
6269 conditions: (a) there was another message waiting for the same server;
6270 (b) the server returned 5xx to all RCPT commands in the first message so
6271 that the message was not completed; (c) the server dropped the connection
6272 or gave a negative response to the RSET that Exim sends to abort the
6273 transaction. The observed case was a dropped connection after DATA that had
6274 been sent in pipelining mode. That is, the server had advertised PIPELINING
6275 but was not implementing it correctly. The effect of the bug was incorrect
6276 behaviour, such as trying another host, and this could lead to a crash.
6282 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6283 and daemon.c were passed as pointers to ints; they should have been
6284 pointers to socklen_t variables (which are typically unsigned ints).
6286 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6289 3. Fixed a really odd bug that affected only the testing scheme; patching a
6290 certain fixed string in the binary changed the value of another string that
6291 happened to be identical to the end of the original first string.
6293 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6294 name", it returns that address as the IP address. On some operating
6295 systems (e.g. Solaris), it also passes back the IP address string as the
6296 "host name". However, on others (e.g. Linux), it passes back an empty
6297 string. Exim wasn't checking for this, and was changing the host name to an
6298 empty string, assuming it had been canonicalized.
6300 5. Although rare, it is permitted to have more than one PTR record for a given
6301 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6302 all the names associated with an address, because they do in Solaris.
6303 However, it seems that they do not in Linux for data that comes from the
6304 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6305 I found this out when I moved to a new Linux workstation and tried to run
6306 the Exim test suite.
6308 To get round this problem I have changed the code so that it now does its
6309 own call to the DNS to look up PTR records when searching for a host name.
6310 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6311 addresses that are only in /etc/hosts are still found.
6313 This behaviour is, however, controlled by an option called host_lookup_
6314 order, which defaults to "bydns:byaddr". If people want to use the other
6315 order, or indeed, just use one or the other means of lookup, they can
6316 specify it in this variable.
6318 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6319 some operating systems, this comes back from gethostbyaddr() as an empty
6320 string, and this is what Exim used to test for. However, it seems that in
6321 other systems, "." is yielded. Exim now tests for this case too.
6323 7. The values of check_spool_space and check_log_space are now held internally
6324 as a number of kilobytes instead of an absolute number of bytes. If a
6325 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6326 kilobyte. This means that much larger values can be stored.
6328 8. Exim monitor: an attempt to get the action menu when not actually pointing
6329 at a message produces an empty menu entitled "No message selected". This
6330 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6331 no entries in it ("Shell widget menu has zero width and/or height"). So I
6332 have added a single, blank menu entry in this case.
6334 9. Added ${quote_local_part.
6336 10. MIME decoding is now applied to the contents of Subject: header lines when
6339 11. Now that a reference to $sender_host_address automatically causes a reverse
6340 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6341 host lookup before query-style lookups in lists that might use this
6342 variable. This has therefore been abolished, and the "net-" prefix is no
6343 longer necessary for query-style lookups.
6345 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6346 have been a typo for LFLAGS, so it has been changed.
6348 13. The install script calls Exim with "-C /dev/null" in order to find the
6349 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6350 to be output. However, since Exim outputs its version number before the
6351 error, it didn't break the script. It just looked ugly. I fixed this by
6352 always allowing "-C /dev/null" if the caller is root.
6354 14. Ignore overlarge ACL variable number when reading spool file - insurance
6355 against a later release with more variables having written the file.
6357 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6358 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6359 this, as well as the form without the "IPv6" on the front.
6361 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6362 OS/Makefile-Darwin file.
6364 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6365 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6367 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6368 variable. However, some OS use ints for inodes. I've added cast to long int
6369 to get rid of the compiler warning.
6371 19. I had forgotten to lock out "/../" in configuration file names when
6372 ALT_CONFIG_PREFIX was set.
6374 20. Routers used for verification do not need to specify transports. However,
6375 if such a router generated a host list, and callout was configured, Exim
6376 crashed, because it could not find a port number from the (non-existent)
6377 transport. It now assumes port 25 in this circumstance.
6379 21. Added the -t option to exigrep.
6381 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6382 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6383 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6384 them is set, LOOKUP_LSEARCH is forced.
6386 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6388 24. Added sender and host information to the "rejected by local_scan()" log
6389 line; previously there was no indication of these.
6391 25. Added .include_if_exists.
6393 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6394 Linux. It turns out that not all file systems support this. Apparently some
6395 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6396 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6397 not supported on the file descriptor, is now ignored when Exim is trying to
6398 sync a directory. This applies only to Linux.
6400 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6402 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6403 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6406 29. Added support for maildirsize files from supplied patch (modified a bit).
6408 30. The use of :fail: followed by an empty string could lead Exim to respond to
6409 sender verification failures with (e.g.):
6411 550 Verification failed for <xxx>
6412 550 Sender verify failed
6414 where the first response line was missing the '-' that indicates it is not
6415 the final line of the response.
6417 31. The loop for finding the name of the user that called Exim had a hardwired
6418 limit of 10; it now uses the value of finduser_retries, which is used for
6419 all other user lookups.
6421 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6424 33. Exim was neglecting to zero errno before one call of strtol() when
6425 expanding a string and expecting an integer value. On some systems this
6426 resulted in spurious "integer overflow" errors. Also, it was casting the
6427 result into an int without checking.
6429 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6430 did not work. The code looks as if it has *never* worked, though it appears
6431 to have been documented since at least release 1.62. I have made it work.
6433 35. The "timeout_DNS" error in retry rules, also documented since at least
6434 1.62, also never worked. As it isn't clear exactly what this means, and
6435 clearly it isn't a major issue, I have abolished the feature by treating it
6436 as "timeout", and writing a warning to the main and panic logs.
6438 36. The display of retry rules for -brt wasn't always showing the error code
6441 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6442 timeout_connect_A, timeout_connect_MX.
6444 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6445 to the empty sender.
6447 39. The daemon was not analysing the content of -oX till after it had closed
6448 stderr and disconnected from the controlling terminal. This meant that any
6449 syntax errors were only noted on the panic log, and the return code from
6450 the command was 0. By re-arranging the code a little, I've made the
6451 decoding happen first, so such errors now appear on stderr, and the return
6452 code is 1. However, the actual setting up of the sockets still happens in
6453 the disconnected process, so errors there are still only recorded on the
6456 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6457 connections (as happens on some IP stacks) was logged at start up time as
6458 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6459 it from "IPv6 and IPv4", which means that two separate sockets are being
6462 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6463 says whether AF_INET or AF_INET6 was passed as an argument.
6465 42. Exiwhat output was messed up when time zones were included in log
6468 43. Exiwhat now gives more information about the daemon's listening ports,
6469 and whether -tls-on-connect was used.
6471 44. The "port" option of the smtp transport is now expanded.
6473 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6474 silently ignored. Now an error message is written to the main and panic
6477 46. There's a new ACL modifier called "logwrite" which writes to a log file
6478 as soon as it is encountered.
6480 47. Added $local_user_uid and $local_user_gid at routing time.
6482 48. Exim crashed when trying to verify a sender address that was being
6485 49. Exim was recognizing only a space character after ".include". It now also
6486 recognizes a tab character.
6488 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6489 extracting the relevant information from the specification. The man page no
6490 longer contains scrambled data for the -d option, and I've added a section
6491 at the front about calling Exim under different names.
6493 51. Added "extra_headers" argument to the "mail" command in filter files.
6495 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6498 53. Installed eximstats 1.29.
6500 54. Added transport_filter_timeout as a generic transport option.
6502 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6503 Cc: header lines. This was required by RFC 822, but it not required by RFC
6506 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6507 incoming messages that did not have them. Now it does so only if the
6508 message originates locally, that is, if there is no associated remote host
6509 address. When Resent- header lines are present, this applies to the Resent-
6510 lines rather than the non-Resent- lines.
6512 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6513 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6515 58. Messages for configuration errors now include the name of the main
6516 configuration file - useful now that there may be more than one file in a
6517 list (.included file names were always shown).
6519 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6520 for those rare installations that do not start the daemon as root or run it
6521 setuid root. I've cut out the call to initgroups() if the daemon is not
6524 60. The Exim user and group can now be bound into the binary as text strings
6525 that are looked up at the start of Exim's processing.
6527 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6529 62. Added $mailstore_basename variable.
6531 63. Installed patch to sieve.c from Michael Haardt.
6533 64. When Exim failed to open the panic log after failing to open the main log,
6534 the original message it was trying to log was written to stderr and debug
6535 output, but if they were not available (the usual case in production), it
6536 was lost. Now it is written to syslog before the two lines that record the
6537 failures to open the logs.
6539 65. Users' Exim filters run in subprocesses under the user's uid. It is
6540 possible for a "deliver" command or an alias in a "personal" command to
6541 provoke an address rewrite. If logging of address rewriting is configured,
6542 this fails because the process is not running as root or exim. There may be
6543 a better way of dealing with this, but for the moment (because 4.30 needs
6544 to be released), I have disabled address rewrite logging when running a
6545 filter in a non-root, non-exim process.
6551 1. The buildconfig auxiliary program wasn't quoting the value set for
6552 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6553 not defined. This bug was masked in 4.22 by the effect that was fixed in
6556 2. Some messages that were rejected after a message id was allocated were
6557 shown as "incomplete" by exigrep. It no longer does this for messages that
6558 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6560 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6561 have allow_domain_literals set, the ID did not get logged in the <= line.
6562 Domain literals are now always recognized in Message-ID: header lines.
6564 4. The first argument for a ${extract expansion item is the key name or field
6565 number. Leading and trailing spaces in this item were not being ignored,
6566 causing some misleading effects.
6568 5. When deliver_drop_privilege was set, single queue runner processes started
6569 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6570 same command in the process it spins off) were not dropping privilege.
6572 6. When the daemon running as "exim" started a queue runner, it always
6573 re-executed Exim in the spun-off process. This is a waste of effort when
6574 deliver_drop_privilege is set. The new process now just calls the
6575 queue-runner function directly.
6581 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6584 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6585 ignored. Though the use of -oP was forcing the writing of a pid file, it
6586 was always written to the default place.
6588 3. If the message "no IP address found for host xxxx" is generated during
6589 incoming verification, it is now followed by identification of the incoming
6590 connection (so you can more easily find what provoked it).
6592 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6594 5. Added some features to "harden" Exim a bit more against certain attacks:
6596 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6597 be put in Local/Makefile. This is like the never_users runtime option,
6598 but it cannot be overridden. The default setting is "root".
6600 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6601 prefix string with which any file named in a -C command line option
6604 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6605 is retained for -C and -D only if the caller of Exim is root. Without
6606 it, the exim user may also use -C and -D and retain privilege.
6608 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6609 command line option is disabled.
6611 6. Macro names set by the -D option must start with an upper case letter, just
6612 like macro names defined in the configuration file.
6614 7. Added "dereference=" facility to LDAP.
6616 8. Two instances of the typo "uknown" in the source files are fixed.
6618 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6619 the Configure-Makefile script screwed up while processing it.
6621 10. Incorporated PCRE 4.4.
6623 11. The SMTP synchronization check was not operating right at the start of an
6624 SMTP session. For example, it could not catch a HELO sent before the client
6625 waited for the greeting. There is now a check for outstanding input at the
6626 point when the greeting is written. Because of the duplex, asynchronous
6627 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6628 way, but not yet received, when the check is performed.
6630 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6631 on TCP/IP sockets, because this apparently causes some broken clients to
6634 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6635 unchanged) from the Cygwin maintainer.
6637 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6638 was supported instead of testing for mbx. Effectively a typo.
6640 15. The spa authenticator server code was not checking that the input it
6641 received was valid base64.
6643 16. The debug output line for the "set" modifier in ACLs was not showing the
6644 name of the variable that was being set.
6646 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6648 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6649 lookup on its own. Something else was needed to trigger the lookup. For
6650 example, a match in host_lookup or the need for a host name in a host list.
6651 Now, if $sender_host_name is referenced and the host name has not yet been
6652 looked up, a lookup is performed. If the lookup fails, the variable remains
6653 empty, and $host_lookup_failed is set to "1".
6655 19. Added "eqi" as a case-independent comparison operator.
6657 20. The saslauthd authentication condition could segfault if neither service
6658 nor realm was specified.
6660 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6661 error message that was logged was misleading, and incoming SMTP
6662 connections were dropped. The message is now more accurate, and temporary
6663 errors are given to SMTP connections.
6665 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6666 (except RSET) with a 421 error, until QUIT is received. However, it was
6667 failing to send a response to QUIT.
6669 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6670 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6671 ACL is now run after the helo_try_verify_hosts code.
6673 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6674 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6675 case-independent, so other case variants are also recognized). Apparently
6676 some systems use these upper case variants.
6678 25. If more than two messages were waiting for the same host, and a transport
6679 filter was specified for the transport, Exim sent two messages over the
6680 same TCP/IP connection, and then failed with "socket operation on non-
6681 socket" when it tried to send the third.
6683 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6685 27. The extern definition of crypt16() in expand.c was not being excluded when
6686 the OS had its own crypt16() function.
6688 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6689 as a preferred synonym for return_size_limit, both as an option and as an
6692 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6694 30. Changed the default configuration ACL to relax the local part checking rule
6695 for addresses that are not in any local domains. For these addresses,
6696 slashes and pipe symbols are allowed within local parts, but the sequence
6697 /../ is explicitly forbidden.
6699 31. SPA server authentication was not clearing the challenge buffer before
6702 32. log_message in a "warn" ACL statement was writing to the reject log as
6703 well as to the main log, which contradicts the documentation and doesn't
6704 seem right (because no rejection is happening). So I have stopped it.
6706 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6707 However, I am unable to do any testing of this.
6709 34. Fixed an infelicity in the appendfile transport. When checking directories
6710 for a mailbox, to see if any needed to be created, it was accidentally
6711 using path names with one or more superfluous leading slashes; tracing
6712 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6714 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6715 contents are added to the log line that is written for every discarded
6716 recipient. (Previously a log_message setting was ignored.)
6718 36. The ${quote: operator now quotes the string if it is empty.
6720 37. The install script runs exim in order to find its version number. If for
6721 some reason other than non-existence or emptiness, which it checks, it
6722 could not run './exim', it was installing it with an empty version number,
6723 i.e. as "exim-". This error state is now caught, and the installation is
6726 38. An argument was missing from the function that creates an error message
6727 when Exim fails to connect to the socket for saslauthd authentication.
6728 This could cause Exim to crash, or give a corrupted message.
6730 39. Added isip, isip4, and isip6 to ${if conditions.
6732 40. The ACL variables $acl_xx are now saved with the message, and can be
6733 accessed later in routers, transports, and filters.
6735 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6736 strings in the file are not string-expanded.
6738 42. If a MAIL command specified a SIZE value that was too large to fit into an
6739 int variable, the check against message_size_limit failed. Such values are
6740 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6741 day this will have to be increased, but I don't think I want to be around
6742 when emails are that large.
6749 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6750 iconv() is not standard in FreeBSD.
6752 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6753 IPv6 enabled. The observed symptom was a segmentation fault on return from
6754 the function os_common_find_running_interfaces() in src/os.c.
6756 3. In the check_special_case() function in daemon.c I had used "errno" as an
6757 argument name, which causes warnings on some systems. This was basically a
6758 typo, since it was named "eno" in the comments!
6760 4. The code that waits for the clock to tick (at a resolution of some fraction
6761 of a second) so as to ensure message-id uniqueness was always waiting for
6762 at least one whole tick, when it could have waited for less. [This is
6763 almost certainly not relevant at current processor speeds, where it is
6764 unlikely to ever wait at all. But we try to future-proof.]
6766 5. The function that sleeps for a time interval that includes fractions of a
6767 second contained a race. It did not block SIGALRM between setting the
6768 timer, and suspending (a couple of lines later). If the interval was short
6769 and the sigsuspend() was delayed until after it had expired, the suspension
6770 never ended. On busy systems this could lead to processes getting stuck for
6773 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6774 process, before it forks any delivery processes. The open lookup caching
6775 mechanism meant that the open file or database connection was passed into
6776 the delivery process. The problem was that delivery processes always tidy
6777 up cached lookup data. This could cause a problem for the next delivery
6778 process started by the queue runner, because the external queue runner
6779 process does not know about the closure. So the next delivery process
6780 still has data in the lookup cache. In the case of a file lookup, there was
6781 no problem because closing a file descriptor in a subprocess doesn't affect
6782 the parent. However, if the lookup was caching a connection to a database,
6783 the connection was closed, and the second delivery process was likely to
6784 see errors such as "PGSQL: query failed: server closed the connection
6785 unexpectedly". The problem has been fixed by closing all cached lookups
6786 in a queue runner before running a delivery process.
6788 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6789 seem to have the "const" qualifier which it has on other OS. I've
6792 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6793 *of the same type* (client or server) with the same public name that an
6794 error should be diagnosed.
6796 9. When Exim looked up a host name for an IP address, but failed to find the
6797 original IP address when looking up the host name (a safety check), it
6798 output the message "<ip address> does not match any IP for NULL", which was
6799 confusing, to say the least. The bug was that the host name should have
6800 appeared instead of "NULL".
6802 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6803 user that is built into the binary, and the -C or -D options is used, root
6804 privilege is dropped before the configuration file is read. In addition,
6805 logging is switched to stderr instead of the normal log files. If the
6806 configuration then re-defines the Exim user, the unprivileged environment
6807 is probably not what is expected, so Exim logs a panic warning message (but
6810 However, if deliver_drop_privilege is set, the unprivileged state may well
6811 be exactly what is intended, so the warning has been cut out in that case,
6812 and Exim is allowed to try to write to its normal log files.
6818 1. smtp_return_error_details was not giving details for temporary sender
6819 or receiver verification errors.
6821 2. Diagnose a configuration error if two authenticators have the same public
6824 3. Exim used not to create the message log file for a message until the first
6825 delivery attempt. This could be confusing when incoming messages were held
6826 for policy or load reasons. The message log file is now created at the time
6827 the message is received, and an initial "Received" line is written to it.
6829 4. The automatically generated man page for command line options had a minor
6830 bug that caused no ill effects; however, a more serious problem was that
6831 the procedure for building the man page automatically didn't always
6832 operate. Consequently, release 4.20 contains an out-of-date version. This
6833 shouldn't happen again.
6835 5. When building Exim with embedded Perl support, the script that builds the
6836 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6837 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6839 6. The freeze_tell option was not being used for messages that were frozen on
6840 arrival, either by an ACL or by local_scan().
6842 7. Added the smtp_incomplete_transaction log selector.
6844 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6845 was accepting AUTH without a new EHLO.
6847 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6848 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6851 10. Logging of TCP/IP connections (when configured) now happens in the main
6852 daemon process instead of the child process, so that the TCP/IP connection
6853 count is more accurate (but it can never be perfect).
6855 11. The use of "drop" in a nested ACL was not being handled correctly in the
6856 outer ACL. Now, if condition failure induced by the nested "drop" causes
6857 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6858 or "require"), the connection is dropped.
6860 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6861 that yield "discard" can now be used with an "accept" or a "discard" verb,
6862 but an error is generated for any others (because I can't see a useful way
6863 to define what should happen).
6865 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6866 are now processed in the same way as lines in the Exim configuration file.
6867 In particular, continuation lines are supported.
6869 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6871 15. Added -ti meaning -t -i.
6873 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6874 domains, and warn by logging if others are found.
6876 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6877 returned from GIFCONF when Exim is trying to find the list of interfaces on
6878 a host. The code in os.c has been modified to copy each ifreq to an aligned
6879 structure in all cases.
6881 Also, in some cases, the returned ifreq's were being copied to a 'struct
6882 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6883 means the last couple of bytes of an IPv6 address could be chopped if the
6884 ifreq contained only a normal sockaddr (14 bytes storage).
6886 18. Named domain lists were not supported in the hosts_treat_as_local option.
6887 An entry such as +xxxx was not recognized, and was treated as a literal
6890 19. Ensure that header lines added by a DATA ACL are included in the reject log
6891 if the ACL subsequently rejects the message.
6893 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6894 MD5 (which is deprecated).
6896 21. When testing a filter file using -bf, Exim was writing a message when it
6897 took the sender from a "From " line in the message, but it was not doing so
6898 when it took $return_path from a Return-Path: header line. It now does.
6900 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6901 with a valid header line field name (a series of printing characters
6902 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6904 23. Changed "disc" in the source to "disk" to conform to the documentation and
6905 the book and for uniformity.
6907 24. Ignore Sendmail's -Ooption=value command line item.
6909 25. When execve() failed while trying to run a command in a pipe transport,
6910 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6911 could be confused with a return value of 69 from the command itself. This
6912 has been changed to 127, the value the shell returns if it is asked to run
6913 a non-existent command. The wording for the related log line suggests a
6914 non-existent command as the problem.
6916 26. If received_header_text expands to an empty string, do not add a Received:
6917 header line to the message. (Well, it adds a token one on the spool, but
6918 marks it "old" so that it doesn't get used or transmitted.)
6920 27. Installed eximstats 1.28 (addition of -nt option).
6922 28. There was no check for failure on the call to getsockname() in the daemon
6923 code. This can fail if there is a shortage of resources on the system, with
6924 ENOMEM, for example. A temporary error is now given on failure.
6926 29. Contrary to the C standard, it seems that in some environments, the
6927 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6928 program. Exim now does this explicitly; it affects the formatting of
6929 timestamps using strftime().
6931 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6932 complaints. I've now initialized all the variables, to avoid this.
6934 32. Header lines added by a system filter were not being "seen" during
6935 transport-time rewrites.
6937 33. The info_callback() function passed to OpenSSL is set up with type void
6938 (*)(SSL *, int, int), as described somewhere. However, when calling the
6939 function (actually a macro) that sets it up, the type void(*)() is
6940 expected. I've put in a cast to prevent warnings from picky compilers.
6942 34. If a DNS black list lookup found a CNAME record, but there were no A
6943 records associated with the domain it pointed at, Exim crashed.
6945 35. If a DNS black list lookup returned more than one A record, Exim ignored
6946 all but the first. It now scans all returned addresses if a particular IP
6947 value is being sought. In this situation, the contents of the
6948 $dnslist_value variable are a list of all the addresses, separated by a
6951 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6952 to accept a host name and all its aliases if the forward lookup for any of
6953 them yielded the IP address of the incoming connection. Now it accepts only
6954 those names whose forward lookup yields the correct IP address. Any other
6955 names are discarded. This closes a loophole whereby a rogue DNS
6956 administrator could create reverse DNS records to break through a
6957 wildcarded host restriction in an ACL.
6959 37. If a user filter or a system filter that ran in a subprocess used any of
6960 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6961 the wrong values were passed to the pipe command ($thisaddress had the
6962 value of $0, $0 had the value of $1, etc). This bug was introduced by
6963 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6965 38. Improved the line breaking for long SMTP error messages from ACLs.
6966 Previously, if there was no break point between 40 and 75 characters, Exim
6967 left the rest of the message alone. Two changes have been made: (a) I've
6968 reduced the minimum length to 35 characters; (b) if it can't find a break
6969 point between 35 and 75 characters, it looks ahead and uses the first one
6970 that it finds. This may give the occasional overlong line, but at least the
6971 remaining text gets split now.
6973 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6974 file descriptors might be low, and that setting 1000 would always raise it.
6975 It turns out that in some environments, the limit is already over 1000 and
6976 that lowering it causes trouble. So now Exim takes care not to decrease it.
6978 40. When delivering a message, the value of $return_path is set to $sender_
6979 address at the start of routing (routers may change the value). By an
6980 oversight, this default was not being set up when an address was tested by
6981 -bt or -bv, which affected the outcome if any router or filter referred to
6984 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6985 log only when "message" or "log_message" are set. However, if one of the
6986 conditions was an address verification, or a call to a nested ACL, the
6987 messages generated by the underlying test were being passed through. This
6988 no longer happens. The underlying message is available in $acl_verify_
6989 message for both "message" and "log_message" expansions, so it can be
6990 passed through if needed.
6992 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6993 new expansion $bh_ to give the encoded byte string without charset
6994 translation. Translation happens only if iconv() is available; HAVE_ICONV
6995 indicates this at build time. HEADERS_CHARSET gives the charset to
6996 translate to; headers_charset can change it in the configuration, and
6997 "headers charset" can change it in an individual filter file.
6999 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7000 that creates RFC 2047 encoded "words" labels them as that charset instead
7001 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7002 expansion operator; (ii) when Exim creates a From: line for a local
7003 message; (iii) when a header line is rewritten to include a "phrase" part.
7005 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7006 buggy, causing it to skip the first lines of messages whose message ID
7007 ended in 'D'. This would not have bitten before Exim release 4.14, because
7008 message IDs were unlikely to end in 'D' before then. The effect was to have
7009 incorrect size information for certain domains.
7011 45. #include "config.h" was missing at the start of the crypt16.c module. This
7012 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7015 46. If there was a timeout during a "random" callout check, Exim treated it as
7016 a failure of the random address, and carried on sending RSET and the real
7017 address. If the delay was just some slowness somewhere, the response to the
7018 original RCPT would be taken as a response to RSET and so on, causing
7019 mayhem of various kinds.
7021 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7022 when I implemented it. It didn't allow for the fact that some option values
7023 may legitimately be negative (e.g. size_addition), and it didn't even do
7024 the right test for positive values.
7026 48. Domain names in DNS records are case-independent. Exim always looks them up
7027 in lower case. Some resolvers return domain names in exactly the case they
7028 appear in the zone file, that is, they may contain uppercase letters. Not
7029 all resolvers do this - some return always lower case. Exim was treating a
7030 change of case by a resolver as a change of domain, similar to a widening
7031 of a domain abbreviation. This triggered its re-routing code and so it was
7032 trying to route what was effectively the same domain again. This normally
7033 caused routing to fail (because the router wouldn't handle the domain
7034 twice). Now Exim checks for this case specially, and just changes the
7035 casing of the domain that it ultimately uses when it transmits the message
7038 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7041 50. If a filter generated a file delivery with a non-absolute name (possible if
7042 no home directory exists for the router), the forbid_file option was not
7045 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7046 the existing equality matching.
7048 52. Exim was using ints instead of ino_t variables in some places where it was
7049 dealing with inode numbers.
7051 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7052 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7053 called TMPDIR, and if it finds it is different, it changes its value.
7055 54. The smtp_printf() function is now made available to local_scan() so
7056 additional output lines can be written before returning. There is also an
7057 smtp_fflush() function to enable the detection of a dropped connection.
7058 The variables smtp_input and smtp_batched_input are exported to
7061 55. Changed the default runtime configuration: the message "Unknown user"
7062 has been removed from the ACL, and instead placed on the localuser router,
7063 using the cannot_route_message feature. This means that any verification
7064 failures that generate their own messages won't get overridden. Similarly,
7065 the "Unrouteable address" message that was in the ACL for unverifiable
7066 relay addresses has also been removed.
7068 56. Added hosts_avoid_esmtp to the smtp transport.
7070 57. The exicyclog script was not checking for the esoteric option
7071 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7072 will work only if exicyclog is run under the appropriate euid.
7074 58. Following a discussion on the list, the rules by which Exim recognises line
7075 endings on incoming messages have been changed. The -dropcr and drop_cr
7076 options are now no-ops, retained only for backwards compatibility. The
7077 following line terminators are recognized: LF CRLF CR. However, special
7078 processing applies to CR:
7080 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7081 nor a local message in the state where . is a terminator.
7083 (ii) If a bare CR is encountered in a header line, an extra space is added
7084 after the line terminator so as not to end the header. The reasoning
7085 behind this is that bare CRs in header lines are most likely either
7086 to be mistakes, or people trying to play silly games.
7088 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7089 was being incorrectly given as 18 bytes larger than it should have been.
7090 This is a VOB (very old bug).
7092 60. This may never have affected anything current, but just in case it has:
7093 When the local host is found other than at the start of a list of hosts,
7094 the local host, those with the same MX, and any that follow, are discarded.
7095 When the list in question was part of a longer list of hosts, the following
7096 hosts (not currently being processed) were also being discarded. This no
7097 longer happens. I'm not sure if this situation could ever has previously
7100 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7103 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7104 header. This is something that is recommended in a new Internet Draft, and
7105 is something that is documented as being done by Sendmail. There are two
7106 possible values. For messages generated by the autoreply transport, Exim
7109 Auto-Submitted: auto-replied
7111 whereas for all other generated messages (e.g. bounces) it adds
7113 Auto-Submitted: auto-generated
7115 63. The "personal" condition in filters now includes a test for the
7116 Auto-Submitted: header. If it contains the string "auto-" the message it
7117 not considered personal.
7119 64. Added rcpt_include_affixes as a generic transport option.
7121 65. Added queue_only_override (default true).
7123 66. Added the syslog_duplication option.
7125 67. If what should have been the first header line of a message consisted of
7126 a space followed by a colon, Exim was mis-interpreting it as a header line.
7127 It isn't of course - it is syntactically invalid and should therefore be
7128 treated as the start of the message body. The misbehaviour could have
7129 caused a number of strange effects, including loss of data in subsequent
7130 header lines, and spool format errors.
7132 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7133 client host had authenticated. This control can now be exercised by an ACL
7134 for more flexibility.
7136 69. By default, callouts do not happen when testing with -bh. There is now a
7137 variant, -bhc, which does actually run the callout code, including
7138 consulting and updating the callout cache.
7140 70. Added support for saslauthd authentication, courtesy of Alexander
7143 71. If statvfs() failed on the spool or log directories while checking their
7144 size for availability, Exim confusingly gave the error "space shortage".
7145 Furthermore, in debugging mode it crashed with a floating point exception.
7146 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7147 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7148 serious problem, Exim now writes to the main and panic logs when this
7149 happens, with details of the failure. It then refuses to accept the
7150 incoming message, giving the message "spool directory problem" or "log
7151 directory problem" with a 421 code for SMTP messages.
7153 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7154 0, 1, and 2 exist, because some OS complain for execs without them (see
7155 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7156 descriptors. However, the code omitted to check that the open succeeded,
7157 causing mysterious errors if for some reason the permissions on /dev/null
7158 got screwed. Now Exim writes a message to the main and panic logs, and
7159 bombs out if it can't open /dev/null.
7161 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7162 interact so that it is all more flexible. It is supposed to remain
7163 backwards compatible. Also added extra_local_interfaces.
7165 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7166 to bomb out with an assertion failure - to the client this appears as a
7167 connection drop. This problem occurs in the part of the code that was taken
7168 from the Samba project. Fortunately, the assertion is in a very simple
7169 function, so I have fixed this by reproducing the function inline in the
7170 one place where it is called, and arranging for authentication to fail
7171 instead of killing the process with assert().
7173 75. The SPA client code was not working when the server requested OEM rather
7174 than Unicode encoding.
7176 76. Added code to make require_files with a specific uid setting more usable in
7177 the case where statting the file as root fails - usually a non-root-mounted
7178 NFS file system. When this happens and the failure is EACCES, Exim now
7179 forks a subprocess and does the per-uid checking as the relevant uid.
7181 77. Added process_log_path.
7183 78. If log_file_path was not explicitly set, a setting of check_log_space or
7184 check_log_inodes was ignored.
7186 79. If a space check for the spool or log partitions fails, the incident is now
7187 logged. Of course, in the latter case the data may get lost...
7189 80. Added the %p formatting code to string_format() so that it can be used to
7190 print addresses in debug_print(). Adjusted all the address printing in the
7191 debugging in store.c to use %p rather than %d.
7193 81. There was a concern that a line of code in smtp_in.c could overflow a
7194 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7195 initially expressed, the concern was not well-founded, because trailing
7196 spaces are removed early. However, if the trailing spaces were followed by
7197 a NULL, they did not get removed, so the overflow was possible. Two fixes
7200 (a) I re-wrote the offending code in a cleaner fashion.
7201 (b) If an incoming SMTP command contains a NULL character, it is rejected
7204 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7205 runs initgroups(), so that if the Exim user is in any additional groups,
7206 they will be used during message reception.
7212 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob