1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
12 non-signal-safe functions being used.
14 JH/03 Bug 2269: When presented with a received message having a stupidly large
15 number of DKIM-Signature headers, disable DKIM verification to avoid
16 a resource-consumption attack. The limit is set at twenty.
18 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
19 files in the spool were present and unlocked. A queue-runner could spot
20 them, resulting in a duplicate delivery. Fix that by doing the unlock
21 after the unlink. Investigation by Tim Stewart.
23 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
24 a queue-runner could start a delivery while other operations were ongoing.
25 Cutthrough delivery was a common victim, resulting in duplicate delivery.
26 Found and investigated by Tim Stewart. Fix by using the open message data
27 file handle rather than opening another, and not locally closing it (which
28 releases a lock) for that case, while creating the temporary .eml format
29 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
31 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
38 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
39 When redis_servers is set to a list of > 1 element, and the Redis servers
40 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
41 case of MOVED into a DEFER case instead, thus moving the query onto the
42 next server in the list. For a cluster of N elements, all N servers must
43 be defined in redis_servers.
45 GF/02 Catch and remove uninitialized value warning in exiqsumm
46 Check for existence of @ARGV before looking at $ARGV[0]
48 JH/01 Replace the store_release() internal interface with store_newblock(),
49 which internalises the check required to safely use the old one, plus
50 the allocate and data copy operations duplicated in both (!) of the
53 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
54 modifier. This matches the restriction on the commandline.
56 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
57 Previously only the last row was returned.
59 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
60 we assumed that tags in the header were well-formed, and parsed the
61 element content after inspecting only the first char of the tag.
62 Assumptions at that stage could crash the receive process on malformed
65 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
66 While running the DKIM ACL we operate on the Permanent memory pool so that
67 variables created with "set" persist to the DATA ACL. Also (at any time)
68 DNS lookups that fail create cache records using the Permanent pool. But
69 expansions release any allocations made on the current pool - so a dnsdb
70 lookup expansion done in the DKIM ACL releases the memory used for the
71 DNS negative-cache, and bad things result. Solution is to switch to the
72 Main pool for expansions.
73 While we're in that code, add checks on the DNS cache during store_reset,
74 active in the testsuite.
75 Problem spotted, and debugging aided, by Wolfgang Breyha.
77 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
78 When none of the hosts presented to a transport match an already-open
79 connection, close it and proceed with the list. Previously we would
80 queue the message. Spotted by Lena with Yahoo, probably involving
83 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
84 Previously a spurious "250 OK id=" response was appended to the proper
87 JH/08 The "support for" informational output now, which built with Content
88 Scanning support, has a line for the malware scanner interfaces compiled
89 in. Interface can be individually included or not at build time.
91 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
92 by the template makefile "src/EDITME". The "STREAM" support for an older
93 ClamAV interface method is removed.
95 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
96 rows affected is given instead).
98 JH/11 The runtime Berkeley DB library version is now additionally output by
99 "exim -d -bV". Previously only the compile-time version was shown.
101 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
102 SMTP connection. Previously, when one had more receipients than the
103 first, an abortive onward connection was made. Move to full support for
104 multiple onward connections in sequence, handling cutthrough connection
105 for all multi-message initiating connections.
107 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
108 routers. Previously, a multi-recipient message would fail to match the
109 onward-connection opened for the first recipient, and cause its closure.
111 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
112 a timeout on read on a GnuTLS initiating connection, resulting in the
113 initiating connection being dropped. This mattered most when the callout
114 was marked defer_ok. Fix to keep the two timeout-detection methods
117 JH/15 Relax results from ACL control request to enable cutthrough, in
118 unsupported situations, from error to silently (except under debug)
119 ignoring. This covers use with PRDR, frozen messages, queue-only and
122 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
124 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
125 metadata, resulting in a crash in free().
127 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
128 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
129 Broken also in d185889f4, with init system revamp.
131 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
132 Previously we abruptly closed the connection after reading a malware-
133 found indication; now we go on to read the "scan ok" response line,
136 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
137 ACL. Previously, a crash would result.
139 JH/19 Speed up macro lookups during configuration file read, by skipping non-
140 macro text after a replacement (previously it was only once per line) and
141 by skipping builtin macros when searching for an uppercase lead character.
143 JH/20 DANE support moved from Experimental to mainline. The Makefile control
144 for the build is renamed.
146 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
147 was allocated for every new TLS startup, meaning one per message. Fix
148 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
150 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
151 reported the original. Fix to report (as far as possible) the ACL
152 result replacing the original.
154 JH/23 Fix memory leak during multi-message connections using STARTTLS under
155 OpenSSL. Certificate information is loaded for every new TLS startup,
156 and the resources needed to be freed.
158 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
160 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
163 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
164 DATA response info to the (existing) per-recipient response info for
165 the "C=" log element. It can have useful tracking info from the
166 destination system. Patch from Simon Arlott.
168 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
169 length value. Previously this would segfault.
171 HS/02 Support Avast multiline protoocol, this allows passing flags to
172 newer versions of the scanner.
174 JH/28 Ensure that variables possibly set during message acceptance are marked
175 dead before release of memory in the daemon loop. This stops complaints
176 about them when the debug_store option is enabled. Discovered specifically
177 for sender_rate_period, but applies to a whole set of variables.
178 Do the same for the queue-runner and queue-list loops, for variables set
179 from spool message files. Do the same for the SMTP per-message loop, for
180 certain variables indirectly set in ACL operations.
182 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
183 as a multi-recipient message from a mailinglist manager). The coding had
184 an arbitrary cutoff number of characters while checking for more input;
185 enforced by writing a NUL into the buffer. This corrupted long / fast
186 input. The problem was exposed more widely when more pipelineing of SMTP
187 responses was introduced, and one Exim system was feeding another.
188 The symptom is log complaints of SMTP syntax error (NUL chars) on the
189 receiving system, and refused recipients seen by the sending system
190 (propating to people being dropped from mailing lists).
191 Discovered and pinpointed by David Carter.
193 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
194 replaced by the ${authresults } expansion.
196 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
198 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
199 allows proper process termination in container environments.
201 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
202 Previously the "final dot" had a newline after it; ensure it is CR,LF.
204 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
205 and "err_perm", deprecated since 4.83 when the RFC-defined words
206 "temperror" and "permerror" were introduced.
208 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
209 transport-filters or DKIM-signing. The restriction was lost in the
210 consolidation of verify-callout and delivery SMTP handling.
211 Extend the restriction to also cover ARC-signing.
213 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
214 in defer=pass mode supply a 450 to the initiator. Previously the message
217 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
218 tls_require_ciphers is used as before.
220 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
221 "pass_unscanned". Only tmpfails from the scanner are written to
222 the paniclog, as they may require admin intervention (permission
223 denied, license issues). Other scanner errors (like decompression
224 bombs) do not cause a paniclog entry.
226 JH/36 Fix reinitialisation of DKIM logging variable between messages.
227 Previously it was possible to log spurious information in receive log
230 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
231 triggered odd behaviour from Outlook Express clients.
233 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
236 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
237 since the IETF WG has not yet settled on that versus the original
238 "bare" representation.
240 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
241 Previously the millisecond value corrupted the output.
242 Fix also for syslog_pid=no and log_selector +pid, for which the pid
243 corrupted the output.
249 JH/01 Rework error string handling in TLS interface so that the caller in
250 more cases is responsible for logging. This permits library-sourced
251 string to be attached to addresses during delivery, and collapses
252 pairs of long lines into single ones.
254 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
255 during configuration. Wildcards are allowed and expanded.
257 JH/02 Rework error string handling in DKIM to pass more info back to callers.
258 This permits better logging.
260 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
261 do not close it down and have the child transport start it up again on
262 the passed-on TCP connection. Instead, proxy the child (and any
263 subsequent ones) for TLS via a unix-domain socket channel. Logging is
264 affected: the continued delivery log lines do not have any DNSSEC, TLS
265 Certificate or OCSP information. TLS cipher information is still logged.
267 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
268 identical IP addresses on different listening ports. Will also affect
271 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
272 add noisy ifdef guards to special-case this sillyness.
273 Patch from Bernd Kuhls.
275 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
276 than 255 are no longer allowed.
278 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
279 Disable the session-cache too, which might reduce our load. Since we
280 currrectly use a new context for every connection, both as server and
281 client, there is no benefit for these.
282 GnuTLS appears to not support tickets server-side by default (we don't
283 call gnutls_session_ticket_enable_server()) but client side is enabled
284 by default on recent versions (3.1.3 +) unless the PFS priority string
287 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
288 <https://reproducible-builds.org/specs/source-date-epoch/>.
290 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
291 the check for any unsuccessful recipients did not notice the limit, and
292 erroneously found still-pending ones.
294 JH/08 Pipeline CHUNKING command and data together, on kernels that support
295 MSG_MORE. Only in-clear (not on TLS connections).
297 JH/09 Avoid using a temporary file during transport using dkim. Unless a
298 transport-filter is involved we can buffer the headers in memory for
299 creating the signature, and read the spool data file once for the
300 signature and again for transmission.
302 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
303 4.77 as the kernel support then wasn't solid, having issues in 64bit
304 mode. Now, it's been long enough. Add support for FreeBSD also.
306 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
307 case where the routing stage had gathered several addresses to send to
308 a host before calling the transport for the first, we previously failed
309 to close down TLS in the old transport process before passing the TCP
310 connection to the new process. The new one sent a STARTTLS command
311 which naturally failed, giving a failed delivery and bloating the retry
312 database. Investigation and fix prototype from Wolfgang Breyha.
314 JH/12 Fix check on SMTP command input synchronisation. Previously there were
315 false-negatives in the check that the sender had not preempted a response
316 or prompt from Exim (running as a server), due to that code's lack of
317 awareness of the SMTP input buffering.
319 PP/04 Add commandline_checks_require_admin option.
320 Exim drops privileges sanely, various checks such as -be aren't a
321 security problem, as long as you trust local users with access to their
322 own account. When invoked by services which pass untrusted data to
323 Exim, this might be an issue. Set this option in main configuration
324 AND make fixes to the calling application, such as using `--` to stop
327 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
328 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
329 responses to those, into a single TLS record each way (this usually means
330 a single packet). As a side issue, smtp_enforce_sync now works on TLS
333 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
334 affects you only if you're dancing at the edge of the param size limits.
335 If you are, and this message makes sense to you, then: raise the
336 configured limit or use OpenSSL 1.1. Nothing we can do for older
339 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
340 cmdline element to get the documented default one. Previously it was
343 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
344 are unsuable for non-message-specific reasons - eg. connection timeout,
345 banner-time rejection.
347 JH/16 Fix logging of delivery remote port, when specified by router, under
350 PP/06 Repair manualroute's ability to take options in any order, even if one
351 is the name of a transport.
354 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
356 JH/17 Change the list-building routines interface to use the expanding-string
357 triplet model, for better allocation and copying behaviour.
359 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
360 Previously it was constructed the first time a possibly-matching string
361 was met in the configuration file input during startup; now it is done
364 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
365 compatible one, to avoid the (poorly documented) possibility of a config
366 file in the working directory redirecting the DB files, possibly correpting
367 some existing file. CVE-2017-10140 assigned for BDB.
369 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
370 cache-hot. Previously, although the result was properly cached, the
371 initial verify call returned a defer.
373 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
374 the main verify for receipient in uncached-mode.
376 JH/22 Retire historical build files to an "unsupported" subdir. These are
377 defined as "ones for which we have no current evidence of testing".
379 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
380 if present. Previously it was ignored.
382 JH/24 Start using specified-initialisers in C structure init coding. This is
383 a C99 feature (it's 2017, so now considered safe).
385 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
386 if was a fixed-sized field and bitmask ops via macros; it is now more
389 PP/07 GitHub PR 56: Apply MariaDB build fix.
390 Patch provided by Jaroslav Škarvada.
392 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
393 during Coverity cleanups [4.87 JH/47]
394 Diagnosis and fix provided by Michael Fischer v. Mollard.
396 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
397 the right size to place the terminating semicolon on its own folded
398 line, the header hash was calculated to an incorrect value thanks to
399 the (relaxed) space the fold became.
401 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
402 and confused the parent.
404 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
405 which could crash as a result. This could lead to undeliverable messages.
407 JH/28 Logging: "next input sent too soon" now shows where input was truncated
410 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
411 matters on fast-turnover and PID-randomising systems, which were getting
412 out-of-order delivery.
414 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
415 a possibly-overlapping copy. The symptom was that "Remote host closed
416 connection in response to HELO" was logged instead of the actual 4xx
419 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
420 Previously only that bufferd was discarded, resulting in SYMTP command
423 JH/32 DKIM: when a message has multiple signatures matching an identity given
424 in dkim_verify_signers, run the dkim acl once for each. Previously only
425 one run was done. Bug 2189.
427 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
428 "panic the current process" to "deliberately defer". The panic log is
429 still written with the problem list name; the mail and reject logs now
430 get a temp-reject line for the message that was being handled, saying
431 something like "domains check lookup or other defer". The SMTP 451
432 message is still "Temporary local problem".
434 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
435 A crafted sequence of BDAT commands could result in in-use memory beeing
436 freed. CVE-2017-16943.
438 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
439 from SMTP input. Previously it was always done; now only done for DATA
440 and not BDAT commands. CVE-2017-16944.
442 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
443 to the message (such as an overlong header line). Previously this was
444 not done and we did not exit BDAT mode. Followon from the previous item
445 though a different problem.
451 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
452 than -2003 did; needs libidn2 in addition to libidn.
454 JH/02 The path option on a pipe transport is now expanded before use.
456 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
457 Patch provided by "Björn", documentation fix added too.
459 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
460 missing a wire-to-host endian conversion.
462 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
463 close after a BDAT command line could be taken as a following command,
464 giving a synch failure. Fix by only checking for synch immediately
465 before acknowledging the chunk.
467 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
468 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
469 macro. Patches provided by Josh Soref.
471 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
472 Previously we did not; the RFC seems ambiguous and VRFY is not listed
473 by IANA as a service extension. However, John Klensin suggests that we
476 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
477 the dkim code may be unix-mode line endings rather than smtp wire-format
478 CRLF, so prepend a CR to any bare LF.
480 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
481 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
483 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
484 the first were themselves being wrongly included in the feed into dkim
485 processing; with most chunk sizes in use this resulted in an incorrect
486 body hash calculated value.
488 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
489 DKIM signature block, for verification. Although advised against by
490 standards it is specifically not ruled illegal.
492 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
494 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
495 missing a body hash (the bh= tag).
497 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
498 It seems that HAProxy sends the Proxy Protocol information in clear and
499 only then does a TLS startup, so do the same.
501 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
502 TCP connections (such as for Spamd) unless the daemon successfully set
503 Fast Open mode on its listening sockets. This fixes breakage seen on
504 too-old kernels or those not configured for Fast Open, at the cost of
505 requiring both directions being enabled for TFO, and TFO never being used
506 by non-daemon-related Exim processes.
508 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
509 endings, at least on the first header line. Try to canonify any that get
510 past that check, despite the cost.
512 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
513 now limited to an arbitrary five deep, while parsing addresses with the
514 strip_excess_angle_brackets option enabled.
516 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
517 instead leave the unprompted TLS handshake in socket buffer for the
518 TLS library to consume.
520 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
522 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
524 JH/16 Drop variables when they go out of scope. Memory management drops a whole
525 region in one operation, for speed, and this leaves assigned pointers
526 dangling. Add checks run only under the testsuite which checks all
527 variables at a store-reset and panics on a dangling pointer; add code
528 explicitly nulling out all the variables discovered. Fixes one known
529 bug: a transport crash, where a dangling pointer for $sending_ip_address
530 originally assigned in a verify callout, is re-used.
532 PP/06 Drop '.' from @INC in various Perl scripts.
534 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
536 PP/08 Reduce a number of compilation warnings under clang; building with
537 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
538 should be warning-free.
540 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
542 HS/01 Fix portability problems introduced by PP/08 for platforms where
543 realloc(NULL) is not equivalent to malloc() [SunOS et al].
545 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
546 chunk. This allows us to accept broken chunked messages. We need a more
547 general solution here.
549 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
550 already-broken messages in the queue.
552 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
554 JH/19 Fix reference counting bug in routing-generated-address tracking.
560 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
561 supports it and a size is available (ie. the sending peer gave us one).
563 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
564 years of being deprecated). The replacements are the ACLs
565 acl_smtp_mime and acl_not_smtp_mime.
567 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
568 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
569 or even an in-clear connection were permitted. Now, if the host lookup
570 was dnssec and dane was requested then the host is only used if the
571 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
572 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
573 if one fails this test.
574 This means that a poorly-configured remote DNS will make it incommunicado;
575 but it protects against a DNS-interception attack on it.
577 JH/04 Bug 1810: make continued-use of an open smtp transport connection
578 non-noisy when a race steals the message being considered.
580 JH/05 If main configuration option tls_certificate is unset, generate a
581 self-signed certificate for inbound TLS connections.
583 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
584 in rewrites and routers.
586 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
587 and logged a warning sing 4.83; now they are a configuration file error.
589 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
590 (lacking @domain). Apply the same qualification processing as RCPT.
592 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
594 JH/10 Support ${sha256:} applied to a string (as well as the previous
597 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
598 a cutthrough deliver is pending, as we always want to make a connection.
599 This also avoids re-routing the message when later placing the cutthrough
600 connection after a verify cache hit.
601 Do not update it with the verify result either.
603 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
604 when routing results in more than one destination address.
606 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
607 signing (which inhibits the cutthrough capability). Previously only
608 the presence of an option was tested; now an expansion evaluating as
609 empty is permissible (obviously it should depend only on data available
610 when the cutthrough connection is made).
612 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
613 the relevant preceding SMTP command did not note the pipelining mode.
615 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
616 Previously they were not counted.
618 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
619 as one having no matching records. Previously we deferred the message
620 that needed the lookup.
622 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
623 distinguished as "(=".
625 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
626 for missing MX records. Previously it only worked for missing A records.
628 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
630 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
631 after the data-go-ahead and data-ack. Patch from Jason Betts.
633 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
634 even for a "none" policy. Patch from Tony Meyer.
636 JH/22 Fix continued use of a connection for further deliveries. If a port was
637 specified by a router, it must also match for the delivery to be
640 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
641 When one of the recipients of a message was unsuitable for the connection
642 (has no matching addresses), we lost track of needing to mark it
643 deferred. As a result mail would be lost.
645 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
647 JH/25 Decoding ACL controls is now done using a binary search; the source code
648 takes up less space and should be simpler to maintain. Merge the ACL
649 condition decode tables also, with similar effect.
651 JH/26 Fix problem with one_time used on a redirect router which returned the
652 parent address unchanged. A retry would see the parent address marked as
653 delivered, so not attempt the (identical) child. As a result mail would
656 JH/27 Fix a possible security hole, wherein a process operating with the Exim
657 UID can gain a root shell. Credit to http://www.halfdog.net/ for
658 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
661 JH/28 Enable {spool,log} filesystem space and inode checks as default.
662 Main config options check_{log,spool}_{inodes,space} are now
663 100 inodes, 10MB unless set otherwise in the configuration.
665 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
666 Previously it only applied to the main-section connection policy
669 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
671 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
672 by me. Added RFC7919 DH primes as an alternative.
674 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
675 are not in the system include path.
677 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
678 GnuTLS, when a session startup failed (eg because the client disconnected)
679 Exim did stdio operations after fclose. This was exposed by a recent
680 change which nulled out the file handle after the fclose.
682 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
683 signed directly by the cert-signing cert, rather than an intermediate
684 OCSP-signing cert. This is the model used by LetsEncrypt.
686 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
688 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
689 an incoming connection.
691 HS/02 Bug 1802: Do not half-close the connection after sending a request
694 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
695 fallback to "prime256v1".
697 JH/34 SECURITY: Use proper copy of DATA command in error message.
698 Could leak key material. Remotely exploitable. CVE-2016-9963.
704 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
705 and 3.4.4 - once the server is enabled to respond to an OCSP request
706 it does even when not requested, resulting in a stapling non-aware
707 client dropping the TLS connection.
709 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
710 support variable-length bit vectors. No functional change.
712 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
713 The I= interface field on outgoing lines is now after the H= remote
714 host field, same as incoming lines. There is a separate
715 outgoing_interface log selector which allows you to disable the
718 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
719 If not running log_selector +smtp_connection the mainlog would be held
720 open indefinitely after a "too many connections" event, including to a
721 deleted file after a log rotate. Leave the per net connection logging
722 leaving it open for efficiency as that will be quickly detected by the
723 check on the next write.
725 HS/01 Bug 1671: Fix post transport crash.
726 Processing the wait-<transport> messages could crash the delivery
727 process if the message IDs didn't exist for some reason. When
728 using 'split_spool_directory=yes' the construction of the spool
729 file name failed already, exposing the same netto behaviour.
731 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
732 mime_regex ACL conditions.
734 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
735 to DSN fail messages (bounces): remote IP, remote greeting, remote response
736 to HELO, local diagnostic string.
738 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
739 log line to debug. Even when configured with a tls authenticator many
740 client connections are expected to not authenticate in this way, so
741 an authenticate fail is not an error.
743 HS/02 Add the Exim version string to the process info. This way exiwhat
744 gives some more detail about the running daemon.
746 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
747 matter for fast-change records such as DNSBLs.
749 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
750 retry record, even if constant. There may be multiple transports with
751 different interface settings and the retry behaviour needs to be kept
754 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
756 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
758 JH/10 Bug 840: fix log_defer_output option of pipe transport
760 JH/11 Bug 830: use same host for all RCPTS of a message, even under
761 hosts_randomize. This matters a lot when combined with mua_wrapper.
763 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
764 ${quote_pgsql:<string>} operator.
766 JH/13 Bug 1708: avoid misaligned access in cached lookup.
768 JH/14 Change header file name for freeradius-client. Relevant if compiling
769 with Radius support; from the Gentoo tree and checked under Fedora.
771 JH/15 Bug 1712: Introduce $prdr_requested flag variable
773 JH/16 Bug 1714: Permit an empty string as expansion result for transport
774 option transport_filter, meaning no filtering.
776 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
778 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
779 defaults to "*" (all hosts). The variable is now available when not built
780 with TLS, default unset, mainly to enable keeping the testsuite sane.
781 If a server certificate is not supplied (via tls_certificate) an error is
782 logged, and clients will find TLS connections fail on startup. Presumably
783 they will retry in-clear.
784 Packagers of Exim are strongly encouraged to create a server certificate
785 at installation time.
787 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
788 with the $config_file variable.
790 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
791 in transport context, after the attempt, and per-recipient. The latter type
792 is per host attempted. The event data is the error message, and the errno
793 information encodes the lookup type (A vs. MX) used for the (first) host,
794 and the trailing two digits of the smtp 4xx response.
796 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
797 to write to mainlog (or rejectlog, paniclog) in the window between file
798 creation and permissions/ownership being changed. Particularly affects
799 installations where exicyclog is run as root, rather than exim user;
800 result is that the running daemon panics and dies.
802 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
804 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
805 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
806 "pri" and "weight". Note that the previous implicit priority given by the
807 list order is no longer honoured.
809 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
812 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
813 by defining SUPPORT_SOCKS.
815 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
816 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
817 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
818 variables are renamed to proxy_{local,external}_{address,port}.
820 JH/27 Move Internationalisation support from Experimental to mainline, enabled
821 for a build by defining SUPPORT_I18N
823 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
824 of the query string, and make ${quote_redis:} do that quoting.
826 JH/29 Move Events support from Experimental to mainline, enabled by default
827 and removable for a build by defining DISABLE_EVENT.
829 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
831 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
832 cached by the daemon.
834 JH/32 Move Redis support from Experimental to mainline, enabled for a build
835 by defining LOOKUP_REDIS. The libhiredis library is required.
837 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
838 keys are given for lookup.
840 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
841 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
842 only supported when built with TLS support. The PolarSSL SHA routines
843 are still used when the TLS library is too old for convenient support.
845 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
846 openssl_options), for security. OpenSSL forces this from version 1.1.0
847 server-side so match that on older versions.
849 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
850 allocation for $value could be released as the expansion processing
851 concluded, but leaving the global pointer active for it.
853 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
854 and to use the domains and local_parts ACL conditions.
856 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
857 incorrectly not doubled on cutthrough transmission, hence seen as a
858 body-termination at the receiving system - resulting in truncated mails.
859 Commonly the sender saw a TCP-level error, and retransmitted the message
860 via the normal store-and-forward channel. This could result in duplicates
861 received - but deduplicating mailstores were liable to retain only the
862 initial truncated version.
864 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
866 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
868 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
869 we're in there, support oversigning also; bug 1309.
871 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
873 HS/04 Add support for keep_environment and add_environment options.
875 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
876 either intentional arithmetic overflow during PRNG, or testing config-
879 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
880 delivery resulted in actual delivery. Cancel cutthrough before DATA
883 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
884 rejects a recipient: pass the reject to the originator.
886 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
887 Many were false-positives and ignorable, but it's worth fixing the
890 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
891 for the new environment-manipulation done at startup. Move the routines
892 from being local to tls.c to being global via the os.c file.
894 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
895 an extract embedded as result-arg for a map, the first arg for extract
896 is unavailable so we cannot tell if this is a numbered or keyed
897 extraction. Accept either.
903 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
906 JH/02 The smtp transport option "multi_domain" is now expanded.
908 JH/03 The smtp transport now requests PRDR by default, if the server offers
911 JH/04 Certificate name checking on server certificates, when exim is a client,
912 is now done by default. The transport option tls_verify_cert_hostnames
913 can be used to disable this per-host. The build option
914 EXPERIMENTAL_CERTNAMES is withdrawn.
916 JH/05 The value of the tls_verify_certificates smtp transport and main options
917 default to the word "system" to access the system default CA bundle.
918 For GnuTLS, only version 3.0.20 or later.
920 JH/06 Verification of the server certificate for a TLS connection is now tried
921 (but not required) by default. The verification status is now logged by
922 default, for both outbound TLS and client-certificate supplying inbound
925 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
928 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
929 Status Notification (bounce) messages are now MIME format per RFC 3464.
930 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
931 under the control of the dsn_advertise_hosts option, and routers may
932 have a dsn_lasthop option.
934 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
935 default, modifiable by a malware= option. The list separator for
936 the options can now be changed in the usual way. Bug 68.
938 JH/10 The smtp_receive_timeout main option is now expanded before use.
940 JH/11 The incoming_interface log option now also enables logging of the
941 local interface on delivery outgoing connections.
943 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
944 if the interface and destination host and port all match.
946 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
949 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
950 Patch from Andrew Lewis.
952 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
953 now supports optional time-restrictions, weighting, and priority
954 modifiers per server. Patch originally by <rommer@active.by>.
956 JH/16 The spamd_address main option now supports a mixed list of local
957 and remote servers. Remote servers can be IPv6 addresses, and
958 specify a port-range.
960 JH/17 Bug 68: The spamd_address main option now supports an optional
961 timeout value per server.
963 JH/18 Bug 1581: Router and transport options headers_add/remove can
964 now have the list separator specified.
966 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
969 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
972 JH/21 Support for the A6 type of dns record is withdrawn.
974 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
975 rather than the verbs used.
977 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
978 from 255 to 1024 chars.
980 JH/24 Verification callouts now attempt to use TLS by default.
982 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
983 are generic router options now. The defaults didn't change.
985 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
986 Original patch from Alexander Shikoff, worked over by JH.
988 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
989 files if ZCAT_COMMAND is not executable.
991 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
993 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
995 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
996 Normally benign, it bites when the pair was led to by a CNAME;
997 modern usage is to not canonicalize the domain to a CNAME target
998 (and we were inconsistent anyway for A-only vs AAAA+A).
1000 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1002 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1003 when evaluating $sender_host_dnssec.
1005 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1006 $sender_helo_dnssec variable.
1008 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1010 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1012 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1014 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1015 documented as working, but never had. Support all but $spam_report.
1017 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1018 added for tls authenticator.
1020 HS/03 Add perl_taintmode main config option
1026 TL/01 When running the test suite, the README says that variables such as
1027 no_msglog_check are global and can be placed anywhere in a specific
1028 test's script, however it was observed that placement needed to be near
1029 the beginning for it to behave that way. Changed the runtest perl
1030 script to read through the entire script once to detect and set these
1031 variables, reset to the beginning of the script, and then run through
1032 the script parsing/test process like normal.
1034 TL/02 The BSD's have an arc4random API. One of the functions to induce
1035 adding randomness was arc4random_stir(), but it has been removed in
1036 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1037 function when detected.
1039 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1040 cause callback expansion.
1042 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1043 syntax errors in an expansion can be treated as a string instead of
1044 logging or causing an error, due to the internal use of bool_lax
1045 instead of bool when processing it.
1047 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1048 server certificates when making smtp deliveries.
1050 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1052 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1054 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1056 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1057 Merged patch from Sebastian Wiedenroth.
1059 JH/05 Fix results-pipe from transport process. Several recipients, combined
1060 with certificate use, exposed issues where response data items split
1061 over buffer boundaries were not parsed properly. This eventually
1062 resulted in duplicates being sent. This issue only became common enough
1063 to notice due to the introduction of connection certificate information,
1064 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1066 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1067 size buffer was used, resulting in syntax errors when an expansion
1070 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1071 version 3.3.6 or later.
1073 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1074 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1075 both become "event_action", the variables become $event_name, $event_data
1076 and $event_defer_errno. There is a new variable $verify_mode, usable in
1077 routers, transports and related events. The tls:cert event is now also
1078 raised for inbound connections, if the main configuration event_action
1081 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1082 early OCSP support, but no stapling (appears to be less than 1.0.0).
1084 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1085 server certificate names available under the smtp transport option
1086 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1089 JH/10 Time-related extraction expansions from certificates now use the main
1090 option "timezone" setting for output formatting, and are consistent
1091 between OpenSSL and GnuTLS compilations. Bug 1541.
1093 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1094 encoded parameter in the incoming message. Bug 1558.
1096 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1097 include certificate info, eximon was claiming there were spoolfile
1100 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1102 JH/14 Log delivery-related information more consistently, using the sequence
1103 "H=<name> [<ip>]" wherever possible.
1105 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1106 are problematic for Debian distribution, omit them from the release
1109 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1111 JH/16 Fix string representation of time values on 64bit time_t architectures.
1114 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1115 output list separator was used.
1120 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1121 checkers that were complaining about end of non-void function with no
1124 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1125 This was a regression introduced in 4.83 by another bugfix.
1127 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1129 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1130 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1136 TF/01 Correctly close the server side of TLS when forking for delivery.
1138 When a message was received over SMTP with TLS, Exim failed to clear up
1139 the incoming connection properly after forking off the child process to
1140 deliver the message. In some situations the subsequent outgoing
1141 delivery connection happened to have the same fd number as the incoming
1142 connection previously had. Exim would try to use TLS and fail, logging
1143 a "Bad file descriptor" error.
1145 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1146 utilities have not been installed.
1148 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1149 temporary space as the ACL may create new global variables.
1151 TL/01 LDAP support uses per connection or global context settings, depending
1152 upon the detected version of the libraries at build time.
1154 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1155 to extract and use the src ip:port in logging and expansions as if it
1156 were a direct connection from the outside internet. PPv2 support was
1157 updated based on HAProxy spec change in May 2014.
1159 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1161 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1162 Properly escape header and check for NULL return.
1164 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1167 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1169 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1170 characters in header names, implemented as a verify condition.
1171 Contributed by Michael Fischer v. Mollard.
1173 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1174 results permerror and temperror. Previous values are deprecated but
1175 still accepted. In a future release, err_perm and err_temp will be
1176 completely removed, which will be a backward incompatibility if the
1177 ACL tests for either of these two old results. Patch contributed by
1178 user bes-internal on the mailing list.
1180 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1182 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1183 selectors, in both main and reject logs.
1185 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1188 JH/07 Add malware type "sock" for talking to simple daemon.
1190 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1192 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1193 routers/transports under cutthrough routing.
1195 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1196 numbers. Touch up "bool" conditional to keep the same definition.
1198 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1200 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1202 JH/12 Expand items in router/transport headers_add or headers_remove lists
1203 individually rather than the list as a whole. Bug 1452.
1205 Required for reasonable handling of multiple headers_ options when
1206 they may be empty; requires that headers_remove items with embedded
1207 colons must have them doubled (or the list-separator changed).
1209 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1210 view the policy declared in the DMARC record. Currently, $dmarc_status
1211 is a combined value of both the record presence and the result of the
1214 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1216 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1217 dnslookup router and the smtp transport (applying to the forward
1220 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1221 of ldap servers used for a specific lookup. Patch provided by Heiko
1224 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1225 New variable $lookup_dnssec_authenticated for observability.
1227 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1228 Patch submitted by Lars Timman.
1230 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1232 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1233 Requires trusted mode and valid format message id, aborts otherwise.
1234 Patch contributed by Heiko Schlichting.
1236 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1237 certextract with support for various fields. Bug 1358.
1239 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1240 is requested by default, modifiable by smtp transport option
1243 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1244 operate on certificate variables to give certificate fingerprints
1245 Also new ${sha256:cert_variable}.
1247 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1249 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1250 Christian Aistleitner.
1252 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1254 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1255 file. Patch from Wolfgang Breyha.
1257 JH/25 Expand the coverage of the delivery $host and $host_address to
1258 client authenticators run in verify callout. Bug 1476.
1260 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1261 align with daemon_smtp_ports. Bug 72.
1263 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1264 support and error reporting did not work properly.
1266 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1267 and is readable. Patch from Andrew Colin Kissa.
1269 TL/14 Enhance documentation of ${run expansion and how it parses the
1270 commandline after expansion, particularly in the case when an
1271 unquoted variable expansion results in an empty value.
1273 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1275 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1276 by renaming away from it. Addresses GH issue 15, reported by
1279 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1281 TL/15 SECURITY: prevent double expansion in math comparison functions
1282 (can expand unsanitized data). Not remotely exploitable.
1289 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1291 PP/02 Make -n do something, by making it not do something.
1292 When combined with -bP, the name of an option is not output.
1294 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1297 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1298 $sender_host_name and config options to manage this, and basic check
1301 PP/05 DSCP support for outbound connections and control modifier for inbound.
1303 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1304 (Only plugin which currently uses this is kerberos4, which nobody should
1305 be using, but we should make it available and other future plugins might
1306 conceivably use it, even though it would break NAT; stuff *should* be
1307 using channel bindings instead).
1309 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1310 name; added for Sendmail compatibility; requires admin caller.
1311 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1312 just ignore it); requires trusted caller.
1313 Also parse but ignore: -Ac -Am -X<logfile>
1316 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1318 TL/02 Add +smtp_confirmation as a default logging option.
1320 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1321 Patch by Magnus Holmgren from 2007-02-20.
1323 TL/04 Bugzilla 1281 - Spec typo.
1324 Bugzilla 1283 - Spec typo.
1325 Bugzilla 1290 - Spec grammar fixes.
1327 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1329 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1331 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1332 mailing list by by Dmitry Isaikin.
1334 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1336 JH/02 Support "G" suffix to numbers in ${if comparisons.
1338 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1340 NM/01 Bugzilla 1197 - Spec typo
1341 Bugzilla 1196 - Spec examples corrections
1343 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1345 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1346 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1349 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1350 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1352 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1353 "acl {{name}{arg}...}", and optional args on acl condition
1356 JH/05 Permit multiple router/transport headers_add/remove lines.
1358 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1360 JH/07 Avoid using a waiting database for a single-message-only transport.
1361 Performance patch from Paul Fisher. Bugzilla 1262.
1363 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1366 JH/09 Add $headers_added variable, with content from use of ACL modifier
1367 add_header (but not yet added to the message). Bugzilla 199.
1369 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1370 Pulled from Bugzilla 817 by Wolfgang Breyha.
1372 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1374 (nb: this is the same fix as in Exim 4.80.1)
1376 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1379 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1382 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1384 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1385 advertise SMTP AUTH mechanism to us, instead of a generic
1386 protocol violation error. Also, make Exim more robust to bad
1387 data from the Dovecot auth socket.
1389 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1391 When a queue runner is handling a message, Exim first routes the
1392 recipient addresses, during which it prunes them based on the retry
1393 hints database. After that it attempts to deliver the message to
1394 any remaining recipients. It then updates the hints database using
1397 So if a recipient address works intermittently, it can get repeatedly
1398 deferred at routing time. The retry hints record remains fresh so the
1399 address never reaches the final cutoff time.
1401 This is a fairly common occurrence when a user is bumping up against
1402 their storage quota. Exim had some logic in its local delivery code
1403 to deal with this. However it did not apply to per-recipient defers
1404 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1406 This change adds a proper retry rule check during routing so that the
1407 final cutoff time is checked against the message's age. We only do
1408 this check if there is an address retry record and there is not a
1409 domain retry record; this implies that previous attempts to handle
1410 the address had the retry_use_local_parts option turned on. We use
1411 this as an approximation for the destination being like a local
1412 delivery, as in LMTP.
1414 I suspect this new check makes the old local delivery cutoff check
1415 redundant, but I have not verified this so I left the code in place.
1417 TF/02 Correct gecos expansion when From: is a prefix of the username.
1419 Test 0254 submits a message to Exim with the header
1423 When I ran the test suite under the user fanf2, Exim expanded
1424 the header to contain my full name, whereas it should have added
1425 a Resent-Sender: header. It erroneously treats any prefix of the
1426 username as equal to the username.
1428 This change corrects that bug.
1430 GF/01 DCC debug and logging tidyup
1431 Error conditions log to paniclog rather than rejectlog.
1432 Debug lines prefixed by "DCC: " to remove any ambiguity.
1434 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1436 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1437 Bug spotted by Jeremy Harris; was flawed since initial commit.
1438 Would have resulted in OCSP responses post-SNI triggering an Exim
1439 NULL dereference and crash.
1441 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1443 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1444 Bug detection, analysis and fix by Samuel Thibault.
1445 Bugzilla 1331, Debian bug #698092.
1447 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1449 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1450 Server implementation by Todd Lyons, client by JH.
1451 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1452 config variable "prdr_enable" controls whether the server
1453 advertises the facility. If the client requests PRDR a new
1454 acl_data_smtp_prdr ACL is called once for each recipient, after
1455 the body content is received and before the acl_smtp_data ACL.
1456 The client is controlled by both of: a hosts_try_prdr option
1457 on the smtp transport, and the server advertisement.
1458 Default client logging of deliveries and rejections involving
1459 PRDR are flagged with the string "PRDR".
1461 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1462 fclose(). Diagnosis by Todd Lyons.
1464 PP/17 Update configure.default to handle IPv6 localhost better.
1465 Patch by Alain Williams (plus minor tweaks).
1468 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1469 This is now consistent with GnuTLS, and is now documented: the
1470 previous undocumented portable approach to treating the option as
1471 unset was to force an expansion failure. That still works, and
1472 an empty string is now equivalent.
1474 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1475 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1476 not performing validation itself.
1478 PP/20 Added force_command boolean option to pipe transport.
1479 Patch from Nick Koston, of cPanel Inc.
1481 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1484 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1486 PP/21 Fix eximon continuous updating with timestamped log-files.
1487 Broken in a format-string cleanup in 4.80, missed when I repaired the
1488 other false fix of the same issue.
1489 Report and fix from Heiko Schlichting.
1492 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1493 Report from Prashanth Katuri.
1495 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1496 It's SecureTransport, so affects any MacOS clients which use the
1497 system-integrated TLS libraries, including email clients.
1499 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1500 using a MIME ACL for non-SMTP local injection.
1501 Report and assistance in diagnosis by Warren Baker.
1503 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1505 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1507 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1508 last id that failed so it may be referenced in subsequent ACL's.
1510 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1513 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1514 ldap library initialization, allowing self-signed CA's to be
1515 used. Also properly sets require_cert option later in code by
1516 using NULL (global ldap config) instead of ldap handle (per
1517 session). Bug diagnosis and testing by alxgomz.
1519 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1520 the src/util/ subdirectory.
1522 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1523 renamed to Transport Post Delivery Action by Jeremy Harris, as
1526 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1527 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1528 redis_servers = needs to be configured which will be used by the redis
1529 lookup. Patch from Warren Baker, of The Packet Hub.
1531 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1533 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1534 hostname or reverse DNS when processing a host list. Used suggestions
1535 from multiple comments on this bug.
1537 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1539 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1540 Missed a few lines, added it to make the runtest require no keyboard
1543 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1544 contains upper case chars. Make router use caseful_local_part.
1546 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1547 support when GnuTLS has been built with p11-kit.
1553 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1555 This, or similar/improved, will also be change PP/11 of 4.82.
1561 PP/01 Handle short writes when writing local log-files.
1562 In practice, only affects FreeBSD (8 onwards).
1563 Bugzilla 1053, with thanks to Dmitry Isaikin.
1565 NM/01 Bugzilla 949 - Documentation tweak
1567 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1570 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1572 PP/02 Implemented gsasl authenticator.
1574 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1576 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1577 `pkg-config foo` for cflags/libs.
1579 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1580 with rest of GSASL and with heimdal_gssapi.
1582 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1583 `pkg-config foo` for cflags/libs for the TLS implementation.
1585 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1586 properties get this fed in as external SSF. A number of robustness
1587 and debugging improvements to the cyrus_sasl authenticator.
1589 PP/08 cyrus_sasl server now expands the server_realm option.
1591 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1592 Patch by Jeremy Harris.
1594 PP/10 Added dbmjz lookup type.
1596 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1598 PP/12 MAIL args handles TAB as well as SP, for better interop with
1599 non-compliant senders.
1600 Analysis and variant patch by Todd Lyons.
1602 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1603 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1604 Patch from Dirk Mueller <dmueller@suse.com>
1606 PP/13 tls_peerdn now print-escaped for spool files.
1607 Observed some $tls_peerdn in wild which contained \n, which resulted
1608 in spool file corruption.
1610 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1611 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1612 or write after TLS renegotiation, which otherwise led to messages
1615 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1616 as a tracking header (ie: a signed header comes before the signature).
1617 Patch from Wolfgang Breyha.
1619 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1620 comma-sep list; embedded commas doubled.
1622 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1624 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1626 Report and patch from Dmitry Banschikov.
1628 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1629 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1630 failures. We appear to now support TLS1.1+ with Exim.
1632 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1633 lets Exim select keys and certificates based upon TLS SNI from client.
1634 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1635 before an outbound SMTP session. New log_selector, +tls_sni.
1637 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1638 NULL dereference. Report and patch from Alun Jones.
1640 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1641 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1642 Not seeing resolver debug output on NetBSD, but suspect this is a
1643 resolver implementation change.
1645 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1646 Left warnings. Added "eximon gdb" invocation mode.
1648 PP/21 Defaulting "accept_8bitmime" to true, not false.
1650 PP/22 Added -bw for inetd wait mode support.
1652 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1653 locate the relevant includes and libraries. Made this the default.
1655 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1656 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1658 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1659 This may cause build issues on older platforms.
1661 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1662 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1663 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1664 Added SNI support via GnuTLS too.
1665 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1667 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1669 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1670 Applied second patch from Janne, implementing suggestion to default
1671 multiple-strings-in-record handling to match SPF spec.
1673 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1675 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1676 read-only, out of scope).
1677 Patch from Wolfgang Breyha, report from Stuart Northfield.
1679 PP/29 Fix three issues highlighted by clang analyser static analysis.
1680 Only crash-plausible issue would require the Cambridge-specific
1681 iplookup router and a misconfiguration.
1682 Report from Marcin Mirosław.
1684 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1686 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1687 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1688 As part of this, removing so much warning spew let me fix some minor
1689 real issues in debug logging.
1691 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1692 assignment on my part. Fixed.
1694 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1695 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1696 Janne Snabb (who went above and beyond: thank you).
1698 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1699 string otherwise requires a connection and a bunch more work and it's
1700 relatively easy to get wrong. Should also expose TLS library linkage
1703 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1704 64-bit ${eval} (JH/03).
1706 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1707 GNU libc to support some of the 64-bit stuff, should not lead to
1708 conflicts. Defined before os.h is pulled in, so if a given platform
1709 needs to override this, it can.
1711 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1712 protection layer was required, which is not implemented.
1713 Bugzilla 1254, patch from Wolfgang Breyha.
1715 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1716 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1717 tls_dhparam take prime identifiers. Also unbreak combination of
1718 OpenSSL+DH_params+TLSSNI.
1720 PP/39 Disable SSLv2 by default in OpenSSL support.
1726 PP/01 Solaris build fix for Oracle's LDAP libraries.
1727 Bugzilla 1109, patch from Stephen Usher.
1729 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1731 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1734 TF/02 Fix a couple more cases where we did not log the error message
1735 when unlink() failed. See also change 4.74-TF/03.
1737 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1738 lock up or crash if it happened to be inside a call to libc when it
1739 got a SIGUSR1 from exiwhat.
1741 The SIGUSR1 handler appends the current process status to the process
1742 log which is later printed by exiwhat. It used to use the general
1743 purpose logging code to do this, but several functions it calls are
1744 not safe for signals.
1746 The new output code in the SIGUSR1 handler is specific to the process
1747 log, and simple enough that it's easy to inspect for signal safety.
1748 Removing some special cases also simplifies the general logging code.
1749 Removing the spurious timestamps from the process log simplifies
1752 TF/04 Improved ratelimit ACL condition.
1754 The /noupdate option has been deprecated in favour of /readonly which
1755 has clearer semantics. The /leaky, /strict, and /readonly update modes
1756 are mutually exclusive. The update mode is no longer included in the
1757 database key; it just determines when the database is updated. (This
1758 means that when you upgrade Exim will forget old rate measurements.)
1760 Exim now checks that the per_* options are used with an update mode that
1761 makes sense for the current ACL. For example, when Exim is processing a
1762 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1763 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1764 must specify per_mail/readonly. If you omit the update mode it defaults to
1765 /leaky where that makes sense (as before) or /readonly where required.
1767 The /noupdate option is now undocumented but still supported for
1768 backwards compatibility. It is equivalent to /readonly except that in
1769 ACLs where /readonly is required you may specify /leaky/noupdate or
1770 /strict/noupdate which are treated the same as /readonly.
1772 A useful new feature is the /count= option. This is a generalization
1773 of the per_byte option, so that you can measure the throughput of other
1774 aggregate values. For example, the per_byte option is now equivalent
1775 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1777 The per_rcpt option has been generalized using the /count= mechanism
1778 (though it's more complicated than the per_byte equivalence). When it is
1779 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1780 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1781 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1782 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1783 Note that using per_rcpt with a non-readonly update mode in more than
1784 one ACL will cause the recipients to be double-counted. (The per_mail
1785 and per_byte options don't have this problem.)
1787 The handling of very low rates has changed slightly. If the computed rate
1788 is less than the event's count (usually one) then this event is the first
1789 after a long gap. In this case the rate is set to the same as this event's
1790 count, so that the first message of a spam run is counted properly.
1792 The major new feature is a mechanism for counting the rate of unique
1793 events. The new per_addr option counts the number of different
1794 recipients that someone has sent messages to in the last time period. It
1795 behaves like per_rcpt if all the recipient addresses are different, but
1796 duplicate recipient addresses do not increase the measured rate. Like
1797 the /count= option this is a general mechanism, so the per_addr option
1798 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1799 example, measure the rate that a client uses different sender addresses
1800 with the options per_mail/unique=$sender_address. There are further
1801 details in the main documentation.
1803 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1805 TF/06 Removed a few PCRE remnants.
1807 TF/07 Automatically extract Exim's version number from tags in the git
1808 repository when doing development or release builds.
1810 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1811 Bugzilla 879. Patch from Paul Fisher.
1813 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1814 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1817 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1819 PP/05 Variable $av_failed, true if the AV scanner deferred.
1820 Bugzilla 1078. Patch from John Horne.
1822 PP/06 Stop make process more reliably on build failure.
1823 Bugzilla 1087. Patch from Heiko Schlittermann.
1825 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1826 Bugzilla 1089. Patch from Heiko Schlittermann.
1828 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1829 Bugzilla 1131. Patch from Holger Weiß.
1831 PP/09 Handle IPv6 addresses with SPF.
1832 Bugzilla 860. Patch from Wolfgang Breyha.
1834 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1836 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1839 PP/11 match_* no longer expand right-hand-side by default.
1840 New compile-time build option, EXPAND_LISTMATCH_RHS.
1841 New expansion conditions, "inlist", "inlisti".
1843 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1845 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1847 PP/14 fix log_write() format string regression from TF/03.
1848 Bugzilla 1152. Patch from Dmitry Isaikin.
1854 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1856 PP/02 Harmonised TLS library version reporting; only show if debugging.
1857 Layout now matches that introduced for other libraries in 4.74 PP/03.
1859 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1861 PP/04 New "dns_use_edns0" global option.
1863 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1866 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1867 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1869 TK/01 Updated PolarSSL code to 0.14.2.
1870 Bugzilla 1097. Patch from Andreas Metzler.
1872 PP/07 Catch divide-by-zero in ${eval:...}.
1873 Fixes bugzilla 1102.
1875 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1878 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1879 format-string attack -- SECURITY: remote arbitrary code execution.
1881 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1882 time unintentionally subject to list matching rules, letting the header
1883 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1884 arbitrary string expansion). This allowed for information disclosure.
1886 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1887 INT_MIN/-1 -- value coerced to INT_MAX.
1893 NM/01 Workaround for PCRE version dependency in version reporting
1896 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1897 This fixes portability to compilers other than gcc, notably
1898 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1900 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1901 makefiles for portability to HP-UX and POSIX correctness.
1903 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1904 Also via indented variable definition in the Makefile.
1905 (Debugging by Oliver Heesakkers).
1907 PP/02 Restore caching of spamd results with expanded spamd_address.
1908 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1910 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1911 Improves build reliability. Fix from: Frank Elsner
1913 NM/02 Fix wide character breakage in the rfc2047 coding
1914 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1916 NM/03 Allow underscore in dnslist lookups
1917 Fixes bug 1026. Patch from Graeme Fowler
1919 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1920 Code patches from Adam Ciarcinski of NetBSD.
1922 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1925 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1926 is logged, to avoid truncation. Patch from John Horne.
1928 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1929 Patch from Jakob Hirsch.
1931 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1932 SQL string expansion failure details.
1933 Patch from Andrey Oktyabrski.
1935 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1936 Patch from Simon Arlott.
1938 PP/09 New lookups functionality failed to compile on old gcc which rejects
1939 extern declarations in function scope.
1940 Patch from Oliver Fleischmann
1942 PP/10 Use sig_atomic_t for flags set from signal handlers.
1943 Check getgroups() return and improve debugging.
1944 Fixed developed for diagnosis in bug 927 (which turned out to be
1947 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1948 Patch from Mark Zealey.
1950 PP/12 Bugzilla 1056: Improved spamd server selection.
1951 Patch from Mark Zealey.
1953 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1954 Based on patch from Heiko Schlittermann.
1956 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1957 Patch from Uwe Doering, sign-off by Michael Haardt.
1959 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1960 variable declaration deep within a block. Bug and patch from
1963 PP/15 lookups-Makefile IRIX compatibility coercion.
1965 PP/16 Make DISABLE_DKIM build knob functional.
1967 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1968 Patch by Simon Arlott
1970 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1971 variable argument macros. Our copy now differs from upstream.
1977 TF/01 Failure to get a lock on a hints database can have serious
1978 consequences so log it to the panic log.
1980 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1981 controlled using the smtp_confirmation log selector.
1983 TF/03 Include the error message when we fail to unlink a spool file.
1985 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1986 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1987 for maintaining out-of-tree patches for some time.
1989 PP/01 Bugzilla 139: Documentation and portability issues.
1990 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1991 Handle per-OS dynamic-module compilation flags.
1993 PP/02 Let /dev/null have normal permissions.
1994 The 4.73 fixes were a little too stringent and complained about the
1995 permissions on /dev/null. Exempt it from some checks.
1996 Reported by Andreas M. Kirchwitz.
1998 PP/03 Report version information for many libraries, including
1999 Exim version information for dynamically loaded libraries. Created
2000 version.h, now support a version extension string for distributors
2001 who patch heavily. Dynamic module ABI change.
2003 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2004 privilege escalation vulnerability whereby the Exim run-time user
2005 can cause root to append content of the attacker's choosing to
2008 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2011 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2012 If dropping privileges for untrusted macros, we disabled normal logging
2013 on the basis that it would fail; for the Exim run-time user, this is not
2014 the case, and it resulted in successful deliveries going unlogged.
2015 Fixed. Reported by Andreas Metzler.
2021 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2022 only prepend for the Resent-* case. Fixes regression introduced in
2023 Exim 4.70 by NM/22 for Bugzilla 607.
2025 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2026 increasing numbers of administrators be bitten by this.
2028 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2030 PP/03 Bugzilla 994: added openssl_options main configuration option.
2032 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2034 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2036 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2038 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2039 without a peer certificate, leading to a segfault because of an
2040 assumption that peers always have certificates. Be a little more
2041 paranoid. Problem reported by Martin Tscholak.
2043 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2044 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2045 NB: ClamAV planning to remove STREAM in "middle of 2010".
2046 CL also introduces -bmalware, various -d+acl logging additions and
2047 more caution in buffer sizes.
2049 PP/09 Implemented reverse_ip expansion operator.
2051 PP/10 Bugzilla 937: provide a "debug" ACL control.
2053 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2055 PP/12 Bugzilla 973: Implement --version.
2057 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2059 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2061 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2063 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2064 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2065 ignore trailing whitespace.
2067 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2069 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2072 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2073 Notification from Dr Andrew Aitchison.
2075 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2076 ExtendedDetectionInfo response format.
2077 Notification from John Horne.
2079 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2082 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2083 XSL and documented dependency on system catalogs, with examples of how
2086 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2089 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2090 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2091 configuration file which is writeable by the Exim user or group.
2093 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2094 of configuration files to cover files specified with the -C option if
2095 they are going to be used with root privileges, not just the default
2098 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2099 option (effectively making it always true).
2101 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2102 files to be used while preserving root privileges.
2104 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2105 that rogue child processes cannot use them.
2107 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2108 run-time user, instead of root.
2110 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2111 Exim run-time user without dropping privileges.
2113 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2114 result string, instead of calling string_vformat() twice with the same
2117 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2118 for other users. Others should always drop root privileges if they use
2119 -C on the command line, even for a whitelisted configure file.
2121 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2123 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2129 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2130 $data_path, and $header_path variables; fixed documentation bugs and
2133 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2134 exipick to access non-standard spools, including the "frozen" queue
2137 NM/01 Bugzilla 965: Support mysql stored procedures.
2138 Patch from Alain Williams
2140 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2142 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2143 Patch from Andreas Metzler
2145 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2146 Patch from Kirill Miazine
2148 NM/05 Bugzilla 671: Added umask to procmail example.
2150 JJ/03 installed exipick 20100323.0, fixing doc bug
2152 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2153 directory. Notification and patch from Dan Rosenberg.
2155 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2157 TK/02 Improve log output when DKIM signing operation fails.
2159 MH/01 Treat the transport option dkim_domain as a colon separated
2160 list, not as a single string, and sign the message with each element,
2161 omitting multiple occurences of the same signer.
2163 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2164 Bugzilla 985, 986. Patch by Simon Arlott
2166 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2167 Patch by Simon Arlott
2169 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2170 MBX locking. Notification from Dan Rosenberg.
2176 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2178 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2180 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2182 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2184 NM/04 Bugzilla 915: Flag broken perl installation during build.
2190 TK/01 Added patch by Johannes Berg that expands the main option
2191 "spamd_address" if it starts with a dollar sign.
2193 TK/02 Write list of recipients to X-Envelope-Sender header when building
2194 the mbox-format spool file for content scanning (suggested by Jakob
2197 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2198 (http://www.dcc-servers.net/) support via dccifd. Activated by
2199 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2201 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2202 by Mark Daniel Reidel <mr@df.eu>.
2204 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2205 When building exim an external PCRE library is now needed -
2206 PCRE is a system library on the majority of modern systems.
2207 See entry on PCRE_LIBS in EDITME file.
2209 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2210 conversation. Added nologin parameter to request.
2211 Patch contributed by Kirill Miazine.
2213 TF/01 Do not log submission mode rewrites if they do not change the address.
2215 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2217 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2218 log files in place. Contributed by Roberto Lima.
2220 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2222 TF/03 Bugzilla 615: When checking the local_parts router precondition
2223 after a local_part_suffix or local_part_prefix option, Exim now
2224 does not use the address's named list lookup cache, since this
2225 contains cached lookups for the whole local part.
2227 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2228 Robert Millan. Documentation is in experimental-spec.txt.
2230 TF/04 Bugzilla 668: Fix parallel build (make -j).
2232 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2234 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2235 Patch provided by Jan Srzednicki.
2237 TF/05 Leading white space used to be stripped from $spam_report which
2238 wrecked the formatting. Now it is preserved.
2240 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2241 that they are available at delivery time.
2243 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2245 TF/08 TLS error reporting now respects the incoming_interface and
2246 incoming_port log selectors.
2248 TF/09 Produce a more useful error message if an SMTP transport's hosts
2249 setting expands to an empty string.
2251 NM/06 Bugzilla 744: EXPN did not work under TLS.
2252 Patch provided by Phil Pennock.
2254 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2255 Patch provided by Richard Godbee.
2257 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2258 acl_smtp_notquit, added index entry.
2260 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2261 Patch provided by Eugene Bujak.
2263 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2264 accept(). Patch provided by Maxim Dounin.
2266 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2267 Patch provided by Phil Pennock.
2269 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2271 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2272 Patch provided by Brad "anomie" Jorsch.
2274 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2275 Patch provided by Dean Brooks.
2277 TK/05 Add native DKIM support (does not depend on external libraries).
2279 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2280 Patch provided by Graeme Fowler.
2282 NM/16 Bugzilla 851: Documentation example syntax fix.
2284 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2286 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2289 NM/19 Bugzilla 745: TLS version reporting.
2290 Patch provided by Phil Pennock.
2292 NM/20 Bugzilla 167: bool: condition support.
2293 Patch provided by Phil Pennock.
2295 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2296 clients. Patch provided by Phil Pennock.
2298 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2299 Patch provided by Brad "anomie" Jorsch.
2301 NM/23 Bugzilla 687: Fix misparses in eximstats.
2302 Patch provided by Heiko Schlittermann.
2304 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2305 Patch provided by Heiko Schlittermann.
2307 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2308 plus update to original patch.
2310 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2312 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2313 Patch provided by David Brownlee.
2315 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2317 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2319 NM/30 Bugzilla 888: TLS documentation bugfixes.
2321 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2323 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2324 Unlike the original bugzilla I have changed all shell scripts in src tree.
2326 NM/33 Bugzilla 898: Transport filter timeout fix.
2327 Patch by Todd Rinaldo.
2329 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2330 Patch by Serge Demonchaux.
2332 NM/35 Bugzilla 39: Base64 decode bug fixes.
2333 Patch by Jakob Hirsch.
2335 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2337 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2339 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2341 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2347 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2348 ALT-N's libdkim that I have put here:
2349 http://duncanthrax.net/exim-experimental/
2351 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2352 were called 'true' and it seems that C99 defines that as a reserved
2353 keyword to be used with 'bool' variable types. That means you could
2354 not include C99-style headers which use bools without triggering
2355 build errors in sieve.c.
2357 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2358 as mailq or other aliases. Changed the --help handling significantly
2359 to do whats expected. exim_usage() emits usage/help information.
2361 SC/01 Added the -bylocaldomain option to eximstats.
2363 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2365 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2367 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2373 PH/01 Another patch from the Sieve maintainer.
2375 PH/02 When an IPv6 address is converted to a string for single-key lookup
2376 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2377 dots are used instead of colons so that keys in lsearch files need not
2378 contain colons. This was done some time before quoting was made available
2379 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2380 (notated using the quote facility) so as to distinguish them from IPv4
2381 keys. This meant that lookups for IP addresses in host lists did not work
2382 for iplsearch lookups.
2384 This has been fixed by arranging for IPv6 addresses to be expressed with
2385 colons if the lookup type is iplsearch. This is not incompatible, because
2386 previously such lookups could never work.
2388 The situation is now rather anomalous, since one *can* have colons in
2389 ordinary lsearch keys. However, making the change in all cases is
2390 incompatible and would probably break a number of configurations.
2392 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2395 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2396 conversion specification without a maximum field width, thereby enabling
2397 a rogue spamd server to cause a buffer overflow. While nobody in their
2398 right mind would setup Exim to query an untrusted spamd server, an
2399 attacker that gains access to a server running spamd could potentially
2400 exploit this vulnerability to run arbitrary code as the Exim user.
2402 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2403 $primary_hostname instead of what libspf2 thinks the hosts name is.
2405 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2406 a directory entry by the name of the lookup key. Previously, if a
2407 symlink pointed to a non-existing file or a file in a directory that
2408 Exim lacked permissions to read, a lookup for a key matching that
2409 symlink would fail. Now it is enough that a matching directory entry
2410 exists, symlink or not. (Bugzilla 503.)
2412 PH/03 The body_linecount and body_zerocount variables are now exported in the
2415 PH/04 Added the $dnslist_matched variable.
2417 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2418 This means they are set thereafter only if the connection becomes
2421 PH/06 Added the client_condition to authenticators so that some can be skipped
2422 by clients under certain conditions.
2424 PH/07 The error message for a badly-placed control=no_multiline_responses left
2425 "_responses" off the end of the name.
2427 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2429 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2430 (without spaces) instead of just copying the configuration text.
2432 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2434 PH/11 Added $max_received_linelength.
2436 PH/12 Added +ignore_defer and +include_defer to host lists.
2438 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2439 way in which PCRE > 7.0 is built.
2441 PH/14 Implemented queue_only_load_latch.
2443 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2444 MAIL command. The effect was to mangle the value on 64-bit systems.
2446 PH/16 Another patch from the Sieve maintainer.
2448 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2450 PH/18 If a system quota error occurred while trying to create the file for
2451 a maildir delivery, the message "Mailbox is full" was not appended to the
2452 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2453 only to a quota excession during the actual writing of the file.
2455 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2456 characters?) which causes problems in log lines. The DN values are now
2457 passed through string_printing() before being added to log lines.
2459 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2460 and InterBase are left for another time.)
2462 PH/21 Added message_body_newlines option.
2464 PH/22 Guard against possible overflow in moan_check_errorcopy().
2466 PH/23 POSIX allows open() to be a macro; guard against that.
2468 PH/24 If the recipient of an error message contained an @ in the local part
2469 (suitably quoted, of course), incorrect values were put in $domain and
2470 $local_part during the evaluation of errors_copy.
2476 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2477 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2480 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2481 issue a MAIL command.
2483 PH/02 In an ACL statement such as
2485 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2487 if a client was not listed at all, or was listed with a value other than
2488 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2489 the condition was not true (as it should be), so access was not denied.
2490 The bug was that the ! inversion was incorrectly passed on to the second
2491 item. This has been fixed.
2493 PH/03 Added additional dnslists conditions == and =& which are different from
2494 = and & when the dns lookup returns more than one IP address.
2496 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2497 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2499 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2500 FSYNC, which compiles an option called disable_fsync that allows for
2501 bypassing fsync(). The documentation is heavily laced with warnings.
2503 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2505 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2506 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2507 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2508 including adding "make clean"; (3) Added -fPIC when compiling the test
2509 dynamically loaded module, to get rid of a warning.
2511 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2512 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2513 = 0s. The bug is otherwise harmless.
2515 PH/07 There was a bug in the dovecot authenticator such that the value of
2516 $auth1 could be overwritten, and so not correctly preserved, after a
2517 successful authentication. This usually meant that the value preserved by
2518 the server_setid option was incorrect.
2520 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2522 PH/09 Installed PCRE release 7.0.
2524 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2525 run for batched SMTP input. It is now run at the start of every message
2526 in the batch. While fixing this I discovered that the process information
2527 (output by running exiwhat) was not always getting set for -bs and -bS
2528 input. This is fixed, and it now also says "batched" for BSMTP.
2530 PH/11 Added control=no_pipelining.
2532 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2533 patch, slightly modified), and move the expansion of helo_data till after
2534 the connection is made in the smtp transport (so it can use these
2537 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2539 PH/14 Added log_selector = +pid.
2541 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2543 PH/16 Add ${if forany and ${if forall.
2545 PH/17 Added dsn_from option to vary the From: line in DSNs.
2547 PH/18 Flush SMTP output before performing a callout, unless control =
2548 no_callout_flush is set.
2550 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2551 was true (the default) a successful delivery failed to delete the retry
2552 item, thus causing premature timeout of the address. The bug is now
2555 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2557 PH/21 Long custom messages for fakedefer and fakereject are now split up
2558 into multiline responses in the same way that messages for "deny" and
2559 other ACL rejections are.
2561 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2562 with slight modification.
2564 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2565 draft, changing the syntax and factoring some duplicate code".
2567 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2568 for deliveries of the second and subsequent messages over the same SMTP
2571 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2572 ${reduce, with only minor "tidies".
2574 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2576 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2577 expansion side effects.
2579 PH/27 When a message times out after an over-quota error from an Exim-imposed
2580 quota, the bounce message says "mailbox is full". This message was not
2581 being given when it was a system quota that was exceeded. It now should
2584 MH/03 Made $recipients available in local_scan(). local_scan() already has
2585 better access to the recipient list through recipients_list[], but
2586 $recipients can be useful in postmaster-provided expansion strings.
2588 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2589 in the case of a MAIL command with additional options following the
2590 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2591 were accidentally chopped off.
2593 PH/29 SMTP synchronization checks are implemented when a command is read -
2594 there is a check that no more input is waiting when there shouldn't be
2595 any. However, for some commands, a delay in an ACL can mean that it is
2596 some time before the response is written. In this time, more input might
2597 arrive, invalidly. So now there are extra checks after an ACL has run for
2598 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2599 pipelining has not been advertised.
2601 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2603 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2604 returned in the SMTP response when smtp_return_error_details was false.
2605 This has been fixed.
2607 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2608 instead of the C I/O that was originally supplied, because problems were
2609 reported on Solaris.
2611 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2612 Exim which did not show up earlier: it was assuming that a call to
2613 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2614 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2615 was a macro that became an assignment, so it seemed to work. This has
2616 changed to a proper function call with a void return, hence the compile
2617 error. Exim's code has been fixed.
2619 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2622 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2624 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2626 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2627 because it turns out to be too big or there is a timeout) there is no
2628 "Completed" line in the log. When some messages of this type were
2629 selected by exigrep, they were listed as "not completed". Others were
2630 picked up by some special patterns. I have improved the selection
2631 criteria to be more general.
2633 PH/38 The host_find_failed option in the manualroute router can now be set
2634 to "ignore", to completely ignore a host whose IP address cannot be
2635 found. If all hosts are ignored, the behaviour is controlled by the new
2636 host_all_ignored option.
2638 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2639 homing or because of multiple MX records with /mx) generated more than
2640 one IP address, and the following item turned out to be the local host,
2641 all the secondary addresses of the first item were incorrectly removed
2642 from the list, along with the local host and any following hosts (which
2643 is what is supposed to happen).
2645 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2646 whoever called Exim into the -H file. In the case of the daemon it was
2647 behaving confusingly. When first started, it used values for whoever
2648 started the daemon, but after a SIGHUP it used the Exim user (because it
2649 calls itself on a restart). I have changed the code so that it now always
2652 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2653 message are rejected with the same error (e.g. no authentication or bad
2654 sender address), and a DATA command is nevertheless sent (as can happen
2655 with PIPELINING or a stupid MUA), the error message that was given to the
2656 RCPT commands is included in the rejection of the DATA command. This is
2657 intended to be helpful for MUAs that show only the final error to their
2660 PH/42 Another patch from the Sieve maintainer.
2662 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2663 Eximstats - Fixed some broken HTML links and added missing column headers
2665 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2666 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2668 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2670 PH/43 Yet another patch from the Sieve maintainer.
2672 PH/44 I found a way to check for a TCP/IP connection going away before sending
2673 the response to the final '.' that terminates a message, but only in the
2674 case where the client has not sent further data following the '.'
2675 (unfortunately, this is allowed). However, in many cases there won't be
2676 any further data because there won't be any more messages to send. A call
2677 to select() can be used: if it shows that the input is "ready", there is
2678 either input waiting, or the socket has been closed. An attempt to read
2679 the next input character can distinguish the two cases. Previously, Exim
2680 would have sent an OK response which the client would never have see.
2681 This could lead to message repetition. This fix should cure that, at
2682 least in a lot of common cases.
2684 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2685 advertised in response to EHLO.
2691 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2692 fixed by 4.65/MH/01 (is this a record?) are fixed:
2694 (i) An empty string was always treated as zero by the numeric comparison
2695 operators. This behaviour has been restored.
2697 (ii) It is documented that the numeric comparison operators always treat
2698 their arguments as decimal numbers. This was broken in that numbers
2699 starting with 0 were being interpreted as octal.
2701 While fixing these problems I realized that there was another issue that
2702 hadn't been noticed. Values of message_size_limit (both the global option
2703 and the transport option) were treated as octal if they started with 0.
2704 The documentation was vague. These values are now always treated as
2705 decimal, and I will make that clear in the documentation.
2711 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2712 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2715 MH/01 Don't check that the operands of numeric comparison operators are
2716 integers when their expansion is in "skipping" mode (fixes bug
2717 introduced by 4.64-PH/07).
2719 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2720 child addresses, Exim now panics and dies. Previously, because the count
2721 is held in a short int, deliveries were likely to be lost. As such a
2722 large number of recipients for a single message is ridiculous
2723 (performance will be very, very poor), I have chosen to impose a limit
2724 rather than extend the field.
2730 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2731 leftover -K file (the existence of which was triggered by #402).
2732 While we were at it, introduced process PID as part of the -K
2733 filename. This should rule out race conditions when creating
2736 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2737 processing considerably. Previous code took too long for large mails,
2738 triggering a timeout which in turn triggers #401.
2740 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2741 in the DK code in transports.c. sendfile() is not really portable,
2742 hence the _LINUX specificness.
2744 TF/01 In the add_headers option to the mail command in an Exim filter,
2745 there was a bug that Exim would claim a syntax error in any
2746 header after the first one which had an odd number of characters
2749 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2750 callout verification, Exim cached a "reject" for the entire domain. This
2751 is correct for most verifications, but it is not correct for a recipient
2752 verification with use_sender or use_postmaster set, because in that case
2753 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2754 case of MAIL FROM:<> rejection from other early rejections (e.g.
2755 rejection of HELO). When verifying a recipient using a non-null MAIL
2756 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2757 Whatever the result of the callout, the value of the domain cache is
2758 left unchanged (for any other kind of callout, getting as far as trying
2759 RCPT means that the domain itself is ok).
2761 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2764 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2765 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2766 session, the daemon ignores ECONNECT errors and logs others; it now
2767 ignores EPIPE as well.
2769 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2770 (quoted-printable decoding).
2772 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2773 later the small subsequent patch to fix an introduced bug.
2775 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2777 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2779 PH/08 An error is now given if message_size_limit is specified negative.
2781 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2782 to be given (somewhat) arbitrary names.
2784 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2787 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2788 miscellaneous code fixes
2790 PH/10 Added the log_reject_target ACL modifier to specify where to log
2793 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2794 hostname. This is wrong, because it relates to the incoming message (and
2795 probably the interface on which it is arriving) and not to the outgoing
2796 callout (which could be using a different interface). This has been
2797 changed to use the value of the helo_data option from the smtp transport
2798 instead - this is what is used when a message is actually being sent. If
2799 there is no remote transport (possible with a router that sets up host
2800 addresses), $smtp_active_hostname is used.
2802 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2803 tweaks were necessary in order to get it to work (see also 21 below):
2804 (a) The code assumed that strncpy() returns a negative number on buffer
2805 overflow, which isn't the case. Replaced with Exim's string_format()
2807 (b) There were several signed/unsigned issues. I just did the minimum
2808 hacking in of casts. There is scope for a larger refactoring.
2809 (c) The code used strcasecmp() which is not a standard C function.
2810 Replaced with Exim's strcmpic() function.
2811 (d) The code set only $1; it now sets $auth1 as well.
2812 (e) A simple test gave the error "authentication client didn't specify
2813 service in request". It would seem that Dovecot has changed its
2814 interface. Fortunately there's a specification; I followed it and
2815 changed what the client sends and it appears to be working now.
2817 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2820 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2821 address A is aliased to B and C, where B exists and C does not. Without
2822 -v the output is "A verified" because verification stops after a
2823 successful redirection if more than one address is generated. However,
2824 with -v the child addresses are also verified. Exim was outputting "A
2825 failed to verify" and then showing the successful verification for C,
2826 with its parentage. It now outputs "B failed to verify", showing B's
2827 parentage before showing the successful verification of C.
2829 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2830 look up a TXT record in a specific list after matching in a combined
2833 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2834 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2835 they consult the DNS. I had assumed they would set it the way they
2836 wanted; and indeed my experiments on Linux seem to show that in some
2837 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2838 To be on the safe side, however, I have now made the interface to
2839 host_find_byname() similar to host_find_bydns(), with an argument
2840 containing the DNS resolver options. The host_find_byname() function now
2841 sets these options at its start, just as host_find_bydns() does. The smtp
2842 transport options dns_qualify_single and dns_search_parents are passed to
2843 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2844 of host_find_byname() use the default settings of RES_DEFNAMES
2845 (qualify_single) but not RES_DNSRCH (search_parents).
2847 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2848 spool_read_header() do less string testing, by means of a preliminary
2849 switch on the second character of optional "-foo" lines. (This is
2850 overdue, caused by the large number of possibilities that now exist.
2851 Originally there were few.) While I was there, I also converted the
2852 str(n)cmp tests so they don't re-test the leading "-" and the first
2853 character, in the hope this might squeeze out yet more improvement.
2855 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2856 flag allowing group syntax was set by the header_syntax check but not
2857 turned off, possible causing trouble later; (2) The flag was not being
2858 set at all for the header_verify test, causing "group"-style headers to
2859 be rejected. I have now set it in this case, and also caused header_
2860 verify to ignore an empty address taken from a group. While doing this, I
2861 came across some other cases where the code for allowing group syntax
2862 while scanning a header line wasn't quite right (mostly, not resetting
2863 the flag correctly in the right place). These bugs could have caused
2864 trouble for malformed header lines. I hope it is now all correct.
2866 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2867 with the "reply" argument non-NULL. The code, however (which originally
2868 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2869 but it didn't always do it. This confused somebody who was copying the
2870 code for some other use. I have removed all the tests.
2872 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2873 feature that was used to support insecure browsers during the U.S. crypto
2874 embargo. It requires special client support, and Exim is probably the
2875 only MTA that supported it -- and would never use it because real RSA is
2876 always available. This code has been removed, because it had the bad
2877 effect of slowing Exim down by computing (never used) parameters for the
2878 RSA_EXPORT functionality.
2880 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2881 authenticator to fail if there's a tab character in the incoming data
2882 (there should never be unless someone is messing about, as it's supposed
2883 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2884 if the connection is using TLS or if the remote IP is the same as the
2885 local IP, and the "valid-client-cert option" if a client certificate has
2888 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2889 authenticators. This can be used for authorization after authentication
2890 succeeds. (In the case of plaintext, it servers for both authentication
2893 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2894 if any retry times were supplied.
2896 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2897 connection, where there is no client IP address to check. In this
2898 situation, the verify now always succeeds.
2900 PH/25 Applied John Jetmore's -Mset patch.
2902 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2904 PH/27 In a string expansion for a processed (not raw) header when multiple
2905 headers of the same name were present, leading whitespace was being
2906 removed from all of them, but trailing whitespace was being removed only
2907 from the last one. Now trailing whitespace is removed from each header
2908 before concatenation. Completely empty headers in a concatenation (as
2909 before) are ignored.
2911 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2912 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2914 PH/29 [Removed. This was a change that I later backed out, and forgot to
2915 correct the ChangeLog entry (that I had efficiently created) before
2916 committing the later change.]
2918 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2919 address errors (4xx response to RCPT) over the same connection as other
2920 messages routed to the same hosts. Such deliveries are always "forced",
2921 so retry times are not inspected. This resulted in far too many retries
2922 for the affected addresses. The effect occurred only when there were more
2923 hosts than the hosts_max_try setting in the smtp transport when it had
2924 the 4xx errors. Those hosts that it had tried were not added to the list
2925 of hosts for which the message was waiting, so if all were tried, there
2926 was no problem. Two fixes have been applied:
2928 (i) If there are any address or message errors in an SMTP delivery, none
2929 of the hosts (tried or untried) are now added to the list of hosts
2930 for which the message is waiting, so the message should not be a
2931 candidate for sending over the same connection that was used for a
2932 successful delivery of some other message. This seems entirely
2933 reasonable: after all the message is NOT "waiting for some host".
2934 This is so "obvious" that I'm not sure why it wasn't done
2935 previously. Hope I haven't missed anything, but it can't do any
2936 harm, as the worst effect is to miss an optimization.
2938 (ii) If, despite (i), such a delivery is accidentally attempted, the
2939 routing retry time is respected, so at least it doesn't keep
2940 hammering the server.
2942 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2943 in ${readsocket because some servers need this prod.
2945 PH/32 Added some extra debug output when updating a wait-xxx database.
2947 PH/33 The hint "could be header name not terminated by colon", which has been
2948 given for certain expansion errors for a long time, was not being given
2949 for the ${if def:h_colon_omitted{... case.
2951 PH/34 The spec says: "With one important exception, whenever a domain list is
2952 being scanned, $domain contains the subject domain." There was at least
2953 one case where this was not true.
2955 PH/35 The error "getsockname() failed: connection reset by peer" was being
2956 written to the panic log as well as the main log, but it isn't really
2957 panic-worthy as it just means the connection died rather early on. I have
2958 removed the panic log writing for the ECONNRESET error when getsockname()
2961 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2962 runs only) independently of the message's sender address. This meant
2963 that, if the 4xx error was in fact related to the sender, a different
2964 message to the same recipient with a different sender could confuse
2965 things. In particular, this can happen when sending to a greylisting
2966 server, but other circumstances could also provoke similar problems.
2967 I have changed the default so that the retry time for these errors is now
2968 based a combination of the sender and recipient addresses. This change
2969 can be overridden by setting address_retry_include_sender=false in the
2972 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2973 remote server are returned as part of bounce messages. This was not
2974 happening for LMTP over a pipe (the lmtp transport), but now it is the
2975 same for both kinds of LMTP.
2977 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2978 in header lines that were in fact CNAMEs. This is no longer the case.
2980 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2981 and queue runs started by the daemon processed all messages. This has
2982 been fixed so that -R and -S can now usefully be given with -q<time>.
2984 PH/40 Import PCRE release 6.7 (fixes some bugs).
2986 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2988 PH/42 Give an error if -q is specified more than once.
2990 PH/43 Renamed the variables $interface_address and $interface_port as
2991 $received_ip_address and $received_port, to make it clear that these
2992 values apply to message reception, and not to the outgoing interface when
2993 a message is delivered. (The old names remain recognized, of course.)
2995 PH/44 There was no timeout on the connect() call when using a Unix domain
2996 socket in the ${readsocket expansion. There now is.
2998 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2999 be meaningful with "accept".
3001 SC/01 Eximstats V1.43
3002 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3004 SC/02 Eximstats V1.44
3005 Use a glob alias rather than an array ref in the generated
3006 parser. This improves both readability and performance.
3008 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3009 Collect SpamAssassin and rejection statistics.
3010 Don't display local sender or destination tables unless
3011 there is data to show.
3012 Added average volumes into the top table text output.
3014 SC/04 Eximstats V1.46
3015 Collect data on the number of addresses (recipients)
3016 as well as the number of messages.
3018 SC/05 Eximstats V1.47
3019 Added 'Message too big' to the list of mail rejection
3020 reasons (thanks to Marco Gaiarin).
3022 SC/06 Eximstats V1.48
3023 Mainlog lines which have GMT offsets and are too short to
3024 have a flag are now skipped.
3026 SC/07 Eximstats V1.49 (Alain Williams)
3027 Added the -emptyok flag.
3029 SC/08 Eximstats V1.50
3030 Fixes for obtaining the IP address from reject messages.
3032 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3033 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3034 whitespace changes from 4.64-PH/27
3036 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3039 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3040 are found, allow negative numbers in numeric criteria)
3042 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3044 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3045 to match changes made in 4.64-PH/43
3047 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3049 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3050 "domain" is actually an IP address, and force a failure. This locks out
3051 those revolvers/nameservers that support "A-for-A" lookups, in
3052 contravention of the specifications.
3054 PH/48 When a host name was looked up from an IP address, and the subsequent
3055 forward lookup of the name timed out, the host name was left in
3056 $sender_host_name, contrary to the specification.
3058 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3059 restricted to single-key lookups, Exim was not diagnosing an error if
3060 * or *@ was used with a query-style lookup.
3062 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3064 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3065 long ago, but noone interested enough thought of it. Let's just say that
3066 the "1.1" means that there are some new functions that weren't there at
3067 some point in the past.
3069 PH/51 Error processing for expansion failure of helo_data from an smtp
3070 transport during callout processing was broken.
3072 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3073 tested/used via the -bh/-bhc/-bs options.
3075 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3076 bug, fixed in subsequent PCRE releases).
3078 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3079 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3081 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3087 SC/01 Use a glob alias rather than an array ref in eximstats generated
3088 parser. This improves both readability and performance.
3090 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3091 Don't display local sender or destination tables in eximstats unless
3092 there is data to show.
3093 Added average volumes into the eximstats top table text output.
3095 SC/03 Collect data on the number of addresses (recipients) as well
3096 as the number of messages in eximstats.
3098 TF/01 Correct an error in the documentation for the redirect router. Exim
3099 does (usually) call initgroups() when daemonizing.
3101 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3102 with consistent privilege compared to when running as a daemon.
3104 TF/03 Note in the spec that $authenticated_id is not set for local
3105 submissions from trusted users.
3107 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3108 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3110 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3111 by adding some example configuration directives to the default
3112 configuration file. A little bit of work is required to uncomment the
3113 directives and define how usernames and passwords are checked, but
3114 there is now a framework to start from.
3116 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3117 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3118 without this. I don't know how relevant this is to other LDAP libraries.
3120 PH/02 Add the verb name to the "unknown ACL verb" error.
3122 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3124 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3126 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3127 directory not expanded when it should be if an expanded home directory
3128 was set for the address (which is overridden by the transport).
3130 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3133 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3134 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3135 because it is too late at that time, and has no effect.
3137 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3138 security issue with \' (bugzilla #107). I could not use the
3139 PQescapeStringConn() function, because it needs a PGconn value as one of
3142 PH/08 When testing addresses using -bt, indicate those final addresses that
3143 are duplicates that would not cause an additional delivery. At least one
3144 person was confused, thinking that -bt output corresponded to deliveries.
3145 (Suppressing duplicates isn't a good idea as you lose the information
3146 about possibly different redirections that led to the duplicates.)
3148 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3149 systems where poll() doesn't work, in particular OS X.
3151 PH/10 Added more information to debugging output for retry time not reached.
3153 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3154 operations in malware.c.
3156 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3159 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3160 syslog_duplication set false, log lines that would normally be written
3161 both the the main log and to the reject log were not written to syslog at
3164 PH/14 In the default configuration, change the use of "message" in ACL warn
3165 statements to "add_header".
3167 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3168 not followed by a command (e.g. "seen endif").
3170 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3171 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3174 PH/17 Added extra conditions to the default value of delay_warning_condition
3178 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3179 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3180 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3183 The Auto-Submitted: and various List- headers are standardised, whereas I
3184 don't think Precedence: ever was.
3186 PH/18 Refactored debugging code in route_finduser() to show more information,
3187 in particular, the error code if getpwnam() issues one.
3189 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3190 This is apparently needed in addition to the PH/07 change above to avoid
3191 any possible encoding problems.
3193 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3194 but not after initializing Perl.
3196 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3197 output them only if debugging. By default they are written stderr,
3198 apparently, which is not desirable.
3200 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3203 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3206 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3208 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3209 authenticated or an ident call has been made. Suppress the default
3210 values for $authenticated_id and $authenticated_sender (but permit -oMai
3211 and -oMas) when testing with -bh.
3213 PH/24 Re-jigged the order of the tests in the default configuration so that the
3214 tests for valid domains and recipients precede the DNS black list and CSA
3215 tests, on the grounds that those ones are more expensive.
3217 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3218 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3219 This bug exists in every version of Exim that I still have, right back to
3222 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3223 However, an attempt to turn on case-sensitivity in a regex key by
3224 including (?-i) didn't work because the subject string was already
3225 lowercased, and the effects were non-intuitive. It turns out that a
3226 one-line patch can be used to allow (?-i) to work as expected.
3232 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3233 other effects) broke the use of negated acl sub-conditions.
3235 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3238 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3239 "Deny" causes Exim to reject the incoming connection with a 554 error.
3240 Unfortunately, if there is a major crisis, such as a disk failure,
3241 tcp-wrappers gives "deny", whereas what one would like would be some
3242 kind of temporary error. A kludge has been added to help with this.
3243 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3244 554 error is used if errno is still zero or contains ENOENT (which occurs
3245 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3248 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3250 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3251 errors. Otherwise a message that provokes a temporary error (when other
3252 messages do not) can cause a whole host to time out.
3254 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3255 addresses were routed directly to files or pipes from a redirect router.
3256 File deliveries just didn't batch; pipe deliveries might have suffered
3259 PH/06 A failure to get a lock for a hints database would erroneously always say
3260 "Failed to get write lock", even when it was really a read lock.
3262 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3263 of 0600. This has been changed to use the value of the lockfile_mode
3264 option (which defaults to 0600).
3266 PH/08 Applied small patch from the Sieve maintainer.
3268 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3269 folder from quota calculations, a direct delivery into this folder messed
3270 up the contents of the maildirsize file. This was because the regex was
3271 used only to exclude .Trash (or whatever) when the size of the mailbox
3272 was calculated. There was no check that a delivery was happening into an
3273 excluded directory. This bug has been fixed by ignoring all quota
3274 processing for deliveries into excluded directories.
3276 PH/10 Added the maildirfolder_create_regex option to appendfile.
3282 PH/01 The code for finding all the local interface addresses on a FreeBSD
3283 system running IPv6 was broken. This may well have applied to all BSD
3284 systems, as well as to others that have similar system calls. The broken
3285 code found IPv4 interfaces correctly, but gave incorrect values for the
3286 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3287 that it would not match correctly against @[] and not recognize the IPv6
3290 PH/02 The ipliteral router was not recognizing addresses of the form user@
3291 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3293 PH/03 Added disable_ipv6.
3295 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3296 decoded form, because it is most often used to construct To: headers
3297 lines in autoreplies, and the decoded form may well be syntactically
3298 invalid. However, $reply_address has leading white space removed, and all
3299 newlines turned into spaces so that the autoreply transport does not
3302 PH/05 If group was specified without a user on a router, and no group or user
3303 was specified on a transport, the group from the router was ignored.
3305 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3306 for visible compile-time settings that can be used to change these
3307 numbers, for those that want even more. Backwards compatibility with old
3308 spool files has been maintained. However, going back to a previous Exim
3309 release will lost any variables that are in spool files.
3311 PH/07 Two small changes when running in the test harness: increase delay when
3312 passing a TCP/IP connection to a new process, in case the original
3313 process has to generate a bounce, and remove special handling of
3314 127.0.0.2 (sic), which is no longer necessary.
3316 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3317 be the same on different OS.
3319 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3322 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3323 whether --show-vars was specified or not
3325 JJ/02 exipick: Added support for new ACL variable spool format introduced
3328 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3329 syntactically invalid From: or Reply-to: line, and a filter used this to
3330 generate an autoreply, and therefore failed to obtain an address for the
3331 autoreply, Exim could try to deliver to a non-existent relative file
3332 name, causing unrelated and misleading errors. What now happens is that
3333 it logs this as a hard delivery error, but does not attempt to create a
3336 PH/11 The exinext utility has a -C option for testing purposes, but although
3337 the given file was scanned by exinext itself; it wasn't being passed on
3338 when Exim was called.
3340 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3341 an end-of-file indication when reading a command response.
3343 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3344 compiled. In many other places in Exim, IPv6 addresses are always
3345 recognized, so I have changed this. It also means that IPv4 domain
3346 literals of the form [IPV4:n.n.n.n] are now always recognized.
3348 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3349 used if the router is not running as root, for example, when verifying at
3350 ACL time, or when using -bh. The debugging output from this situation was
3351 non-existent - all you got was a failure to exec. I have made two
3354 (a) Failures to set uid/gid, the current directory, or a process leader
3355 in a subprocess such as that created by queryprogram now generate
3356 suitable debugging output when -d is set.
3358 (b) The queryprogram router detects when it is not running as root,
3359 outputs suitable debugging information if -d is set, and then runs
3360 the subprocess without attempting to change uid/gid.
3362 PH/15 Minor change to Makefile for building test_host (undocumented testing
3365 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3366 additional section of a DNS packet that returns MX or SRV records.
3367 Instead, it always explicitly searches for A/AAAA records. This avoids
3368 major problems that occur when a DNS server includes only records of one
3369 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3370 fixed another bug: if SRV records were looked up and the corresponding
3371 address records were *not* found in the additional section, the port
3372 values from the SRV records were lost.
3374 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3375 using the correct key (the original address) when searching the retry
3376 rules in order to find which one to use for generating the retry hint.
3378 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3379 adding the default one. Similarly, if it contains a Reply-To: header, the
3380 errors_reply_to option, if set, is not used.
3382 PH/19 When calculating a retry time, Exim used to measure the "time since
3383 failure" by looking at the "first failed" field in the retry record. Now
3384 it does not use this if it is later than than the arrival time of the
3385 message. Instead it uses the arrival time. This makes for better
3386 behaviour in cases where some deliveries succeed, thus re-setting the
3387 "first failed" field. An example is a quota failure for a huge message
3388 when small messages continue to be delivered. Without this change, the
3389 "time since failure" will always be short, possible causing more frequent
3390 delivery attempts for the huge message than are intended.
3391 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3393 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3394 $1, $2, $3) because the numerical variables can be reset during some
3395 expansion items (e.g. "match"), thereby losing the authentication data.
3397 PH/21 Make -bV show the size of off_t variables so that the test suite can
3398 decide whether to run tests for quotas > 2G.
3400 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3401 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3402 filecount value is greater than 2G or if a quota value is greater than 2G
3403 on a system where the size of off_t is not greater than 4, a panic error
3406 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3407 never match. The debug and -bh output now contains an explicit error
3408 message indicating a malformed IPv4 address or mask.
3410 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3411 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3412 PH/23 above applies.
3414 PH/25 Do not write to syslog when running in the test harness. The only
3415 occasion when this arises is a failure to open the main or panic logs
3416 (for which there is an explicit test).
3418 PH/26 Added the /no_tell option to "control=freeze".
3420 PH/27 If a host name lookup failed very early in a connection, for example, if
3421 the IP address matched host_lookup and the reverse lookup yielded a name
3422 that did not have a forward lookup, an error message of the form "no IP
3423 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3424 could be logged. Now it outputs the IP address instead of "NULL".
3426 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3427 allows the sender and the authenticated sender to be set when
3428 submitting a message from within Exim. Since child_open_exim() is
3429 documented for local_scan(), the new function should be too.
3431 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3432 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3433 results in an empty string is now treated as unset.
3435 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3437 PH/31 Added sender_verify_fail logging option.
3439 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3440 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3441 tidied the source and removed it altogether.
3443 PH/33 When a queue run was abandoned because the load average was too high, a
3444 log line was always written; now it is written only if the queue_run log
3445 selector is set. In addition, the log line for abandonment now contains
3446 information about the queue run such as the pid. This is always present
3447 in "start" and "stop" lines but was omitted from the "abandon" line.
3449 PH/34 Omit spaces between a header name and the colon in the error message that
3450 is given when verify = headers_syntax fails (if there are lots of them,
3451 the message gets confusing).
3453 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3454 names, as there are now some PTR records that contain slashes. This check
3455 is only to protect against broken name servers that fall over on strange
3456 characters, so the fact that it applies to all lookups doesn't matter.
3458 PH/36 Now that the new test suite is complete, we can remove some of the
3459 special code in Exim that was needed for the old test suite. For example,
3460 sorting DNS records because real resolvers return them in an arbitrary
3461 order. The new test suite's fake resolver always returns records in the
3464 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3465 bounces) except when queue_only is set, to avoid logging races between
3466 the different processes.
3468 PH/38 Panic-die if .include specifies a non-absolute path.
3470 PH/39 A tweak to the "H" retry rule from its user.
3472 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3473 a label. They prevented compilation on older perls.
3475 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3476 a warning to be raised on newish perls.
3478 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3479 on queue. Changes to match documented behaviour of showing count of
3480 messages matching specified criteria.
3482 PH/40 Changed the default ident timeout from 30s to 5s.
3484 PH/41 Added support for the use of login_cap features, on those BSD systems
3485 that have them, for controlling the resources used by pipe deliveries.
3487 PH/42 The content-scanning code uses fopen() to create files in which to put
3488 message data. Previously it was not paying any attention to the mode of
3489 the files. Exim runs with umask(0) because the rest of the code creates
3490 files with open(), and sets the required mode explicitly. Thus, these
3491 files were ending up world-writeable. This was not a big issue, because,
3492 being within the spool directory, they were not world-accessible. I have
3493 created a function called modefopen, which takes an additional mode
3494 argument. It sets umask(777), creates the file, chmods it to the required
3495 mode, then resets the umask. All the relevant calls to fopen() in the
3496 content scanning code have been changed to use this function.
3498 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3499 to 24 hours. This avoids potential overflow problems when processing G
3500 and H retry rules. I suspect nobody ever tinkers with this value.
3502 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3504 PH/45 When the plaintext authenticator is running as a client, the server's
3505 challenges are checked to ensure they are valid base64 strings. By
3506 default, the authentication attempt is cancelled if an invalid string is
3507 received. Setting client_ignore_invalid_base64 true ignores these errors.
3508 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3509 they are received. Thus, the responses can be made to depend on the
3510 challenges. If an invalid string is ignored, an empty string is placed in
3513 PH/46 Messages that are created by the autoreply transport now contains a
3514 References: header, in accordance with RFCs 2822 and 3834.
3516 PH/47 Added authenticated_sender_force to the smtp transport.
3518 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3520 PH/49 Installed latest patch from the Sieve maintainer.
3522 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3523 was also set, the appendfile transport was unnecessarily scanning a
3524 directory of message files (e.g. for maildir delivery) to find the count
3525 of files (along with the size), even though it did not need this
3526 information. It now does the scan only if it needs to find either the
3527 size of the count of files.
3529 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3531 PH/52 Two bugs concerned with error handling when the smtp transport is
3534 (i) Exim was not creating retry information for temporary errors given
3535 for individual recipients after the DATA command when the smtp transport
3536 was used in LMTP mode. This meant that they could be retried too
3537 frequently, and not timed out correctly.
3539 (ii) Exim was setting the flag that allows error details to be returned
3540 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3541 recipients that were returned after the DATA command.
3543 PH/53 This is related to PH/52, but is more general: for any failing address,
3544 when detailed error information was permitted to be returned to the
3545 sender, but the error was temporary, then after the final timeout, only
3546 "retry timeout exceeded" was returned. Now it returns the full error as
3547 well as "retry timeout exceeded".
3549 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3550 do this, and (what is worse) MTAs that accept it.
3552 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3553 will now be deprecated.
3555 PH/56 New os.c-cygwin from the Cygwin maintainer.
3557 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3558 formats (previously only available in exim formats via -bpr, -bpru,
3559 and -bpra. Now also available in native and exiqgrep formats)
3561 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3562 with very large, slow to parse queues
3564 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3566 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3568 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3569 responses to authentication challenges, though it was showing the
3570 challenges; (ii) I've removed the CR characters from the debug output for
3573 PH/58 Allow for the insertion of a newline as well as a space when a string
3574 is turned into more than one encoded-word during RFC 2047 encoding. The
3575 Sieve code now uses this.
3577 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3578 data_4xx, lost_connection, tls_required.
3580 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3581 message was being sent as an SMTP response.
3583 PH/61 Add -l and -k options to exicyclog.
3585 PH/62 When verifying, if an address was redirected to one new address, so that
3586 verification continued, and the new address failed or deferred after
3587 having set something in $address_data, the value of $address_data was not
3588 passed back to the ACL. This was different to the case when no
3589 redirection occurred. The value is now passed back in both cases.
3591 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3592 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3593 use login_cap.h, so on its own it isn't the distinguishing feature. The
3594 new name refers directly to the setclassresources() function.
3596 PH/65 Added configuration files for NetBSD3.
3598 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3600 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3601 is preferred over IPv4.
3603 PH/68 The bounce_return_message and bounce_return_body options were not being
3604 honoured for bounces generated during the reception of non-SMTP messages.
3605 In particular, this applied to messages rejected by the ACL. This bug has
3606 been fixed. However, if bounce_return_message is true and bounce_return_
3607 body is false, the headers that are returned for a non-SMTP message
3608 include only those that have been read before the error was detected.
3609 (In the case of an ACL rejection, they have all been read.)
3611 PH/69 The HTML version of the specification is now built in a directory called
3612 spec_html instead of spec.html, because the latter looks like a path with
3613 a MIME-type, and this confuses some software.
3615 PH/70 Catch two compiler warnings in sieve.c.
3617 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3618 function verify_get_ident() calls ip_connect() to connect a socket, but
3619 if the "connect()" function timed out, ip_connect() used to close the
3620 socket. However, verify_get_ident() also closes the socket later, and in
3621 between Exim writes to the log, which may get opened at this point. When
3622 the socket was closed in ip_connect(), the log could get the same file
3623 descriptor number as the socket. This naturally causes chaos. The fix is
3624 not to close the socket in ip_connect(); the socket should be closed by
3625 the function that creates it. There was only one place in the code where
3626 this was missing, in the iplookup router, which I don't think anybody now
3627 uses, but I've fixed it anyway.
3629 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3630 well as to direct DNS lookups. Otherwise the handling of names in host
3631 lists is inconsistent and therefore confusing.
3637 PH/01 Two changes to the default runtime configuration:
3639 (1) Move the checks for relay_from_hosts and authenticated clients from
3640 after to before the (commented out) DNS black list checks.
3642 (2) Add control=submission to the relay_from_hosts and authenticated
3643 clients checks, on the grounds that messages accepted by these
3644 statements are most likely to be submissions.
3646 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3648 (1) Generate an error if the third argument for the ${prvs expansion is
3651 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3654 (3) Reset the variables that are obtained from the first argument of
3655 ${prvscheck and used in the second argument before leaving the code,
3656 because their memory is reclaimed, so using them afterwards may do
3659 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3660 one (it's much easier than Tom thought :-).
3662 (5) Because of (4), we can now allow for the use of $prvscheck_result
3663 inside the third argument.
3665 PH/03 For some reason, the default setting of PATH when running a command from
3666 a pipe transport was just "/usr/bin". I have changed it to
3669 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3670 anything to be listed in the output from -bV.
3672 PH/05 When a filter generated an autoreply, the entire To: header line was
3673 quoted in the delivery log line, like this:
3675 => >A.N.Other <ano@some.domain> <original@ddress> ...
3677 This has been changed so that it extracts the operative address. There
3678 may be more than one such address. If so, they are comma-separated, like
3681 => >ano@some.domain,ona@other.domain <original@ddress> ...
3683 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3684 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3685 not being looked up in the rDNS to get a host name, Exim was showing the
3686 IP address twice in Received: lines, even though the IP addresses were
3687 identical. For example:
3689 Received: from [1.2.3.4] (helo=[1.2.3.4])
3691 However, if the real host name was known, it was omitting the HELO data
3692 if it matched the actual IP address. This has been tidied up so that it
3693 doesn't show the same IP address twice.
3695 PH/07 When both +timestamp and +memory debugging was on, the value given by
3696 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3697 called by the debug printing, thereby overwriting the timestamp buffer.
3698 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3700 PH/08 When the original message was included in an autoreply transport, it
3701 always said "this is a copy of the message, including all the headers",
3702 even if body_only or headers_only was set. It now gives an appropriate
3705 PH/09 Applied a patch from the Sieve maintainer which:
3707 o fixes some comments
3708 o adds the (disabled) notify extension core
3709 o adds some debug output for the result of if/elsif tests
3710 o points to the current vacation draft in the documentation
3711 and documents the missing references header update
3715 o fixes a bug in processing the envelope test (when testing
3716 multiple envelope elements, the last element determined the
3719 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3720 Electronic Mail") by including:
3722 Auto-submitted: auto-generated
3724 in the messages that it generates (bounce messages and others, such as
3725 warnings). In the case of bounce messages for non-SMTP messages, there was
3726 also a typo: it was using "Auto_submitted" (underscore instead of
3727 hyphen). Since every message generated by Exim is necessarily in response
3728 to another message, thes have all been changed to:
3730 Auto-Submitted: auto-replied
3732 in accordance with these statements in the RFC:
3734 The auto-replied keyword:
3736 - SHOULD be used on messages sent in direct response to another
3737 message by an automatic process,
3739 - MUST NOT be used on manually-generated messages,
3741 - MAY be used on Delivery Status Notifications (DSNs) and Message
3742 Disposition Notifications (MDNs),
3744 - MUST NOT be used on messages generated by automatic or periodic
3745 processes, except for messages which are automatic responses to
3748 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3749 to the default Received: header definition.
3751 PH/12 Added log selector acl_warn_skipped (default on).
3753 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3754 variables because (a) they are in the wrong storage pool and (b) even if
3755 they were copied, it wouldn't work properly because of the caching.
3757 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3758 checking when decoding. Apparently there are clients that generate
3759 overlong encoded strings. Why am I not surprised?
3761 PH/15 If the first argument of "${if match_address" was not empty, but did not
3762 contain an "@" character, Exim crashed. Now it writes a panic log message
3763 and treats the condition as false.
3765 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3767 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3768 extension "envelope-auth". The code is finished and in agreement with
3769 other implementations, but there is no documentation so far and in fact,
3770 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3771 not changing the active code.
3773 Print executed "if" and "elsif" statements when debugging is used. This
3774 helps a great deal to understand what a filter does.
3776 Document more things not specified clearly in RFC3028. I had all this
3777 sorted out, when out of a sudden new issues came to my mind. Oops."
3779 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3782 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3783 canonical form (as documented). However, after a host name lookup from
3784 the IP address, check_host() was doing a simple string comparison with
3785 addresses acquired from the DNS when checking that the found name did
3786 have the original IP as one of its addresses. Since any found IPv6
3787 addresses are likely to be in abbreviated form, the comparison could
3788 fail. Luckily, there already exists a function for doing the comparison
3789 by converting both addresses to binary, so now that is used instead of
3790 the text comparison.
3792 PH/20 There was another similar case to PH/19, when a complete host name was
3793 given in a host list; looking up its IP address could give an abbreviated
3794 form, whereas the current host's name might or might not be abbreviated.
3795 The same fix has been applied.
3801 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3802 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3805 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3806 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3808 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3810 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3811 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3812 Also, for "Auto-Submitted", treat anything other than "no" as
3813 non-personal, in accordance with RFC 3834. (Previously it treated
3814 anything starting "auto-" as non-personal.)
3816 TF/01 The control=submission/name=... option had a problem with syntax
3817 errors if the name included a slash character. The /name= option
3818 now slurps the rest of the string, so it can include any characters
3819 but it must come last in the list of options (after /sender_retain
3822 PH/05 Some modifications to the interface to the fake nameserver for the new
3830 TK/01 Added the "success_on_redirect" address verification option. See
3831 NewStuff for rationale and an example.
3833 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3835 PH/02 Patch to exigrep to allow it to work on syslog lines.
3837 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3838 fread() to skip over the body file's header line, because in Cygwin the
3839 header line is locked and is inaccessible.
3841 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3842 co-exist for some time) to make it clear that it is the Exim ID that is
3843 referenced, not the Message-ID: header line.
3845 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3846 string_format() function, because snprintf() does not exist on all
3849 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3850 ${stat: expansion item.
3852 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3853 protocol synchronization error", to keep the pedants happy.
3855 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3856 well as for IRIX systems, when gcc is being used. See the host.c source
3859 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3861 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3864 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3865 transport and to the smtp transport in LMTP mode.
3867 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3869 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3870 run a filter in a subprocess. This could lead to confusion in subsequent
3871 lookups in the parent process. There should also be a search_tidyup() at
3872 the end of the subprocess.
3874 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3875 only if the host matched helo_try_verify_hosts, which caused the
3876 verification to occur when the EHLO/HELO command was issued. The ACL just
3877 tested the remembered result. Now, if a previous verification attempt has
3878 not happened, "verify = helo" does it there and then.
3880 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3882 TK/03 Fix log output including CR from clamd.
3884 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3885 exist provoked a memory error which could cause a segfault.
3887 PH/15 Installed PCRE 6.2
3889 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3891 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3892 of the problem. Specifically, suggested +O2 rather than +O1 for the
3895 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3897 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3898 record, the port was not correctly logged when the outgoing_port log
3899 selector was set (it logged the transort's default port).
3901 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3902 fallback_hosts, and "hosts" in the smtp transport.
3904 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3905 host errors such as "Connection refused".
3907 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3908 authentication with radiusclient 0.4.9:
3910 - Error returned from rc_read_config was caught wrongly
3911 - Username/password not passed on to radius server due to wrong length.
3913 The presumption is that some radiusclient API changes for 4.51/PH/17
3914 were not taken care of correctly. The code is still untested by me (my
3915 Linux distribution still has 0.3.2 of radiusclient), but it was
3916 contributed by a Radius user.
3918 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3919 expanding the "port" option of the smtp transport.
3921 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3922 while reading a MIME header. Thanks to Tom Hughes for a patch.
3924 PH/24 Include config.h inside local_scan.h so that configuration settings are
3927 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3928 that in an ACL for RCPT (for example), you can examine exactly what was
3931 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3932 commands, but it was not correctly comparing the address with the actual
3933 client host address. Thus, it would show the EHLO address in Received:
3934 header lines when this was not necessary.
3936 PH/27 Added the % operator to ${eval:}.
3938 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3939 it should be ignoring failures (because with -C, for example, it has lost
3940 privilege). It wasn't ignoring creation failures other than "already
3943 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3946 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3947 because an input line was too long, either on its own, or by virtue of
3948 too many continuations, the temporary file was not being removed, and the
3949 return code was incorrect.
3951 PH/31 Missing "BOOL" in function definition in filtertest.c.
3953 PH/32 Applied Sieve patches from the maintainer.
3955 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3957 PH/33 Added "verify = not_blind".
3959 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3960 Local/Makefile (with some defaults set). These are used in built scripts
3961 such as exicyclog, but they have never been used in the exim_install
3962 script (though there are many overriding facilities there). I have
3963 arranged that the exim_install script now takes note of these two
3966 PH/35 Installed configuration files for Dragonfly.
3968 PH/36 When a locally submitted message by a trusted user did not contain a
3969 From: header, and the sender address was obtained from -f or from an SMTP
3970 MAIL command, and the trusted user did not use -F to supply a sender
3971 name, $originator_name was incorrectly used when constructing a From:
3972 header. Furthermore, $originator_name was used for submission mode
3973 messages from external hosts without From: headers in a similar way,
3974 which is clearly wrong.
3976 PH/37 Added control=suppress_local_fixups.
3978 PH/38 When log_selector = +received_sender was set, and the addition of the
3979 sender made the log line's construction buffer exactly full, or one byte
3980 less than full, an overflow happened when the terminating "\n" was
3983 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3984 when the result of a list match is failure because a DNS lookup failed.
3986 PH/40 RM_COMMAND is now used in the building process.
3988 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3989 the "build-* directories that it finds.
3991 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3992 address in a domain literal was a prefix of an interface address.
3994 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3995 when verifying a sender address, unless rewrite_headers is false.
3997 PH/44 Wrote a long comment about why errors_to addresses are verified as
3998 recipients, not senders.
4000 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4001 the ratelimit ACL was added.
4003 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4005 PH/46 Added extra information about PostgreSQL errors to the error string.
4007 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4008 avoiding the need to install special zones in a real server. This is
4009 backwards compatible; if it can't find the fake resolver, it drops back.
4010 Thus, both old and new test suites can be run.
4012 TF/02 Added util/ratelimit.pl
4014 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4015 clock is set back in time.
4017 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4018 Candler <B.Candler@pobox.com>.
4020 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4021 OK for addresses that are the result of redirections.
4023 PH/48 A number of further additions for the benefit of the new test suite,
4024 including a fake gethostbyname() that interfaces to the fake DNS resolver
4027 TF/06 The fix for widen_domains has also been applied to qualify_single and
4028 search_parents which are the other dnslookup options that can cause
4031 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4034 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4036 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4037 service" if it encountered an MX record with an empty target hostname.
4038 The message is now "an MX or SRV record indicated no SMTP service".
4040 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4041 if the DNS of the sending site is misconfigured. This is quite a
4042 common situation. This change restores the behaviour of treating a
4043 helo verification defer as a failure.
4045 PH/51 If self=fail was set on a router, the bounce message did not include the
4046 actual error message.
4052 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4054 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4055 command itself ended in error, the underlying message about the transport
4056 filter timeout was being overwritten with the pipe command error. Now the
4057 underlying error message should be appended to the second error message.
4059 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4061 PH/02 Reduce the amount of output that "make" produces by default. Full output
4062 can still be requested.
4064 PH/03 The warning log line about a condition test deferring for a "warn" verb
4065 was being output only once per connection, rather than after each
4066 occurrence (because it was using the same function as for successful
4067 "warn" verbs). This seems wrong, so I have changed it.
4069 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4070 it should not have, which might have caused a crash in the right
4071 circumstances, but probably never did.
4073 PH/04 Installed a modified version of Tony Finch's patch to make submission
4074 mode fix the return path as well as the Sender: header line, and to
4075 add a /name= option so that you can make the user's friendly name appear
4078 TF/03 Added the control = fakedefer ACL modifier.
4080 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4081 Mark Lowes for thorough testing.
4083 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4085 TK/03 Merged latest SRS patch from Miles Wilton.
4087 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4088 with the definition in sysexits.h (which is #included earlier).
4089 Fortunately, Exim does not actually use EX_OK. The code used to try to
4090 preserve the sysexits.h value, by assuming that macro definitions were
4091 scanned for macro replacements. I have been disabused of this notion,
4092 so now the code just undefines EX_OK before #including unistd.h.
4094 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4095 in the smtp transport. When a block could not be written in a single
4096 write() function, the timeout was being re-applied to each part-write.
4097 This seems wrong - if the receiver was accepting one byte at a time it
4098 would take for ever. The timeout is now adjusted when this happens. It
4099 doesn't have to be particularly precise.
4101 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4102 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4104 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4105 without a domain if the check to <postmaster@domain> fails.
4107 SC/01 Eximstats: added -xls and the ability to specify output files
4108 (patch written by Frank Heydlauf).
4110 SC/02 Eximstats: use FileHandles for outputting results.
4112 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4114 SC/04 Eximstats: fixed display of large numbers with -nvr option
4116 SC/05 Eximstats: fixed merging of reports with empty tables.
4118 SC/06 Eximstats: added the -include_original_destination flag
4120 SC/07 Eximstats: removed tabs and trailing whitespace.
4122 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4124 TK/06 MBOX spool code: Add real "From " MBOX separator line
4125 so the .eml file is really in mbox format (even though
4126 most programs do not really care). Patch from Alex Miller.
4128 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4129 The latter is generated from $received_to and is only set if the
4130 message has one envelope recipient. SA can use these headers,
4131 obviously out-of-the-box. Patch from Alex Miller.
4133 PH/08 The ${def test on a variable was returning false if the variable's
4134 value was "0", contrary to what the specification has always said!
4135 The result should be true unless the variable is empty.
4137 PH/09 The syntax error of a character other than { following "${if
4138 def:variable_name" (after optional whitespace) was not being diagnosed.
4139 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4140 accidental colon was present, for example, could give incorrect results.
4142 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4143 result is used (not including appendfile, where other changes are about
4146 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4147 This involved changing a lot of size variables from int to off_t. It
4148 should work with maildirs and everything.
4150 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4151 spamd dying while we are connected to it.
4153 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4156 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4157 function for BDB 4.3.
4159 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4161 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4162 the sources for PCRE have been re-arranged and more files are now
4165 PH/15 The code I had for printing potentially long long variables in PH/11
4166 above was not the best (it lost precision). The length of off_t variables
4167 is now inspected at build time, and an appropriate printing format (%ld
4168 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4169 to be "long long int" or "long int". This is needed for the internal
4170 formatting function string_vformat().
4172 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4173 the configuration file to be ":syslog", then the script "guesses" where
4174 the logs files are, rather than using the compiled in default. In our
4175 case the guess is not the same as the compiled default, so the script
4176 suddenly stopped working when I started to use syslog. The patch checks
4177 to see if log_file_path is "". If so, it attempts to read it from exim
4178 with no configuration file to get the compiled in version, before it
4179 falls back to the previous guessing code."
4181 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4182 implementing BATV in an Exim configuration. See NewStuff for the gory
4185 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4186 Makefile that are specific to HP-UX.
4188 PH/18 If the "use_postmaster" option was set for a recipient callout together
4189 with the "random" option, the postmaster address was used as the MAIL
4190 FROM address for the random test, but not for the subsequent recipient
4191 test. It is now used for both.
4193 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4194 patch removes a few documentation additions to RFC 3028, because the
4195 latest draft now contains them. It adds the new en;ascii-case comparator
4196 and a new error check for 8bit text in MIME parts. Comparator and
4197 require names are now matched exactly. I enabled the subaddress
4198 extension, but it is not well tested yet (read: it works for me)."
4200 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4201 rework some of the code of TK/09 above to avoid the hardwired use of
4202 "%lld" and "long long". Replaced the call to snprintf() with a call to
4205 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4206 records point to non-existent hosts", "retry timeout exceeded", and
4207 "retry time not reached for any host after a long failure period".
4209 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4210 experimental DomainKeys support:
4212 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4213 (2) On an error such as an illegally used "control", the wrong name for
4214 the control was given.
4216 These problems did NOT occur unless DomainKeys support was compiled.
4218 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4220 PH/24 Added ${if match_ip condition.
4222 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4223 databases so that it will be absolutely obvious if a crash occurs in the
4224 DB library. This is a regular occurrence (often caused by mis-matched
4227 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4228 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4229 picked up on a user's system that detects such things. There doesn't seem
4230 to be a gcc warning option for this - only an attribute that has to be
4231 put on the function's prototype. It seems that in Fedora Core 4 they have
4232 set this on a number of new functions. No doubt there will be more in due
4235 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4236 specify a transport. However, if an address that was verified by such a
4237 router was the subject of a callout, Exim crashed because it tried to
4238 read the rcpt_include_affixes from the non-existent transport. Now it
4239 just assumes that the setting of that option is false. This bug was
4240 introduced by 4.51/PH/31.
4242 PH/28 Changed -d+all to exclude +memory, because that information is very
4243 rarely of interest, but it makes the output a lot bigger. People tend to
4244 do -d+all out of habit.
4246 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4247 code in os-type was giving problems when libc.so lives in lib64, like on
4250 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4251 aren't the modern standard, and it seems that some systems' include files
4252 don't always have them. Exim was already checking for some of the newer
4253 ones like T_AAAA, and defining it itself. I've added checks for all the
4254 record types that Exim uses.
4256 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4257 not automatically generating a new one, as it is supposed to. This
4258 prevented TLS from working. If the file did exist, but contained invalid
4259 data, a new version was generated, as expected. It was only the case of a
4260 non-existent file that was broken.
4262 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4263 with a change in libdomainkeys > 0.64.
4265 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4266 from DNS. If the selector record carries the flag, it now has
4267 precedence over the domain-wide flag.
4269 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4271 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4272 the use of a port name that isn't defined in /etc/services) occurred, the
4273 message was deferred as in a normal delivery, and thus remained on the
4274 spool, instead of being failed because of the mua_wrapper setting. This
4275 is now fixed, and I tidied up some of the mua_wrapper messages at the
4278 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4279 the messages in a hash of arrays rather than using individual hashes.
4280 This is a bit cleaner and results in dramatic memory savings, albeit
4281 at a slight CPU cost.
4283 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4284 as requested by Marc Sherman.
4286 SC/10 Eximstats: added histograms for user specified patterns as requested
4289 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4291 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4292 fopen() in the content-scanning modules that did not already have it.
4298 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4299 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4301 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4303 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4305 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4306 to test Sieve filters that use "vacation".
4308 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4309 that changes the way the GnuTLS parameters are stored in the cache file.
4310 The new format can be generated externally. For backward compatibility,
4311 if the data in the cache doesn't make sense, Exim assumes it has read an
4312 old-format file, and it generates new data and writes a new file. This
4313 means that you can't go back to an older release without removing the
4316 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4317 work if there are any delivery delays because "one_time" forces the
4318 parent to be marked "delivered", so its unseen clone is never tried
4319 again. For this reason, Exim now forbids the simultaneous setting of
4322 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4323 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4324 entry if you want to know the details. The fix, however, neglected to
4325 consider the case when local delivery batching is involved. The test for
4326 "previously delivered" was not happening when checking to see if an
4327 address could be batched with a previous (undelivered) one; under
4328 certain circumstances this could lead to multiple deliveries to the same
4331 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4332 in its include files, and this causes problems building Exim.
4334 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4335 header_syntax) but Exim was just ignoring anything given after a slash.
4336 In particular, this caused confusion with an attempt to use "verify =
4337 reverse_host_lookup/defer_ok". An error is now given when options are
4338 supplied for verify items that do not have them. (Maybe reverse_host_
4339 lookup should have a defer_ok option, but that's a different point.)
4341 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4342 defined by RFC 821) to 2048, because there were problems with some AUTH
4343 commands, and RFC 1869 says the size should be increased for extended
4344 SMTP commands that take arguments.
4346 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4349 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4350 "unknown" error; now it says that the functionality isn't in the binary.
4352 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4353 an address' error message when a string expansion fails (syntax or
4354 whatever). Otherwise the password may appear in the log. Following change
4355 PH/42 below, there is no longer a chance of it appearing in a bounce
4358 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4360 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4361 of its addresses was ever tried. (Bugzilla bug #2.)
4363 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4364 the result incorrectly in the debug output. (It correctly added a newline
4365 to what was transported.)
4367 TF/01 Added $received_time.
4369 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4370 commented out examples of how to interface to a virus scanner and to
4371 SpamAssassin. Also added commented examples of av_scanner and
4372 spamd_address settings.
4374 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4375 and controls are allowed in which ACLs. There were a couple of minor
4376 errors. Some of the entries in the conditions table (which is a table of
4377 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4378 negation of where the condition IS allowed.
4380 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4382 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4383 header file does not have a version number, so I've had to invent a new
4384 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4385 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4386 radiusclient), but it was contributed by a Radius user.
4388 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4389 files or directories, for OpenSSL.
4391 PH/19 When an Exim process that is running as root has to create an Exim log
4392 file, it does so in a subprocess that runs as exim:exim so as to get the
4393 ownership right at creation (otherwise, other Exim processes might see
4394 the file with the wrong ownership). There was no test for failure of this
4395 fork() call, which would lead to the process getting stuck as it waited
4396 for a non-existent subprocess. Forks do occasionally fail when resources
4397 run out. I reviewed all the other calls to fork(); they all seem to check
4400 PH/20 When checking for unexpected SMTP input at connect time (before writing
4401 the banner), Exim was not dealing correctly with a non-positive return
4402 from the read() function. If the client had disconnected by this time,
4403 the result was a log entry for a synchronization error with an empty
4404 string after "input=" when read() returned zero. If read() returned -1
4405 (an event I could not check), uninitialized data bytes were printed.
4406 There were reports of junk text (parts of files, etc) appearing after
4409 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4411 PH/22 Added support for macro redefinition, and (re)definition in between
4412 driver and ACL definitions.
4414 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4415 forgetting to use the resulting value; it was using the unexpanded value.
4417 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4418 hadn't been configured. The fix is from Juergen Kreileder, who
4419 understands it better than I do:
4421 "Here's what I see happening with three configured cyrus_sasl
4422 authenticators configured (plain, login, cram-md5):
4424 On startup auth_cyrus_sasl_init() gets called for each of these.
4425 This means three calls to sasl_listmech() without a specified mech_list.
4426 => SASL tests which mechs of all available mechs actually work
4427 => three warnings about OTP not working
4428 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4430 With the patch, sasl_listmech() also gets called three times. But now
4431 SASL's mech_list option is set to the server_mech specified in the the
4432 authenticator. Or in other words, the answer from sasl_listmech()
4433 gets limited to just the mech you're testing for (which is different
4435 => the return list contains just 'plain' or 'login', 'cram-md5' or
4436 nothing depending on the value of ob->server_mech.
4438 I've just tested the patch: Authentication still works fine,
4439 unavailable mechs specified in the exim configuration are still
4440 caught, and the auth.log warnings about OTP are gone."
4442 PH/25 When debugging is enabled, the contents of the command line are added
4443 to the debugging output, even when log_selector=+arguments is not
4446 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4447 answer is "GNU", and only if the return is "GNU/something" is the answer
4450 PH/27 $acl_verify_message is now set immediately after the failure of a
4451 verification in an ACL, and so is available in subsequent modifiers. In
4452 particular, the message can be preserved by coding like this:
4454 warn !verify = sender
4455 set acl_m0 = $acl_verify_message
4457 Previously, $acl_verify_message was set only while expanding "message"
4458 and "log_message" when a very denied access.
4460 PH/28 Modified OS/os.c-Linux with
4462 -#ifndef OS_LOAD_AVERAGE
4463 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4465 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4466 nomenclature these days.)
4468 PH/29 Installed patch from the Sieve maintainer that adds the options
4469 sieve_useraddress and sieve_subaddress to the redirect router.
4471 PH/30 In these circumstances:
4472 . Two addresses routed to the same list of hosts;
4473 . First host does not offer TLS;
4474 . First host accepts first address;
4475 . First host gives temporary error to second address;
4476 . Second host offers TLS and a TLS session is established;
4477 . Second host accepts second address.
4478 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4479 and peerdn, if requested) that were in fact used only for the second
4482 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4483 attention to any local part prefix or suffix that was matched by the
4484 router that accepted the address. It now behaves in the same way as it
4485 does for delivery: the affixes are removed from the local part unless
4486 rcpt_include_affixes is set on the transport.
4488 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4489 timeout during the DATA phase of an incoming message.
4491 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4492 applied a patch sanctioned by the Sieve maintainer.
4494 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4495 the uid or gid is negative. A case of a negative gid caused this to be
4496 noticed. The fix allows for either to be negative.
4498 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4499 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4502 PH/36 The change PH/12 above was broken. Fixed it.
4504 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4505 the grounds that routing the same address twice would always produce the
4506 same answer. This might have been true once, but it is certainly no
4507 longer true now. Routing a child address may depend on the previous
4508 routing that produced that child. Some complicated redirection strategies
4509 went wrong when messages had multiple recipients, and made Exim's
4510 behaviour dependent on the order in which the addresses were given.
4512 I have moved the duplicate checking until after the routing is complete.
4513 Exim scans the addresses that are assigned to local and remote
4514 transports, and removes any duplicates. This means that more work will be
4515 done, as duplicates will always all be routed, but duplicates are
4516 presumably rare, so I don't expect this is of any significance.
4518 For deliveries to pipes, files, and autoreplies, the duplicate checking
4519 still happens during the routing process, since they are not going to be
4522 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4523 It corrects a timeout issue with spamd. This is Ian's comment: "The
4524 background is that sometimes spamd either never reads data from a
4525 connection it has accepted, or it never writes response data. The exiscan
4526 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4527 blindly assumes that writes won't block so it may never time out."
4529 PH/39 Allow G after quota size as well as K and M.
4531 PH/40 The value set for $authenticated_id in an authenticator may not contain
4532 binary zeroes or newlines because the value is written to log lines and
4533 to spool files. There was no check on this. Now the value is run through
4534 the string_printing() function so that such characters are converted to
4535 printable escape sequences.
4537 PH/41 $message_linecount is a new variable that contains the total number of
4538 lines in the message. Compare $body_linecount, which is the count for the
4541 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4542 bounce and delay warning messages, except in certain special cases, which
4545 (a) An SMTP error message from a remote host;
4546 (b) A message specified in a :fail: redirection;
4547 (c) A message specified in a "fail" command in a system filter;
4548 (d) A message specified in a FAIL return from the queryprogram router;
4549 (e) A message specified by the cannot_route_message router option.
4551 In these cases only, Exim does include the error details in bounce and
4552 warning messages. There are also a few cases where bland messages such
4553 as "unrouteable address" or "local delivery error" are given.
4555 PH/43 $value is now also set for the "else" part of a ${run expansion.
4557 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4558 being worked on, but at least Exim now implements the latest version to
4561 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4562 process to complete was treated as a delivery failure, a timeout while
4563 writing the message to the pipe was logged, but erroneously treated as a
4564 successful delivery. Such timeouts include transport filter timeouts. For
4565 consistency with the overall process timeout, these timeouts are now
4566 treated as errors, giving rise to delivery failures by default. However,
4567 there is now a new Boolean option for the pipe transport called
4568 timeout_defer, which, if set TRUE, converts the failures into defers for
4569 both kinds of timeout. A transport filter timeout is now identified in
4572 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4573 systems where "make" and "gmake" are different, calling "gmake" at top
4574 level broke things. I've arranged for the value of $(MAKE) to be passed
4575 from the Makefile to this script so that it can call the same version of
4579 A note about Exim versions 4.44 and 4.50
4580 ----------------------------------------
4582 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4583 changes of various kinds. As a consequence, a big documentation update was
4584 needed. This delayed the release for rather longer than seemed good, especially
4585 in the light of a couple of (minor) security issues. Therefore, the changes
4586 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4587 release. So 4.44 and 4.50 are in effect two different branches that both start
4590 I have left the 4.50 change log unchanged; it contains all the changes since
4591 4.43. The change log for 4.44 is below; many of its items are identical to
4592 those for 4.50. This seems to be the most sensible way to preserve the
4593 historical information.
4599 1. Minor wording change to the doc/README.SIEVE file.
4601 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4602 computation of the current number of files was incorrect.
4604 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4605 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4608 4. Give more explanation in the error message when the command for a transport
4609 filter fails to execute.
4611 5. There are several places where Exim runs a non-Exim command in a
4612 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4613 was being done only for the command run by the queryprogram router. It is
4614 now done for all such subprocesses. The other cases are: ${run, transport
4615 filters, and the commands run by the lmtp and pipe transports.
4617 6. Added CONFIGURE_GROUP build-time option.
4619 7. Some older OS have a limit of 256 on the maximum number of file
4620 descriptors. Exim was using setrlimit() to set 1000 as a large value
4621 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4622 systems. I've change it so that if it can't get 1000, it tries for 256.
4624 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4625 was an oversight, and furthermore, ever since the addition of extra
4626 controls (e.g. 4.43/32), the checks on when to allow different forms of
4627 "control" were broken. There should now be diagnostics for all cases when a
4628 control that does not make sense is encountered.
4630 9. Added the /retain_sender option to "control=submission".
4632 10. $recipients is now available in the predata ACL (oversight).
4634 11. Tidy the search cache before the fork to do a delivery from a message
4635 received from the command line. Otherwise the child will trigger a lookup
4636 failure and thereby defer the delivery if it tries to use (for example) a
4637 cached ldap connection that the parent has called unbind on.
4639 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4640 of $address_data from the recipient verification was clobbered by the
4641 sender verification.
4643 13. The value of address_data from a sender verification is now available in
4644 $sender_address_data in subsequent conditions in the ACL statement.
4646 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4648 15. Added a new option "connect=<time>" to callout options, to set a different
4651 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4652 was its contents. (It was OK if the option was not defined at all.)
4654 17. A "Completed" log line is now written for messages that are removed from
4655 the spool by the -Mrm option.
4657 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4658 information about exactly what failed.
4660 19. Added -dd to debug only the daemon process.
4662 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4663 handles timeouts, both on the server side and network timeouts. Renamed the
4664 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4666 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4667 It is now set to "smtps".
4669 22. $host_address is now set to the target address during the checking of
4670 ignore_target_hosts.
4672 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4673 being passed; this would have caused $sender_host_name to have been used if
4674 matching the list had actually called for a host name (not very likely,
4675 since this list is usually IP addresses). A host name is now passed as
4678 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4679 code that specifies a non-restarting handler (typically sigaction() in
4680 modern systems) in an attempt to fix a rare and obscure crash bug.
4682 25. Narrowed the window for a race in the daemon that could cause it to ignore
4683 SIGCHLD signals. This is not a major problem, because they are used only to
4684 wake it up if nothing else does.
4686 26. A malformed maildirsize file could cause Exim to calculate negative values
4687 for the mailbox size or file count. Odd effects could occur as a result.
4688 The maildirsize information is now recalculated if the size or filecount
4691 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4692 support for a long time. Removed HAVE_SYS_VFS_H.
4694 28. Installed the latest version of exipick from John Jetmore.
4696 29. In an address list, if the pattern was not a regular expression, an empty
4697 subject address (from a bounce message) matched only if the pattern was an
4698 empty string. Non-empty patterns were not even tested. This was the wrong
4699 because it is perfectly reasonable to use an empty address as part of a
4700 database query. An empty address is now tested by patterns that are
4701 lookups. However, all the other forms of pattern expect the subject to
4702 contain a local part and a domain, and therefore, for them, an empty
4703 address still always fails if the pattern is not itself empty.
4705 30. Exim went into a mad DNS loop when attempting to do a callout where the
4706 host was specified on an smtp transport, and looking it up yielded more
4707 than one IP address.
4709 31. Re-factored the code for checking spool and log partition space into a
4710 function that finds that data and another that does the check. The former
4711 is then used to implement four new variables: $spool_space, $log_space,
4712 $spool_inodes, and $log_inodes.
4714 32. The RFC2047 encoding function was originally intended for short strings
4715 such as real names; it was not keeping to the 75-character limit for
4716 encoded words that the RFC imposes. It now respects the limit, and
4717 generates multiple encoded words if necessary. To be on the safe side, I
4718 have increased the buffer size for the ${rfc2047: expansion operator from
4721 33. It is now permitted to omit both strings after an "if" condition; if the
4722 condition is true, the result is "true". As before, when the second string
4723 is omitted, a false condition yields an empty string. This makes it less
4724 cumbersome to write custom ACL and router conditions.
4726 34. Failure to deliver a bounce message always caused it to be frozen, even if
4727 there was an errors_to setting on the router. The errors_to setting is now
4730 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4731 canonical form (fully expanded) before being placed in
4732 $sender_host_address.
4734 36. The table in the code that translates DNS record types into text (T_A to
4735 "A" for instance) was missing entries for NS and CNAME. It is just possible
4736 that this could have caused confusion if both these types were looked up
4737 for the same domain, because the text type is used as part of Exim's
4738 per-process caching. But the chance of anyone hitting this buglet seems
4741 37. The dnsdb lookup has been extended in a number of ways.
4743 (1) There is a new type, "zns", which walks up the domain tree until it
4744 finds some nameserver records. It should be used with care.
4746 (2) There is a new type, "mxh", which is like "mx" except that it returns
4747 just the host names, not the priorities.
4749 (3) It is now possible to give a list of domains (or IP addresses) to be
4750 looked up. The behaviour when one of the lookups defers can be
4751 controlled by a keyword.
4753 (4) It is now possible to specify the separator character for use when
4754 multiple records are returned.
4756 38. The dnslists ACL condition has been extended: it is now possible to supply
4757 a list of IP addresses and/or domains to be looked up in a particular DNS
4760 39. Added log_selector=+queue_time_overall.
4762 40. When running the queue in the test harness, wait just a tad after forking a
4763 delivery process, to get repeatability of debugging output.
4765 41. Include certificate and key file names in error message when GnuTLS fails
4766 to set them up, because the GnuTLS error message doesn't include the name
4767 of the failing file when there is a problem reading it.
4769 42. Allow both -bf and -bF in the same test run.
4771 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4773 44. The "Exiscan patch" is now merged into the mainline Exim source.
4775 45. Sometimes the final signoff response after QUIT could fail to get
4776 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4777 < 0 before doing a fflush(). This bug looks as though it goes back to the
4778 introduction of TLS in release 3.20, but "sometimes" must have been rare
4779 because the tests only now provoked it.
4781 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4782 (this can affect the format of dates).
4784 47. exim_tidydb, when checking for the continued existence of a message for
4785 which it has found a message-specific retry record, was not finding
4786 messages that were in split spool directories. Consequently, it was
4787 deleting retry records that should have stayed in existence.
4789 48. Steve fixed some bugs in eximstats.
4791 49. The SPA authentication driver was not abandoning authentication and moving
4792 on to the next authenticator when an expansion was forced to fail,
4793 contradicting the general specification for all authenticators. Instead it
4794 was generating a temporary error. It now behaves as specified.
4796 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4797 (the order specifies the preference for clients). The order is now AES256,
4798 AES128, 3DES, ARCFOUR128.
4800 51. Small patch to Sieve code - explicitly set From: when generating an
4803 52. Exim crashed if a remote delivery caused a very long error message to be
4804 recorded - for instance if somebody sent an entire SpamAssassin report back
4805 as a large number of 550 error lines. This bug was coincidentally fixed by
4806 increasing the size of one of Exim's internal buffers (big_buffer) that
4807 happened as part of the Exiscan merge. However, to be on the safe side, I
4808 have made the code more robust (and fixed the comments that describe what
4811 53. Now that there can be additional text after "Completed" in log lines (if
4812 the queue_time_overall log selector is set), a one-byte patch to exigrep
4813 was needed to allow it to recognize "Completed" as not the last thing in
4816 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4817 patch that reportedly fixes this has been added. I am not expert enough to
4818 create a test for it. This is what the patch creator wrote:
4820 "I found a little strange behaviour of ldap code when working with
4821 Windows 2003 AD Domain, where users was placed in more than one
4822 Organization Units. When I tried to give exim partial DN, the exit code
4823 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4824 But simultaneously result of request was absolutely normal ldap result,
4825 so I produce this patch..."
4827 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4828 REFERENCE, so I have modified the code to exclude the patch when that macro
4831 55. Some experimental protocols are using DNS PTR records for new purposes. The
4832 keys for these records are domain names, not reversed IP addresses. The
4833 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4834 leaves it alone. Component reversal etc. now happens only for IP addresses.
4837 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4839 57. Double the size of the debug message buffer (to 2048) so that more of very
4840 long debug lines gets shown.
4842 58. The exicyclog utility now does better if the number of log files to keep
4843 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4845 59. Two changes related to the smtp_active_hostname option:
4847 (1) $smtp_active_hostname is now available as a variable.
4848 (2) The default for smtp_banner uses $smtp_active_hostname instead
4849 of $primary_hostname.
4851 60. The host_aton() function is supposed to be passed a string that is known
4852 to be a valid IP address. However, in the case of IPv6 addresses, it was
4853 not checking this. This is a hostage to fortune. Exim now panics and dies
4854 if the condition is not met. A case was found where this could be provoked
4855 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4856 components; fortuitously, this particular loophole had already been fixed
4857 by change 4.50/55 above.
4859 If there are any other similar loopholes, the new check in host_aton()
4860 itself should stop them being exploited. The report I received stated that
4861 data on the command line could provoke the exploit when Exim was running as
4862 exim, but did not say which command line option was involved. All I could
4863 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4864 running as the user.
4867 61. There was a buffer overflow vulnerability in the SPA authentication code
4868 (which came originally from the Samba project). I have added a test to the
4869 spa_base64_to_bits() function which I hope fixes it.
4872 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4873 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4875 63. The daemon start-up calls getloadavg() while still root for those OS that
4876 need the first call to be done as root, but it missed one case: when
4877 deliver_queue_load_max is set with deliver_drop_privilege. This is
4878 necessary for the benefit of the queue runner, because there is no re-exec
4879 when deliver_drop_privilege is set.
4881 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4882 This has been fixed.
4884 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4885 was in use, was not putting the data itself into the right store pool;
4886 consequently, it could be overwritten for a subsequent message in the same
4887 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4890 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4892 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4893 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4894 the function were treating the return as a boolean value, which happened to
4895 work because 0=false and not-0=true, but is not correct code.
4897 68. The host_aton() function was not handling scoped IPv6 addresses (those
4898 with, for example, "%eth0" on the end) correctly.
4900 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4901 negated items (that is, ~something) in unsigned ints. Some compilers
4902 apparently mutter when there is no cast.
4904 70. If an address verification called from an ACL failed, and did not produce a
4905 user-specific message (i.e. there was only a "system" message), nothing was
4906 put in $acl_verify_message. In this situation, it now puts the system
4909 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4910 session; change 4.31/43 added the unwanted input to the log line - except
4911 that it did not do this in the start of session case. It now does.
4913 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4914 This is wrong and can cause the other end to generate a synchronization
4915 error if it is another Exim or anything else that does the synchronization
4916 check. A QUIT command is no longer sent after a timeout.
4918 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4919 during host lookups.
4921 74. The defer_ok option of callout verification was not working if it was used
4922 when verifying addresses in header lines, that is, for this case:
4924 verify = header_sender/callout=defer_ok
4926 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4927 those file descriptors could be used for SMTP connections. If anything
4928 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4929 could be sent to the SMTP client, causing chaos. The daemon now opens
4930 stdin, stdout, and stderr to /dev/null when it puts itself into the
4933 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4934 log by default. The user can override this with suitable Perl magic.
4936 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4937 the log message when discard triggers, was not working for the DATA ACL or
4938 for the non-SMTP ACL.
4940 78. Error message wording change in sieve.c.
4942 79. If smtp_accept_max_per_host was set, the number of connections could be
4943 restricted to fewer than expected, because the daemon was trying to set up
4944 a new connection before checking whether the processes handling previous
4945 connections had finished. The check for completed processes is now done
4946 earlier. On busy systems, this bug wouldn't be noticed because something
4947 else would have woken the daemon, and it would have reaped the completed
4950 80. If a message was submitted locally by a user whose login name contained one
4951 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4952 It caused a spool format error. I have fixed the spool reading code. A
4953 related problem was that the "from" clause in the Received: line became
4954 illegal because of the space(s). It is now covered by ${quote_local_part.
4956 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4959 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4961 83. Patch from Sieve maintainer to fix unterminated string problem in
4962 "vacation" handling.
4964 84. Some minor changes to the Linux configuration files to help with other
4965 OS variants using glibc.
4967 85. One more patch for Sieve to update vacation handling to latest spec.
4970 ----------------------------------------------------
4971 See the note above about the 4.44 and 4.50 releases.
4972 ----------------------------------------------------
4978 1. Change 4.43/35 introduced a bug that caused file counts to be
4979 incorrectly computed when quota_filecount was set in an appendfile
4982 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4983 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4986 3. Give more explanation in the error message when the command for a transport
4987 filter fails to execute.
4989 4. There are several places where Exim runs a non-Exim command in a
4990 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4991 was being done only for the command run by the queryprogram router. It is
4992 now done for all such subprocesses. The other cases are: ${run, transport
4993 filters, and the commands run by the lmtp and pipe transports.
4995 5. Some older OS have a limit of 256 on the maximum number of file
4996 descriptors. Exim was using setrlimit() to set 1000 as a large value
4997 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4998 systems. I've change it so that if it can't get 1000, it tries for 256.
5000 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5001 was an oversight, and furthermore, ever since the addition of extra
5002 controls (e.g. 4.43/32), the checks on when to allow different forms of
5003 "control" were broken. There should now be diagnostics for all cases when a
5004 control that does not make sense is encountered.
5006 7. $recipients is now available in the predata ACL (oversight).
5008 8. Tidy the search cache before the fork to do a delivery from a message
5009 received from the command line. Otherwise the child will trigger a lookup
5010 failure and thereby defer the delivery if it tries to use (for example) a
5011 cached ldap connection that the parent has called unbind on.
5013 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5014 of $address_data from the recipient verification was clobbered by the
5015 sender verification.
5017 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5018 was its contents. (It was OK if the option was not defined at all.)
5020 11. A "Completed" log line is now written for messages that are removed from
5021 the spool by the -Mrm option.
5023 12. $host_address is now set to the target address during the checking of
5024 ignore_target_hosts.
5026 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5027 being passed; this would have caused $sender_host_name to have been used if
5028 matching the list had actually called for a host name (not very likely,
5029 since this list is usually IP addresses). A host name is now passed as
5032 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5033 code that specifies a non-restarting handler (typically sigaction() in
5034 modern systems) in an attempt to fix a rare and obscure crash bug.
5036 15. Narrowed the window for a race in the daemon that could cause it to ignore
5037 SIGCHLD signals. This is not a major problem, because they are used only to
5038 wake it up if nothing else does.
5040 16. A malformed maildirsize file could cause Exim to calculate negative values
5041 for the mailbox size or file count. Odd effects could occur as a result.
5042 The maildirsize information is now recalculated if the size or filecount
5045 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5046 support for a long time. Removed HAVE_SYS_VFS_H.
5048 18. Updated exipick to current release from John Jetmore.
5050 19. Allow an empty sender to be matched against a lookup in an address list.
5051 Previously the only cases considered were a regular expression, or an
5054 20. Exim went into a mad DNS lookup loop when doing a callout where the
5055 host was specified on the transport, if the DNS lookup yielded more than
5058 21. The RFC2047 encoding function was originally intended for short strings
5059 such as real names; it was not keeping to the 75-character limit for
5060 encoded words that the RFC imposes. It now respects the limit, and
5061 generates multiple encoded words if necessary. To be on the safe side, I
5062 have increased the buffer size for the ${rfc2047: expansion operator from
5065 22. Failure to deliver a bounce message always caused it to be frozen, even if
5066 there was an errors_to setting on the router. The errors_to setting is now
5069 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5070 canonical form (fully expanded) before being placed in
5071 $sender_host_address.
5073 24. Updated eximstats to version 1.33
5075 25. Include certificate and key file names in error message when GnuTLS fails
5076 to set them up, because the GnuTLS error message doesn't include the name
5077 of the failing file when there is a problem reading it.
5079 26. Expand error message when OpenSSL has problems setting up cert/key files.
5082 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5083 (this can affect the format of dates).
5085 28. exim_tidydb, when checking for the continued existence of a message for
5086 which it has found a message-specific retry record, was not finding
5087 messages that were in split spool directories. Consequently, it was
5088 deleting retry records that should have stayed in existence.
5090 29. eximstats updated to version 1.35
5091 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5092 1.35 - bugfix such that pie charts by volume are generated correctly
5094 30. The SPA authentication driver was not abandoning authentication and moving
5095 on to the next authenticator when an expansion was forced to fail,
5096 contradicting the general specification for all authenticators. Instead it
5097 was generating a temporary error. It now behaves as specified.
5099 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5100 (the order specifies the preference for clients). The order is now AES256,
5101 AES128, 3DES, ARCFOUR128.
5103 31. Small patch to Sieve code - explicitly set From: when generating an
5106 32. Exim crashed if a remote delivery caused a very long error message to be
5107 recorded - for instance if somebody sent an entire SpamAssassin report back
5108 as a large number of 550 error lines. This bug was coincidentally fixed by
5109 increasing the size of one of Exim's internal buffers (big_buffer) that
5110 happened as part of the Exiscan merge. However, to be on the safe side, I
5111 have made the code more robust (and fixed the comments that describe what
5114 33. Some experimental protocols are using DNS PTR records for new purposes. The
5115 keys for these records are domain names, not reversed IP addresses. The
5116 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5117 leaves it alone. Component reversal etc. now happens only for IP addresses.
5120 34. The host_aton() function is supposed to be passed a string that is known
5121 to be a valid IP address. However, in the case of IPv6 addresses, it was
5122 not checking this. This is a hostage to fortune. Exim now panics and dies
5123 if the condition is not met. A case was found where this could be provoked
5124 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5125 components; fortuitously, this particular loophole had already been fixed
5126 by change 4.50/55 or 4.44/33 above.
5128 If there are any other similar loopholes, the new check in host_aton()
5129 itself should stop them being exploited. The report I received stated that
5130 data on the command line could provoke the exploit when Exim was running as
5131 exim, but did not say which command line option was involved. All I could
5132 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5133 running as the user.
5136 35. There was a buffer overflow vulnerability in the SPA authentication code
5137 (which came originally from the Samba project). I have added a test to the
5138 spa_base64_to_bits() function which I hope fixes it.
5141 36. The daemon start-up calls getloadavg() while still root for those OS that
5142 need the first call to be done as root, but it missed one case: when
5143 deliver_queue_load_max is set with deliver_drop_privilege. This is
5144 necessary for the benefit of the queue runner, because there is no re-exec
5145 when deliver_drop_privilege is set.
5147 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5148 was in use, was not putting the data itself into the right store pool;
5149 consequently, it could be overwritten for a subsequent message in the same
5150 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5153 38. Sometimes the final signoff response after QUIT could fail to get
5154 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5155 < 0 before doing a fflush(). This bug looks as though it goes back to the
5156 introduction of TLS in release 3.20, but "sometimes" must have been rare
5157 because the tests only now provoked it.
5163 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5164 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5165 Now it returns an integer. A number of calls were still expecting a T/F
5166 return. Fortuitously, in all cases, the tests worked in OK situations,
5167 which is the norm. However, things would have gone wrong on any write
5168 failures on the smtp file descriptor. This function is used when sending
5169 messages over SMTP and also when doing verify callouts.
5171 2. When Exim is called to do synchronous delivery of a locally submitted
5172 message (the -odf or -odi options), it no longer closes stderr before doing
5175 3. Implemented the mua_wrapper option.
5177 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5179 5. Implemented the functions header_remove(), header_testname(),
5180 header_add_at_position(), and receive_remove_recipient(), and exported them
5183 6. If an ACL "warn" statement specified the addition of headers, Exim already
5184 inserted X-ACL-Warn: at the start if there was no header name. However, it
5185 was not making this test for the second and subsequent header lines if
5186 there were newlines in the string. This meant that an invalid header could
5187 be inserted if Exim was badly configured.
5189 7. Allow an ACL "warn" statement to add header lines at the start or after all
5190 the Received: headers, as well as at the end.
5192 8. Added the rcpt_4xx retry error code.
5194 9. Added postmaster_mailfrom=xxx to callout verification option.
5196 10. Added mailfrom=xxxx to the callout verification option, for verify=
5199 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5200 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5202 12. Inserted some casts to stop certain compilers warning when using pointer
5203 differences as field lengths or precisions in printf-type calls (mostly
5204 affecting debugging statements).
5206 13. Added optional readline() support for -be (dynamically loaded).
5208 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5209 same clock tick as a message's arrival, so that its received time was the
5210 same as the "first fail" time on the retry record, and that message
5211 remained on the queue past the ultimate address timeout, every queue runner
5212 would try a delivery (because it was past the ultimate address timeout) but
5213 after another failure, the ultimate address timeout, which should have then
5214 bounced the address, did not kick in. This was a "< instead of <=" error;
5215 in most cases the first failure would have been in the next clock tick
5216 after the received time, and all would be well.
5218 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5219 being recognized when the domain list was tested by the match_domain
5220 condition in an expansion string.
5222 16. Added the ${str2b64: operator.
5224 17. Exim was always calling setrlimit() to set a large limit for the number of
5225 processes, without checking whether the existing limit was already
5226 adequate. (It did check for the limit on file descriptors.) Furthermore,
5227 errors from getrlimit() and setrlimit() were being ignored. Now they are
5228 logged to the main and panic logs, but Exim does carry on, to try to do its
5229 job under whatever limits there are.
5231 18. Imported PCRE 5.0.
5233 19. Trivial typo in log message " temporarily refused connection" (the leading
5236 20. If the log selector return_path_on_delivery was set and an address was
5237 redirected to /dev/null, the delivery process crashed because it assumed
5238 that a return path would always be set for a "successful" delivery. In this
5239 case, the whole delivery is bypassed as an optimization, and therefore no
5242 21. Internal re-arrangement: the function for sending a challenge and reading
5243 a response while authentication was assuming a zero-terminated challenge
5244 string. It's now changed to take a pointer and a length, to allow for
5245 binary data in such strings.
5247 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5249 23. Exim was not respecting finduser_retries when seeking the login of the
5250 uid under which it was called; it was always trying 10 times. (The default
5251 setting of finduser_retries is zero.) Also, it was sleeping after the final
5252 failure, which is pointless.
5254 24. Implemented tls_on_connect_ports.
5256 25. Implemented acl_smtp_predata.
5258 26. If the domain in control=submission is set empty, Exim assumes that the
5259 authenticated id is a complete email address when it generates From: or
5260 Sender: header lines.
5262 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5263 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5264 chown and chgrp in /bin and hostname in /usr/bin.
5266 28. Exim was keeping the "process log" file open after each use, just as it
5267 does for the main log. This opens the possibility of it remaining open for
5268 long periods when the USR1 signal hits a daemon. Occasional processlog
5269 errors were reported, that could have been caused by this. Anyway, it seems
5270 much more sensible not to leave this file open at all, so that is what now
5273 29. The long-running daemon process does not normally write to the log once it
5274 has entered its main loop, and it closes the log before doing so. This is
5275 so that log files can straightforwardly be renamed and moved. However,
5276 there are a couple of unusual error situations where the daemon does write
5277 log entries, and I had neglected to close the log afterwards.
5279 30. The text of an SMTP error response that was received during a remote
5280 delivery was being truncated at 512 bytes. This is too short for some of
5281 the long messages that one sometimes sees. I've increased the limit to
5284 31. It is now possible to make retry rules that apply only when a message has a
5285 specific sender, in particular, an empty sender.
5287 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5288 it possible to be selective about when SMTP synchronization is enforced.
5290 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5292 32. Implemented hosts_connection_nolog.
5294 33. Added an ACL for QUIT.
5296 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5299 35. Added mailbox_size and mailbox_filecount to appendfile.
5301 36. Added control = no_multiline_responses to ACLs.
5303 37. There was a bug in the logic of the code that waits for the clock to tick
5304 in the case where the clock went backwards by a substantial amount such
5305 that the microsecond fraction of "now" was more than the microsecond
5306 fraction of "then" (but the whole seconds number was less).
5308 38. Added support for the libradius Radius client library this is found on
5309 FreeBSD (previously only the radiusclient library was supported).
5315 1. When certain lookups returned multiple values in the form name=value, the
5316 quoting of the values was not always being done properly. Specifically:
5317 (a) If the value started with a double quote, but contained no whitespace,
5319 (b) If the value contained whitespace other than a space character (i.e.
5320 tabs or newlines or carriage returns) it was not quoted.
5321 This fix has been applied to the mysql and pgsql lookups by writing a
5322 separate quoting function and calling it from the lookup code. The fix
5323 should probably also be applied to nisplus, ibase and oracle lookups, but
5324 since I cannot test any of those, I have not disturbed their existing code.
5326 2. A hit in the callout cache for a specific address caused a log line with no
5327 reason for rejecting RCPT. Now it says "Previous (cached) callout
5328 verification failure".
5330 3. There was an off-by-one bug in the queryprogram router. An over-long
5331 return line was truncated at 256 instead of 255 characters, thereby
5332 overflowing its buffer with the terminating zero. As well as fixing this, I
5333 have increased the buffer size to 1024 (and made a note to document this).
5335 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5336 when Exim is waiting for an SMTP response from a remote server, Exim
5337 restarts its select() call on the socket, thereby resetting its timeout.
5338 This is not a problem when such interrupts are rare. Somebody set up a cron
5339 job to run exiwhat every 2 minutes, which is less than the normal select()
5340 timeout (5 or 10 minutes). This meant that the select() timeout never
5341 kicked in because it was always reset. I have fixed this by comparing the
5342 time when an interrupt arrives with the time at the start of the first call
5343 to select(). If more time than the timeout has elapsed, the interrupt is
5344 treated as a timeout.
5346 5. Some internal re-factoring in preparation for the addition of Sieve
5347 extensions (by MH). In particular, the "personal" test is moved to a
5348 separate function, and given an option for scanning Cc: and Bcc: (which is
5349 not set for Exim filters).
5351 6. When Exim created an email address using the login of the caller as the
5352 local part (e.g. when creating a From: or Sender: header line), it was not
5353 quoting the local part when it contained special characters such as @.
5355 7. Installed new OpenBSD configuration files.
5357 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5358 try to make them clearer.
5360 9. Callout options, other than the timeout value, were being ignored when
5361 verifying sender addresses in header lines. For example, when using
5363 verify = header_sender/callout=no_cache
5365 the cache was (incorrectly) being used.
5367 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5368 only the Cygwin environment.
5370 11. When return_path_on_delivery was set as a log selector, if different remote
5371 addresses in the same message used different return paths and parallel
5372 remote delivery occurred, the wrong values would sometimes be logged.
5373 (Whenever a remote delivery process finished, the return path value from
5374 the most recently started remote delivery process was logged.)
5376 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5377 lines when AUTH and/or TLS are in use. This is the "received protocol"
5378 field. Exim used to use "asmtp" for authenticated SMTP, without any
5379 indication (in the protocol name) for TLS use. Now it follows the RFC and
5380 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5381 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5382 names appear in log lines as well as in Received: header lines.
5384 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5385 extensions, and comparison tests, and to fix some bugs.
5387 14. Changes to the "personal" filter test:
5389 (1) The test was buggy in that it was just doing the equivalent of
5390 "contains" tests on header lines. For example, if a user's address was
5391 anne@some.where, the "personal" test would incorrectly be true for
5393 To: susanne@some.where
5395 This test is now done by extracting each address from the header in turn,
5396 and checking the entire address. Other tests that are part of "personal"
5397 are now done using regular expressions (for example, to check local parts
5398 of addresses in From: header lines).
5400 (2) The list of non-personal local parts in From: addresses has been
5401 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5402 taken from the Sieve specification recommendations.
5404 (3) If the message contains any header line starting with "List-" it is
5405 treated as non-personal.
5407 (4) The test for "circular" in the Subject: header line has been removed
5408 because it now seems ill-conceived.
5410 15. Minor typos in src/EDITME comments corrected.
5412 16. Installed latest exipick from John Jetmore.
5414 17. If headers_add on a router specified a text string that was too long for
5415 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5416 of string_sprintf() is now avoided.
5418 18. $message_body_size was not set (it was always zero) when running the DATA
5419 ACL and the local_scan() function.
5421 19. For the "mail" command in an Exim filter, no default was being set for
5422 the once_repeat time, causing a random time value to be used if "once" was
5423 specified. (If the value happened to be <= 0, no repeat happened.) The
5424 default is now 0s, meaning "never repeat". The "vacation" command was OK
5425 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5426 (I found it when inspecting the code).
5428 20. There is now an overall timeout for performing a callout verification. It
5429 defaults to 4 times the callout timeout, which applies to individual SMTP
5430 commands during the callout. The overall timeout applies when there is more
5431 than one host that can be tried. The timeout is checked before trying the
5432 next host. This prevents very long delays if there are a large number of
5433 hosts and all are timing out (e.g. when the network connections are timing
5434 out). The value of the overall timeout can be changed by specifying an
5435 additional sub-option for "callout", called "maxwait". For example:
5437 verify = sender/callout=5s,maxwait=20s
5439 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5440 to the end before writing, but this should make it even safer).
5442 22. Exim was forgetting that it had advertised PIPELINING for the second and
5443 subsequent messages on an SMTP connection. It was also not resetting its
5444 memory on STARTTLS and an internal HELO.
5446 23. When Exim logs an SMTP synchronization error within a session, it now
5447 records whether PIPELINING has been advertised or not.
5449 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5450 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5451 rather than long int.
5453 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5455 26. Added the never_mail option to autoreply.
5461 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5462 crash if the getsockname() call failed; this can happen if a connection is
5463 closed very soon after it is established. The problem was simply in the
5464 order in which certain operations were done, causing Exim to try to write
5465 to the SMTP stream before it had set up the file descriptor. The bug has
5466 been fixed by making things happen in the correct order.
5472 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5473 before the connection was closed, thus losing the rejection response.
5475 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5476 some early Solaris releases, but causes trouble in current releases where
5477 socklen_t is defined.
5479 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5482 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5485 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5486 that releasing the top of it at the end releases what was used for sub-
5487 expansions (unless the block got too big). However, discard this block if
5488 the first thing is a variable or header, so that we can use its block when
5489 it is dynamic (useful for very large $message_headers, for example).
5491 6. Lookups now cache *every* query, not just the most recent. A new, separate
5492 store pool is used for this. It can be recovered when all lookup caches are
5493 flushed. Lookups now release memory at the end of their result strings.
5494 This has involved some general refactoring of the lookup sources.
5496 7. Some code has been added to the store_xxx() functions to reduce the amount
5497 of flapping under certain conditions.
5499 8. log_incoming_interface used to affect only the <= reception log lines. Now
5500 it causes the local interface and port to be added to several more SMTP log
5501 lines, for example "SMTP connection from", and rejection lines.
5503 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5505 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5507 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5508 could be overwritten at the end of the current message (or the start of a
5509 new message if it was set in a HELO ACL). The value is now preserved for
5510 the duration of the SMTP connection.
5512 12. If a transport had a headers_rewrite setting, and a matching header line
5513 contained an unqualified address, that address was qualified, even if it
5514 did not match any rewriting rules. The underlying bug was that the values
5515 of the flags that permit the existence of unqualified sender and recipient
5516 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5517 non-local messages, and by -bnq for local messages) were not being
5518 preserved with the message after it was received.
5520 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5521 "next input=" as part of the text comprising the host identity instead of
5522 the correct text. The code was using the same buffer for two different
5523 strings. However, depending on which order the printing function evaluated
5524 its arguments, the bug did not always show up. Under Linux, for example, my
5525 test suite worked just fine.
5527 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5528 doesn't work with some older versions of Perl. It has been changed to "my",
5529 which in any case is probably the better facility to use.
5531 15. A really picky compiler found some instances of statements for creating
5532 error messages that either had too many or two few arguments for the format
5535 16. The size of the buffer for calls to the DNS resolver has been increased
5536 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5537 for addresses that have a lot of PTR records. This alleviates a problem; it
5538 does not fully solve it.
5540 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5541 buffer now truncates the list and logs the incident, which is the same
5542 action as happens when Exim is looking up a host name and its aliases.
5543 Previously in this situation something unpredictable would happen;
5544 sometimes it was "internal error: store_reset failed".
5546 18. If a server dropped the connection unexpectedly when an Exim client was
5547 using GnuTLS and trying to read a response, the client delivery process
5548 crashed while trying to generate an error log message.
5550 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5551 string, for example:
5553 warn message = H1: something\nH2: something
5555 the text was added as a single header line from Exim's point of view
5556 though it ended up OK in the delivered message. However, searching for the
5557 second and subsequent header lines using $h_h2: did not work. This has been
5558 fixed. Similarly, if a system filter added multiple headers in this way,
5559 the routers could not see them.
5561 20. Expanded the error message when iplsearch is called with an invalid key to
5562 suggest using net-iplsearch in a host list.
5564 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5565 ACLs are no longer actually imposed (and a message to that effect is
5568 22. If a "gecos" field in a passwd entry contained escaped characters, in
5569 particular, if it contained a \" sequence, Exim got it wrong when building
5570 a From: or a Sender: header from that name. A second bug also caused
5571 incorrect handling when an unquoted " was present following a character
5572 that needed quoting.
5574 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5575 was not being matched caselessly.
5577 24. Arranged for all hyphens in the exim.8 source to be escaped with
5580 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5581 properly. Recipient callout cache records were still being keyed to include
5582 the sender, even when use_sender was set false. This led to far more
5583 callouts that were necessary. The sender is no longer included in the key
5584 when use_sender is false.
5586 26. Added "control = submission" modifier to ACLs.
5588 27. Added the ${base62d: operator to decode base 62 numbers.
5590 28. dnsdb lookups can now access SRV records.
5592 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5593 the configuration file.
5595 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5596 (-v) mode. This makes the output for a verbose queue run more intelligible.
5598 31. Added a use_postmaster feature to recipient callouts.
5600 32. Added the $body_zerocount variable, containing the number of binary zero
5601 bytes in the message body.
5603 33. The time of last modification of the "new" subdirectory is now used as the
5604 "mailbox time last read" when there is a quota error for a maildir
5607 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5609 35. Added +ignore_unknown as a special item in host lists.
5611 36. Code for decoding IPv6 addresses in host lists is now included, even if
5612 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5613 address was recognized as an IP address, but was then not correctly decoded
5614 into binary, causing unexpected and incorrect effects when compared with
5621 1. Very minor rewording of debugging text in manualroute to say "list of
5622 hosts" instead of "hostlist".
5624 2. If verify=header_syntax was set, and a header line with an unqualified
5625 address (no domain) and a large number of spaces between the end of the
5626 name and the colon was received, the reception process suffered a buffer
5627 overflow, and (when I tested it) crashed. This was caused by some obsolete
5628 code that should have been removed. The fix is to remove it!
5630 3. When running in the test harness, delay a bit after writing a bounce
5631 message to get a bit more predictability in the log output.
5633 4. Added a call to search_tidyup() just before forking a reception process. In
5634 theory, someone could use a lookup in the expansion of smtp_accept_max_
5635 per_host which, without the tidyup, could leave open a database connection.
5637 5. Added the variables $recipient_data and $sender_data which get set from a
5638 lookup success in an ACL "recipients" or "senders" condition, or a router
5639 "senders" option, similar to $domain_data and $local_part_data.
5641 6. Moved the writing of debug_print from before to after the "senders" test
5644 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5645 problems for message scanning, either using a data ACL, or using
5646 local_scan() because the Received: header was not generated till after they
5647 were called (in order to set the time as the time of reception completion).
5648 I have revised the way this works. The header is now generated after the
5649 body is received, but before the ACL or local_scan() are called. After they
5650 are run, the timestamp in the header is updated.
5656 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5657 before starting a queue runner without re-exec. This happened only when
5658 deliver_drop_privilege was set or when the Exim user was set to root. The
5659 effect of the bug was that timeouts during subsequent deliveries caused
5660 crashes instead of being properly handled. The handler is now left at its
5661 default (and expected) setting.
5663 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5664 message, again when deliver_drop_privilege is set or Exim is run as root.
5665 The bug described in (1) was not present in this case, but the tidying up
5666 of the other signals was missing. I have made the two cases consistent.
5668 3. The ignore_target_hosts setting on a manualroute router was being ignored
5669 for hosts that were looked up using the /MX notation.
5671 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5674 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5675 operated on the sender address. After changing the $sender_address to <>
5676 for the sender address verify, Exim was re-instated it as the original
5677 (before rewriting) address, but remembering that it had rewritten it, so it
5678 wasn't rewriting it again. This bug also had the effect of breaking the
5679 sender address verification caching when the sender address was rewritten.
5681 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5682 This has been changed so that if the ip literal address matches
5683 ignore_target_hosts, the router declines.
5685 7. Added expansion conditions match_domain, match_address, and match_local_
5686 part (NOT match_host).
5688 8. The placeholder for the Received: header didn't have a length field set.
5690 9. Added code to Exim itself and to exim_lock to test for a specific race
5691 condition that could lead to file corruption when using MBX delivery. The
5692 issue is with the lockfile that is created in /tmp. If this file is removed
5693 after a process has opened it but before that process has acquired a lock,
5694 there is the potential for a second process to recreate the file and also
5695 acquire a lock. This could lead to two Exim processes writing to the file
5696 at the same time. The added code performs the same test as UW imapd; it
5697 checks after acquiring the lock that its file descriptor still refers to
5698 the same named file.
5700 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5701 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5704 11. Added the smtp_active_hostname option. If used, this will typically be made
5705 to depend on the incoming interface address. Because $interface_address is
5706 not set up until the daemon has forked a reception process, error responses
5707 that can happen earlier (such as "too many connections") no longer contain
5710 12. If an expansion in a condition on a "warn" statement fails because a lookup
5711 defers, the "warn" statement is abandoned, and the next ACL statement is
5712 processed. Previously this caused the whole ACL to be aborted.
5714 13. Added the iplsearch lookup type.
5716 14. Added ident_timeout as a log selector.
5718 15. Added tls_certificate_verified as a log selector.
5720 16. Added a global option tls_require_ciphers (compare the smtp transport
5721 option of the same name). This controls incoming TLS connections.
5723 17. I finally figured out how to make tls_require_ciphers do a similar thing
5724 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5725 before starting the TLS session.
5727 18. Tabs are now shown as \t in -bP output.
5729 19. If the log selector return_path_on_delivery was set, Exim crashed when
5730 bouncing a message because it had too many Received: header lines.
5732 20. If two routers both had headers_remove settings, and the first one included
5733 a superfluous trailing colon, the final name in the first list and the
5734 first name in the second list were incorrectly joined into one item (with a
5735 colon in the middle).
5741 1. Added -C and -D options to the exinext utility, mainly to make it easier
5742 to include in the automated testing, but these could be helpful when
5743 multiple configurations are in use.
5745 2. The exinext utility was not formatting the output nicely when there was
5746 an alternate port involved in the retry record key, nor when there was a
5747 message id as well (for retries that were specific to a specific message
5748 and a specific host). It was also confused by IPv6 addresses, because of
5749 the additional colons they contain. I have fixed the IPv4 problem, and
5750 patched it up to do a reasonable job for IPv6.
5752 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5753 delivery, the log line now contains "pipelined" if PIPELINING was used.
5755 4. An SMTP transport process used to panic and die if the bind() call to set
5756 an explicit outgoing interface failed. This has been changed; it is now
5757 treated in the same way as a connect() failure.
5759 5. A reference to $sender_host_name in the part of a conditional expansion
5760 that was being skipped was still causing a DNS lookup. This no longer
5763 6. The def: expansion condition was not recognizing references to header lines
5764 that used bh_ and bheader_.
5766 7. Added the _cache feature to named lists.
5768 8. The code for checking quota_filecount in the appendfile transport was
5769 allowing one more file than it should have been.
5771 9. For compatibility with Sendmail, the command line option
5779 and sets the incoming protocol and host name (for trusted callers). The
5780 host name and its colon can be omitted when only the protocol is to be set.
5781 Note the Exim already has two private options, -pd and -ps, that refer to
5782 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5783 "s", but I don't think that's a major issue.
5785 10. A number of refactoring changes to the code, none of which should affect
5788 (a) The number of logging options was getting close to filling up the
5789 32-bit word that was used as a bit map. I have split them into two classes:
5790 those that are passed in the argument to log_write(), and those that are
5791 only ever tested independently outside of that function. These are now in
5792 separate 32-bit words, so there is plenty of room for expansion again.
5793 There is no change in the user interface or the logging behaviour.
5795 (b) When building, for example, log lines, the code previously used a
5796 macro that called string_cat() twice, in order to add two strings. This is
5797 not really sufficiently general. Furthermore, there was one instance where
5798 it was actually wrong because one of the argument was used twice, and in
5799 one call a function was used. (As it happened, calling the function twice
5800 did not affect the overall behaviour.) The macro has been replaced by a
5801 function that can join an arbitrary number of extra strings onto a growing
5804 (c) The code for expansion conditions now uses a table and a binary chop
5805 instead of a serial search (which was left over from when there were very
5806 few conditions). Also, it now recognizes conditions like "pam" even when
5807 the relevant support is not compiled in: a suitably worded error message is
5808 given if an attempt is made to use such a condition.
5810 11. Added ${time_interval:xxxxx}.
5812 12. A bug was causing one of the ddress fields not to be passed back correctly
5813 from remote delivery subprocesses. The field in question was not being
5814 subsequently used, so this caused to problems in practice.
5816 13. Added new log selectors queue_time and deliver_time.
5818 14. Might have fixed a bug in maildirsizefile handling that threw up
5819 "unexpected character" debug warnings, and recalculated the data
5820 unnecessarily. In any case, I expanded the warning message to give more
5823 15. Added the message "Restricted characters in address" to the statements in
5824 the default ACL that block characters like @ and % in local parts.
5826 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5827 Three changes have been made:
5829 (a) There was a serious bug; a negative response to MAIL caused the whole
5830 recipient domain to be cached as invalid, thereby blocking all messages
5831 to all local parts at the same domain, from all senders. This bug has
5832 been fixed. The domain is no longer cached after a negative response to
5833 MAIL if the sender used is not empty.
5835 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5838 (c) A new callout option, "use_sender" has been added for people who want
5839 the modified behaviour.
5845 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5848 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5849 indeed breaks things for older releases.
5851 3. Added additional logging to the case where there is a problem reading data
5852 from a filter that is running in a subprocess using a pipe, in order to
5853 try to track down a specific problem.
5855 4. Testing facility fudge: when running in the test harness and attempting
5856 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5857 getting "No route to host". Convert this to a timeout.
5859 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5862 6. Some OS don't have socklen_t but use size_t instead. This affects the
5863 fifth argument of getsockopt() amongst other things. This is now
5864 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5865 can be set for individual OS. I have set it for SunOS5, OSF1, and
5866 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5867 some earlier ones do not.
5869 7. Change 4.30/15 was not doing the test caselessly.
5871 8. The standard form for an IPv6 address literal was being rejected by address
5872 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5873 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5874 this, as well as the form without the "IPv6" on the front (but only when
5875 address literals are enabled, of course).
5877 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5879 10. Exim crashed if a message with an empty sender address specified by -f
5880 encountered a router with an errors_to setting. This could be provoked only
5881 by a command such as
5885 where an empty string was supplied; "<>" did not hit this bug.
5887 11. Installed PCRE release 4.5.
5889 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5890 remained set. It is now erased.
5892 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5893 times from message ids (which are base 36 rather than the normal 62).
5895 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5896 were being counted as actual protocol errors, and logged if the log
5897 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5898 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5899 and DATA following a set of rejected RCPTs do not count as protocol errors.
5900 In other words, Exim assumes they were pipelined, though this may not
5901 actually be the case. Of course, in all cases the client gets an
5902 appropriate error code.
5904 15. If a lookup fails in an ACL condition, a message about the failure may
5905 be available; it is used if testing the ACL cannot continue, because most
5906 such messages specify what the cause of the deferral is. However, some
5907 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5908 that caused an old message to be retained and used if a later statement
5909 caused a defer, replacing the real cause of the deferral.
5911 16. If an IP address had so many PTR records that the DNS lookup buffer
5912 was not large enough to hold them, Exim could crash while trying to process
5913 the truncated data. It now detects and logs this case.
5915 17. Further to 4.21/58, another change has been made: if (and only if) the
5916 first line of a message (the first header line) ends with CRLF, a bare LF
5917 in a subsequent header line has a space inserted after it, so as not to
5918 terminate the header.
5920 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5921 unnecessarily, used atoi() instead of strtol(), and didn't check the
5922 termination when getting file sizes from file names by regex.
5924 19. Completely re-implemented the support for maildirsize files, in the light
5925 of a number of problems with the previous contributed implementation
5926 (4.30/29). In particular:
5928 . If the quota is zero, the maildirsize file is maintained, but no quota is
5931 . If the maildir directory does not exist, it is created before any attempt
5932 to write a maildirsize file.
5934 . The quota value in the file is just a cache; if the quota is changed in
5935 the transport, the new value overrides.
5937 . A regular expression is available for excluding directories from the
5940 20. The autoreply transport checks the characters in options that define the
5941 message's headers; it allows continued headers, but it was checking with
5942 isspace() after an embedded newline instead of explicitly looking for a
5945 21. If all the "regular" hosts to which an address was routed had passed their
5946 expiry times, and had not reached their retry times, the address was
5947 bounced, even if fallback hosts were defined. Now Exim should go on to try
5950 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5951 equivalent code in the SMTP transport. Some hosts send humungous responses
5952 to HELO/EHLO, more than 1024 it seems.
5954 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5955 gives compiler warnings in some environments. I've now done it "properly",
5958 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5959 (because of problems with the built-in one) was declared to return uschar *
5960 instead of char *, causing compiler failure.
5962 25. Fixed a file descriptor leak when processing alias/forward files.
5964 26. Fixed a minor format string issue in dbfn.c.
5966 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5968 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5969 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5972 29. When a sender address is verified, it is cached, to save repeating the test
5973 when there is more than one recipient in a message. However, when the
5974 verification involves a callout, it is possible for different callout
5975 options to be set for different recipients. It is too complicated to keep
5976 track of this in the cache, so now Exim always runs a verification when a
5977 callout is required, relying on the callout cache for the optimization.
5978 The overhead is duplication of the address routing, but this should not be
5981 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5982 to be verified with callout=postmaster, and the main callout worked but the
5983 postmaster check failed, the verification correctly failed. However, if a
5984 subsequent RCPT command asked for sender verification *without* the
5985 postmaster check, incorrect caching caused this verification also to fail,
5988 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5989 it was not caching the DNS options (qualify_single, search_parents) that
5990 were used when the lookup failed. A subsequent lookup with different
5991 options therefore always gave the same answer, though there were cases
5992 where it should not have. (Example: a "domains = !$mx_any" option on a
5993 dnslookup router: the "domains" option is always processed without any
5994 widening, but the router might have qualify_single set.) Now Exim uses the
5995 cached value only when the same options are set.
5997 32. Added John Jetmore's "exipick" utility to the distribution.
5999 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6000 than a timeout (e.g. a certificate is required, and is not provided), an
6001 Exim server now closes the connection immediately. Previously it waited for
6002 the client to close - but if the client is SSL, it seems that they each
6003 wait for each other, leading to a delay before one of them times out.
6005 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6006 maintained 0.8.x compatibility because I don't think many are using it, and
6007 it is clearly obsolete.
6009 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6012 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6013 client certificate was expired. A simple patch fixes this, though I don't
6014 understand the full logic of why the verify callback is called multiple
6017 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6018 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6019 which causes problems with some clients (such as the Certicom SSL Plus
6020 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6021 disables the coutermeasure allowing Eudora to connect."
6023 38. Exim was not checking that a write() to a log file succeeded. This could
6024 lead to Bad Things if a log got too big, in particular if it hit a file
6025 size limit. Exim now panics and dies if it cannot write to a log file, just
6026 as it does if it cannot open a log file.
6028 39. Modified OS/Makefile-Linux so that it now contains
6030 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6032 The two -D definitions ensure that Exim is compiled with large file
6033 support, which makes it possible to handle log files that are bigger than
6036 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6037 instance) a domain was checked against a named list that involved a lookup,
6038 causing $domain_data to be set, then another domain was checked against the
6039 same list, then the first domain was re-checked, the value of $domain_data
6040 after the final check could be wrong. In particular, if the second check
6041 failed, it could be set empty. This bug probably also applied to
6044 41. The strip_trailing_dot option was not being applied to the address given
6045 with the -f command-line option.
6047 42. The code for reading a message's header from the spool was incrementing
6048 $received_count, but never initializing it. This meant that the value was
6049 incorrect (doubled) while delivering a message in the same process in which
6050 it was received. In the most common configuration of Exim, this never
6051 happens - a fresh exec is done - but it can happen when
6052 deliver_drop_privilege is set.
6054 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6055 it now includes up to 150 characters of the unexpected data in the log
6058 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6059 and building data strings. The size of both of these buffers was 10 000
6060 bytes - far larger than anybody would *ever* want, thought I. Needless to
6061 say, somebody hit the limit. I have increased the maximum line length to
6062 20 000 and the maximum data length of concatenated lines to 100 000. I have
6063 also fixed two bugs, because there was no checking on these buffers. Tsk,
6064 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6065 buffer is too small.
6067 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6068 lsearch lookups. Now it does.
6070 46. When parsing a route_list item in a manualroute router, a fixed-length
6071 buffer was used for the list of hosts. I made this 1024 bytes long,
6072 thinking that nobody would ever have a list of hosts that long. Wrong.
6073 Somebody had a whole pile of complicated expansion conditions, and the
6074 string was silently truncated, leading to an expansion error. It turns out
6075 that it is easier to change to an unlimited length (owing to other changes
6076 that have happened since this code was originally written) than to build
6077 structure for giving a limitation error. The length of the item that
6078 expands into the list of hosts is now unlimited.
6080 47. The lsearch lookup could not handle data where the length of text line was
6081 more than 4095 characters. Such lines were truncated, leading to shortened
6082 data being returned. It should now handle lines of any length.
6084 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6085 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6088 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6089 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6090 no longer happens in comment lines. A list of replacements is now placed
6091 at the head of all of the source files, except those whose only change is
6092 to replace PERL_COMMAND in the very first #! line.
6094 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6095 messages on the queue, with a bottom-up merge sort, using code contributed
6096 by Michael Haardt. This should make operations like -bp somewhat faster on
6097 large queues. It won't affect queue runners, except when queue_run_in_order
6100 51. Installed eximstats 1.31 in the distribution.
6102 52. Added support for SRV lookups to the dnslookup router.
6104 53. If an ACL referred to $message_body or $message_body_end, the value was not
6105 reset for any messages that followed in the same SMTP session.
6107 54. The store-handling optimization for building very long strings was not
6108 differentiating between the different store pools. I don't think this
6109 actually made any difference in practice, but I've tidied it.
6111 55. While running the routers to verify a sender address, $sender_address
6112 was still set to the sender address. This is wrong, because when routing to
6113 send a bounce to the sender, it would be empty. Therefore, I have changed
6114 it so that, while verifying a sender address, $sender_address is set to <>.
6115 (There is no change to what happens when verifying a recipient address.)
6117 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6118 target A or AAAA records (if not already returned) without resetting the
6119 qualify_single or search_parents options of the DNS resolver. These are
6120 inappropriate in this case because the targets of MX and SRV records must
6121 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6122 target that, when qualified, matched something in the local domain. These
6123 two options are now turned off when doing these lookups.
6125 57. It seems that at least some releases of Reiserfs (which does not have the
6126 concept of a fixed number of inodes) returns zero and not -1 for the
6127 number of available inodes. This interacted badly with check_spool_inodes,
6128 which assumed that -1 was the "no such thing" setting. What I have done is
6129 to check that the total number of inodes is greater than zero before doing
6130 the test of how many are available.
6132 58. When a "warn" ACL statement has a log_message modifier, the message is
6133 remembered, and not repeated. This is to avoid a lot of repetition when a
6134 message has many recipients that cause the same warning to be written.
6135 However, Exim was preserving the list of already written lines for an
6136 entire SMTP session, which doesn't seem right. The memory is now reset if a
6137 new message is started.
6139 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6140 debugging output unless log_rewrite was also set.
6142 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6143 of (int)(handle) when we know that handle contains (void *)(-1).
6145 61. The Exim daemon panic-logs an error return when it closes the incoming
6146 connection. However "connection reset by peer" seems to be common, and
6147 isn't really an error worthy of noting specially, so that particular error
6150 62. When Exim is trying to find all the local interfaces, it used to panic and
6151 die if the ioctl to get the interface flags failed. However, it seems that
6152 on at least one OS (Solaris 9) it is possible to have an interface that is
6153 included in the list of interfaces, but for which you get a failure error
6154 for this call. This happens when the interface is not "plumbed" into a
6155 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6156 failure of the "get flags" call assumes that the interface is down.
6158 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6159 makes life easier for people who are doing arithmetic on fields extracted
6160 from dates, where you often get leading zeros that should not be
6161 interpreted as octal.
6163 64. Added qualify_domain to the redirect router, to override the global
6166 65. If a pathologically long header line contained very many addresses (the
6167 report of this problem mentioned 10 000) and each of them was rewritten,
6168 Exim could use up a very large amount of memory. (It kept on making new
6169 copies of the header line as it rewrote, and never released the old ones.)
6170 At the expense of a bit more processing, the header rewriting function has
6171 been changed so that it no longer eats memory in this way.
6173 66. The generation of the Received: header has been moved from the time that a
6174 message starts to be received, to the time that it finishes. The timestamp
6175 in the Received: header should now be very close to that of the <= log
6176 line. There are two side-effects of this change:
6178 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6179 logged header lines no longer include the local Received: line, because
6180 it has not yet been created. The same applies to a copy of the message
6181 that is returned to a non-SMTP sender when a message is rejected.
6183 (b) When a filter file is tested using -bf, no additional Received: header
6184 is added to the test message. After some thought, I decided that this
6187 This change does not affect the value of $received_for. It is still set
6188 after address rewriting, but before local_scan() is called.
6190 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6192 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6193 gave an unhelpful panic error message, and a defer error. I have managed to
6194 change this behaviour so that it now rejects any supplied certificate,
6195 which seems right, as the list of acceptable certificates is empty.
6197 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6198 gave an unhelpful defer error. I have not managed to make this reject any
6199 supplied certificates, but the error message it gives is "no certificate
6200 supplied", which is not helpful.
6202 70. exigrep's output now also includes lines that are not associated with any
6203 message, but which match the given pattern. Implemented by a patch from
6204 Martin Sluka, which also tidied up the Perl a bit.
6206 71. Recipient callout verification, like sender verification, was using <> in
6207 the MAIL FROM command. This isn't really the right thing, since the actual
6208 sender may affect whether the remote host accepts the recipient or not. I
6209 have changed it to use the actual sender in the callout; this means that
6210 the cache record is now keyed on a recipient/sender pair, not just the
6211 recipient address. There doesn't seem to be a real danger of callout loops,
6212 since a callout by the remote host to check the sender would use <>.
6213 [SEE ABOVE: changed after hitting problems.]
6215 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6216 temporary errors. However, in the case of such a code being given after
6217 the end of a data transmission (i.e. after ".") Exim was failing to write
6218 a retry record for the message. (Yes, there was some broken host that was
6219 actually sending 8xx at this point.)
6221 73. An unknown lookup type in a host list could cause Exim to panic-die when
6222 the list was checked. (An example that provoked this was putting <; in the
6223 middle of a list instead of at the start.) If this happened during a DATA
6224 ACL check, a -D file could be left lying around. This kind of configuration
6225 error no longer causes Exim to die; instead it causes a defer error. The
6226 incident is still logged to the main and panic logs.
6228 74. Buglet left over from Exim 3 conversion. The message "too many messages
6229 in one connection" was written to the rejectlog but not the mainlog, except
6230 when address rewriting (yes!) was being logged.
6232 75. Added write_rejectlog option.
6234 76. When a system filter was run not as root (that is, when system_filter_user
6235 was set), the values of the $n variables were not being returned to the
6236 main process; thus, they were not subsequently available in the $sn
6239 77. Added +return_path_on_delivery log selector.
6241 78. A connection timeout was being treated differently from recipients deferred
6242 when testing hosts_max_try with a message that was older than the host's
6243 retry timeout. (The host should not be counted, thus allowing all hosts to
6244 be tried at least once before bouncing.) This may have been the cause of an
6245 occasionally reported bug whereby a message would remain on the queue
6246 longer than the retry timeout, but would be bounced if a delivery was
6247 forced. I say "may" because I never totally pinned down the problem;
6248 setting up timeout/retry tests is difficult. See also the next item.
6250 79. The ultimate address timeout was not being applied to errors that involved
6251 a combination of host plus message (for example, a timeout on a MAIL
6252 command). When an address resolved to a number of possible hosts, and they
6253 were not all tried for each delivery (e.g. because of hosts_max_try), a
6254 message could remain on the queue longer than the retry timeout.
6256 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6259 81. Fixed an obscure SMTP outgoing bug which required at least the following
6260 conditions: (a) there was another message waiting for the same server;
6261 (b) the server returned 5xx to all RCPT commands in the first message so
6262 that the message was not completed; (c) the server dropped the connection
6263 or gave a negative response to the RSET that Exim sends to abort the
6264 transaction. The observed case was a dropped connection after DATA that had
6265 been sent in pipelining mode. That is, the server had advertised PIPELINING
6266 but was not implementing it correctly. The effect of the bug was incorrect
6267 behaviour, such as trying another host, and this could lead to a crash.
6273 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6274 and daemon.c were passed as pointers to ints; they should have been
6275 pointers to socklen_t variables (which are typically unsigned ints).
6277 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6280 3. Fixed a really odd bug that affected only the testing scheme; patching a
6281 certain fixed string in the binary changed the value of another string that
6282 happened to be identical to the end of the original first string.
6284 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6285 name", it returns that address as the IP address. On some operating
6286 systems (e.g. Solaris), it also passes back the IP address string as the
6287 "host name". However, on others (e.g. Linux), it passes back an empty
6288 string. Exim wasn't checking for this, and was changing the host name to an
6289 empty string, assuming it had been canonicalized.
6291 5. Although rare, it is permitted to have more than one PTR record for a given
6292 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6293 all the names associated with an address, because they do in Solaris.
6294 However, it seems that they do not in Linux for data that comes from the
6295 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6296 I found this out when I moved to a new Linux workstation and tried to run
6297 the Exim test suite.
6299 To get round this problem I have changed the code so that it now does its
6300 own call to the DNS to look up PTR records when searching for a host name.
6301 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6302 addresses that are only in /etc/hosts are still found.
6304 This behaviour is, however, controlled by an option called host_lookup_
6305 order, which defaults to "bydns:byaddr". If people want to use the other
6306 order, or indeed, just use one or the other means of lookup, they can
6307 specify it in this variable.
6309 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6310 some operating systems, this comes back from gethostbyaddr() as an empty
6311 string, and this is what Exim used to test for. However, it seems that in
6312 other systems, "." is yielded. Exim now tests for this case too.
6314 7. The values of check_spool_space and check_log_space are now held internally
6315 as a number of kilobytes instead of an absolute number of bytes. If a
6316 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6317 kilobyte. This means that much larger values can be stored.
6319 8. Exim monitor: an attempt to get the action menu when not actually pointing
6320 at a message produces an empty menu entitled "No message selected". This
6321 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6322 no entries in it ("Shell widget menu has zero width and/or height"). So I
6323 have added a single, blank menu entry in this case.
6325 9. Added ${quote_local_part.
6327 10. MIME decoding is now applied to the contents of Subject: header lines when
6330 11. Now that a reference to $sender_host_address automatically causes a reverse
6331 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6332 host lookup before query-style lookups in lists that might use this
6333 variable. This has therefore been abolished, and the "net-" prefix is no
6334 longer necessary for query-style lookups.
6336 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6337 have been a typo for LFLAGS, so it has been changed.
6339 13. The install script calls Exim with "-C /dev/null" in order to find the
6340 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6341 to be output. However, since Exim outputs its version number before the
6342 error, it didn't break the script. It just looked ugly. I fixed this by
6343 always allowing "-C /dev/null" if the caller is root.
6345 14. Ignore overlarge ACL variable number when reading spool file - insurance
6346 against a later release with more variables having written the file.
6348 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6349 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6350 this, as well as the form without the "IPv6" on the front.
6352 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6353 OS/Makefile-Darwin file.
6355 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6356 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6358 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6359 variable. However, some OS use ints for inodes. I've added cast to long int
6360 to get rid of the compiler warning.
6362 19. I had forgotten to lock out "/../" in configuration file names when
6363 ALT_CONFIG_PREFIX was set.
6365 20. Routers used for verification do not need to specify transports. However,
6366 if such a router generated a host list, and callout was configured, Exim
6367 crashed, because it could not find a port number from the (non-existent)
6368 transport. It now assumes port 25 in this circumstance.
6370 21. Added the -t option to exigrep.
6372 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6373 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6374 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6375 them is set, LOOKUP_LSEARCH is forced.
6377 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6379 24. Added sender and host information to the "rejected by local_scan()" log
6380 line; previously there was no indication of these.
6382 25. Added .include_if_exists.
6384 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6385 Linux. It turns out that not all file systems support this. Apparently some
6386 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6387 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6388 not supported on the file descriptor, is now ignored when Exim is trying to
6389 sync a directory. This applies only to Linux.
6391 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6393 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6394 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6397 29. Added support for maildirsize files from supplied patch (modified a bit).
6399 30. The use of :fail: followed by an empty string could lead Exim to respond to
6400 sender verification failures with (e.g.):
6402 550 Verification failed for <xxx>
6403 550 Sender verify failed
6405 where the first response line was missing the '-' that indicates it is not
6406 the final line of the response.
6408 31. The loop for finding the name of the user that called Exim had a hardwired
6409 limit of 10; it now uses the value of finduser_retries, which is used for
6410 all other user lookups.
6412 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6415 33. Exim was neglecting to zero errno before one call of strtol() when
6416 expanding a string and expecting an integer value. On some systems this
6417 resulted in spurious "integer overflow" errors. Also, it was casting the
6418 result into an int without checking.
6420 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6421 did not work. The code looks as if it has *never* worked, though it appears
6422 to have been documented since at least release 1.62. I have made it work.
6424 35. The "timeout_DNS" error in retry rules, also documented since at least
6425 1.62, also never worked. As it isn't clear exactly what this means, and
6426 clearly it isn't a major issue, I have abolished the feature by treating it
6427 as "timeout", and writing a warning to the main and panic logs.
6429 36. The display of retry rules for -brt wasn't always showing the error code
6432 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6433 timeout_connect_A, timeout_connect_MX.
6435 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6436 to the empty sender.
6438 39. The daemon was not analysing the content of -oX till after it had closed
6439 stderr and disconnected from the controlling terminal. This meant that any
6440 syntax errors were only noted on the panic log, and the return code from
6441 the command was 0. By re-arranging the code a little, I've made the
6442 decoding happen first, so such errors now appear on stderr, and the return
6443 code is 1. However, the actual setting up of the sockets still happens in
6444 the disconnected process, so errors there are still only recorded on the
6447 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6448 connections (as happens on some IP stacks) was logged at start up time as
6449 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6450 it from "IPv6 and IPv4", which means that two separate sockets are being
6453 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6454 says whether AF_INET or AF_INET6 was passed as an argument.
6456 42. Exiwhat output was messed up when time zones were included in log
6459 43. Exiwhat now gives more information about the daemon's listening ports,
6460 and whether -tls-on-connect was used.
6462 44. The "port" option of the smtp transport is now expanded.
6464 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6465 silently ignored. Now an error message is written to the main and panic
6468 46. There's a new ACL modifier called "logwrite" which writes to a log file
6469 as soon as it is encountered.
6471 47. Added $local_user_uid and $local_user_gid at routing time.
6473 48. Exim crashed when trying to verify a sender address that was being
6476 49. Exim was recognizing only a space character after ".include". It now also
6477 recognizes a tab character.
6479 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6480 extracting the relevant information from the specification. The man page no
6481 longer contains scrambled data for the -d option, and I've added a section
6482 at the front about calling Exim under different names.
6484 51. Added "extra_headers" argument to the "mail" command in filter files.
6486 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6489 53. Installed eximstats 1.29.
6491 54. Added transport_filter_timeout as a generic transport option.
6493 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6494 Cc: header lines. This was required by RFC 822, but it not required by RFC
6497 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6498 incoming messages that did not have them. Now it does so only if the
6499 message originates locally, that is, if there is no associated remote host
6500 address. When Resent- header lines are present, this applies to the Resent-
6501 lines rather than the non-Resent- lines.
6503 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6504 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6506 58. Messages for configuration errors now include the name of the main
6507 configuration file - useful now that there may be more than one file in a
6508 list (.included file names were always shown).
6510 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6511 for those rare installations that do not start the daemon as root or run it
6512 setuid root. I've cut out the call to initgroups() if the daemon is not
6515 60. The Exim user and group can now be bound into the binary as text strings
6516 that are looked up at the start of Exim's processing.
6518 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6520 62. Added $mailstore_basename variable.
6522 63. Installed patch to sieve.c from Michael Haardt.
6524 64. When Exim failed to open the panic log after failing to open the main log,
6525 the original message it was trying to log was written to stderr and debug
6526 output, but if they were not available (the usual case in production), it
6527 was lost. Now it is written to syslog before the two lines that record the
6528 failures to open the logs.
6530 65. Users' Exim filters run in subprocesses under the user's uid. It is
6531 possible for a "deliver" command or an alias in a "personal" command to
6532 provoke an address rewrite. If logging of address rewriting is configured,
6533 this fails because the process is not running as root or exim. There may be
6534 a better way of dealing with this, but for the moment (because 4.30 needs
6535 to be released), I have disabled address rewrite logging when running a
6536 filter in a non-root, non-exim process.
6542 1. The buildconfig auxiliary program wasn't quoting the value set for
6543 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6544 not defined. This bug was masked in 4.22 by the effect that was fixed in
6547 2. Some messages that were rejected after a message id was allocated were
6548 shown as "incomplete" by exigrep. It no longer does this for messages that
6549 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6551 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6552 have allow_domain_literals set, the ID did not get logged in the <= line.
6553 Domain literals are now always recognized in Message-ID: header lines.
6555 4. The first argument for a ${extract expansion item is the key name or field
6556 number. Leading and trailing spaces in this item were not being ignored,
6557 causing some misleading effects.
6559 5. When deliver_drop_privilege was set, single queue runner processes started
6560 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6561 same command in the process it spins off) were not dropping privilege.
6563 6. When the daemon running as "exim" started a queue runner, it always
6564 re-executed Exim in the spun-off process. This is a waste of effort when
6565 deliver_drop_privilege is set. The new process now just calls the
6566 queue-runner function directly.
6572 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6575 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6576 ignored. Though the use of -oP was forcing the writing of a pid file, it
6577 was always written to the default place.
6579 3. If the message "no IP address found for host xxxx" is generated during
6580 incoming verification, it is now followed by identification of the incoming
6581 connection (so you can more easily find what provoked it).
6583 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6585 5. Added some features to "harden" Exim a bit more against certain attacks:
6587 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6588 be put in Local/Makefile. This is like the never_users runtime option,
6589 but it cannot be overridden. The default setting is "root".
6591 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6592 prefix string with which any file named in a -C command line option
6595 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6596 is retained for -C and -D only if the caller of Exim is root. Without
6597 it, the exim user may also use -C and -D and retain privilege.
6599 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6600 command line option is disabled.
6602 6. Macro names set by the -D option must start with an upper case letter, just
6603 like macro names defined in the configuration file.
6605 7. Added "dereference=" facility to LDAP.
6607 8. Two instances of the typo "uknown" in the source files are fixed.
6609 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6610 the Configure-Makefile script screwed up while processing it.
6612 10. Incorporated PCRE 4.4.
6614 11. The SMTP synchronization check was not operating right at the start of an
6615 SMTP session. For example, it could not catch a HELO sent before the client
6616 waited for the greeting. There is now a check for outstanding input at the
6617 point when the greeting is written. Because of the duplex, asynchronous
6618 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6619 way, but not yet received, when the check is performed.
6621 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6622 on TCP/IP sockets, because this apparently causes some broken clients to
6625 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6626 unchanged) from the Cygwin maintainer.
6628 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6629 was supported instead of testing for mbx. Effectively a typo.
6631 15. The spa authenticator server code was not checking that the input it
6632 received was valid base64.
6634 16. The debug output line for the "set" modifier in ACLs was not showing the
6635 name of the variable that was being set.
6637 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6639 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6640 lookup on its own. Something else was needed to trigger the lookup. For
6641 example, a match in host_lookup or the need for a host name in a host list.
6642 Now, if $sender_host_name is referenced and the host name has not yet been
6643 looked up, a lookup is performed. If the lookup fails, the variable remains
6644 empty, and $host_lookup_failed is set to "1".
6646 19. Added "eqi" as a case-independent comparison operator.
6648 20. The saslauthd authentication condition could segfault if neither service
6649 nor realm was specified.
6651 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6652 error message that was logged was misleading, and incoming SMTP
6653 connections were dropped. The message is now more accurate, and temporary
6654 errors are given to SMTP connections.
6656 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6657 (except RSET) with a 421 error, until QUIT is received. However, it was
6658 failing to send a response to QUIT.
6660 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6661 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6662 ACL is now run after the helo_try_verify_hosts code.
6664 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6665 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6666 case-independent, so other case variants are also recognized). Apparently
6667 some systems use these upper case variants.
6669 25. If more than two messages were waiting for the same host, and a transport
6670 filter was specified for the transport, Exim sent two messages over the
6671 same TCP/IP connection, and then failed with "socket operation on non-
6672 socket" when it tried to send the third.
6674 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6676 27. The extern definition of crypt16() in expand.c was not being excluded when
6677 the OS had its own crypt16() function.
6679 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6680 as a preferred synonym for return_size_limit, both as an option and as an
6683 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6685 30. Changed the default configuration ACL to relax the local part checking rule
6686 for addresses that are not in any local domains. For these addresses,
6687 slashes and pipe symbols are allowed within local parts, but the sequence
6688 /../ is explicitly forbidden.
6690 31. SPA server authentication was not clearing the challenge buffer before
6693 32. log_message in a "warn" ACL statement was writing to the reject log as
6694 well as to the main log, which contradicts the documentation and doesn't
6695 seem right (because no rejection is happening). So I have stopped it.
6697 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6698 However, I am unable to do any testing of this.
6700 34. Fixed an infelicity in the appendfile transport. When checking directories
6701 for a mailbox, to see if any needed to be created, it was accidentally
6702 using path names with one or more superfluous leading slashes; tracing
6703 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6705 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6706 contents are added to the log line that is written for every discarded
6707 recipient. (Previously a log_message setting was ignored.)
6709 36. The ${quote: operator now quotes the string if it is empty.
6711 37. The install script runs exim in order to find its version number. If for
6712 some reason other than non-existence or emptiness, which it checks, it
6713 could not run './exim', it was installing it with an empty version number,
6714 i.e. as "exim-". This error state is now caught, and the installation is
6717 38. An argument was missing from the function that creates an error message
6718 when Exim fails to connect to the socket for saslauthd authentication.
6719 This could cause Exim to crash, or give a corrupted message.
6721 39. Added isip, isip4, and isip6 to ${if conditions.
6723 40. The ACL variables $acl_xx are now saved with the message, and can be
6724 accessed later in routers, transports, and filters.
6726 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6727 strings in the file are not string-expanded.
6729 42. If a MAIL command specified a SIZE value that was too large to fit into an
6730 int variable, the check against message_size_limit failed. Such values are
6731 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6732 day this will have to be increased, but I don't think I want to be around
6733 when emails are that large.
6740 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6741 iconv() is not standard in FreeBSD.
6743 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6744 IPv6 enabled. The observed symptom was a segmentation fault on return from
6745 the function os_common_find_running_interfaces() in src/os.c.
6747 3. In the check_special_case() function in daemon.c I had used "errno" as an
6748 argument name, which causes warnings on some systems. This was basically a
6749 typo, since it was named "eno" in the comments!
6751 4. The code that waits for the clock to tick (at a resolution of some fraction
6752 of a second) so as to ensure message-id uniqueness was always waiting for
6753 at least one whole tick, when it could have waited for less. [This is
6754 almost certainly not relevant at current processor speeds, where it is
6755 unlikely to ever wait at all. But we try to future-proof.]
6757 5. The function that sleeps for a time interval that includes fractions of a
6758 second contained a race. It did not block SIGALRM between setting the
6759 timer, and suspending (a couple of lines later). If the interval was short
6760 and the sigsuspend() was delayed until after it had expired, the suspension
6761 never ended. On busy systems this could lead to processes getting stuck for
6764 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6765 process, before it forks any delivery processes. The open lookup caching
6766 mechanism meant that the open file or database connection was passed into
6767 the delivery process. The problem was that delivery processes always tidy
6768 up cached lookup data. This could cause a problem for the next delivery
6769 process started by the queue runner, because the external queue runner
6770 process does not know about the closure. So the next delivery process
6771 still has data in the lookup cache. In the case of a file lookup, there was
6772 no problem because closing a file descriptor in a subprocess doesn't affect
6773 the parent. However, if the lookup was caching a connection to a database,
6774 the connection was closed, and the second delivery process was likely to
6775 see errors such as "PGSQL: query failed: server closed the connection
6776 unexpectedly". The problem has been fixed by closing all cached lookups
6777 in a queue runner before running a delivery process.
6779 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6780 seem to have the "const" qualifier which it has on other OS. I've
6783 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6784 *of the same type* (client or server) with the same public name that an
6785 error should be diagnosed.
6787 9. When Exim looked up a host name for an IP address, but failed to find the
6788 original IP address when looking up the host name (a safety check), it
6789 output the message "<ip address> does not match any IP for NULL", which was
6790 confusing, to say the least. The bug was that the host name should have
6791 appeared instead of "NULL".
6793 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6794 user that is built into the binary, and the -C or -D options is used, root
6795 privilege is dropped before the configuration file is read. In addition,
6796 logging is switched to stderr instead of the normal log files. If the
6797 configuration then re-defines the Exim user, the unprivileged environment
6798 is probably not what is expected, so Exim logs a panic warning message (but
6801 However, if deliver_drop_privilege is set, the unprivileged state may well
6802 be exactly what is intended, so the warning has been cut out in that case,
6803 and Exim is allowed to try to write to its normal log files.
6809 1. smtp_return_error_details was not giving details for temporary sender
6810 or receiver verification errors.
6812 2. Diagnose a configuration error if two authenticators have the same public
6815 3. Exim used not to create the message log file for a message until the first
6816 delivery attempt. This could be confusing when incoming messages were held
6817 for policy or load reasons. The message log file is now created at the time
6818 the message is received, and an initial "Received" line is written to it.
6820 4. The automatically generated man page for command line options had a minor
6821 bug that caused no ill effects; however, a more serious problem was that
6822 the procedure for building the man page automatically didn't always
6823 operate. Consequently, release 4.20 contains an out-of-date version. This
6824 shouldn't happen again.
6826 5. When building Exim with embedded Perl support, the script that builds the
6827 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6828 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6830 6. The freeze_tell option was not being used for messages that were frozen on
6831 arrival, either by an ACL or by local_scan().
6833 7. Added the smtp_incomplete_transaction log selector.
6835 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6836 was accepting AUTH without a new EHLO.
6838 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6839 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6842 10. Logging of TCP/IP connections (when configured) now happens in the main
6843 daemon process instead of the child process, so that the TCP/IP connection
6844 count is more accurate (but it can never be perfect).
6846 11. The use of "drop" in a nested ACL was not being handled correctly in the
6847 outer ACL. Now, if condition failure induced by the nested "drop" causes
6848 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6849 or "require"), the connection is dropped.
6851 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6852 that yield "discard" can now be used with an "accept" or a "discard" verb,
6853 but an error is generated for any others (because I can't see a useful way
6854 to define what should happen).
6856 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6857 are now processed in the same way as lines in the Exim configuration file.
6858 In particular, continuation lines are supported.
6860 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6862 15. Added -ti meaning -t -i.
6864 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6865 domains, and warn by logging if others are found.
6867 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6868 returned from GIFCONF when Exim is trying to find the list of interfaces on
6869 a host. The code in os.c has been modified to copy each ifreq to an aligned
6870 structure in all cases.
6872 Also, in some cases, the returned ifreq's were being copied to a 'struct
6873 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6874 means the last couple of bytes of an IPv6 address could be chopped if the
6875 ifreq contained only a normal sockaddr (14 bytes storage).
6877 18. Named domain lists were not supported in the hosts_treat_as_local option.
6878 An entry such as +xxxx was not recognized, and was treated as a literal
6881 19. Ensure that header lines added by a DATA ACL are included in the reject log
6882 if the ACL subsequently rejects the message.
6884 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6885 MD5 (which is deprecated).
6887 21. When testing a filter file using -bf, Exim was writing a message when it
6888 took the sender from a "From " line in the message, but it was not doing so
6889 when it took $return_path from a Return-Path: header line. It now does.
6891 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6892 with a valid header line field name (a series of printing characters
6893 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6895 23. Changed "disc" in the source to "disk" to conform to the documentation and
6896 the book and for uniformity.
6898 24. Ignore Sendmail's -Ooption=value command line item.
6900 25. When execve() failed while trying to run a command in a pipe transport,
6901 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6902 could be confused with a return value of 69 from the command itself. This
6903 has been changed to 127, the value the shell returns if it is asked to run
6904 a non-existent command. The wording for the related log line suggests a
6905 non-existent command as the problem.
6907 26. If received_header_text expands to an empty string, do not add a Received:
6908 header line to the message. (Well, it adds a token one on the spool, but
6909 marks it "old" so that it doesn't get used or transmitted.)
6911 27. Installed eximstats 1.28 (addition of -nt option).
6913 28. There was no check for failure on the call to getsockname() in the daemon
6914 code. This can fail if there is a shortage of resources on the system, with
6915 ENOMEM, for example. A temporary error is now given on failure.
6917 29. Contrary to the C standard, it seems that in some environments, the
6918 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6919 program. Exim now does this explicitly; it affects the formatting of
6920 timestamps using strftime().
6922 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6923 complaints. I've now initialized all the variables, to avoid this.
6925 32. Header lines added by a system filter were not being "seen" during
6926 transport-time rewrites.
6928 33. The info_callback() function passed to OpenSSL is set up with type void
6929 (*)(SSL *, int, int), as described somewhere. However, when calling the
6930 function (actually a macro) that sets it up, the type void(*)() is
6931 expected. I've put in a cast to prevent warnings from picky compilers.
6933 34. If a DNS black list lookup found a CNAME record, but there were no A
6934 records associated with the domain it pointed at, Exim crashed.
6936 35. If a DNS black list lookup returned more than one A record, Exim ignored
6937 all but the first. It now scans all returned addresses if a particular IP
6938 value is being sought. In this situation, the contents of the
6939 $dnslist_value variable are a list of all the addresses, separated by a
6942 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6943 to accept a host name and all its aliases if the forward lookup for any of
6944 them yielded the IP address of the incoming connection. Now it accepts only
6945 those names whose forward lookup yields the correct IP address. Any other
6946 names are discarded. This closes a loophole whereby a rogue DNS
6947 administrator could create reverse DNS records to break through a
6948 wildcarded host restriction in an ACL.
6950 37. If a user filter or a system filter that ran in a subprocess used any of
6951 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6952 the wrong values were passed to the pipe command ($thisaddress had the
6953 value of $0, $0 had the value of $1, etc). This bug was introduced by
6954 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6956 38. Improved the line breaking for long SMTP error messages from ACLs.
6957 Previously, if there was no break point between 40 and 75 characters, Exim
6958 left the rest of the message alone. Two changes have been made: (a) I've
6959 reduced the minimum length to 35 characters; (b) if it can't find a break
6960 point between 35 and 75 characters, it looks ahead and uses the first one
6961 that it finds. This may give the occasional overlong line, but at least the
6962 remaining text gets split now.
6964 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6965 file descriptors might be low, and that setting 1000 would always raise it.
6966 It turns out that in some environments, the limit is already over 1000 and
6967 that lowering it causes trouble. So now Exim takes care not to decrease it.
6969 40. When delivering a message, the value of $return_path is set to $sender_
6970 address at the start of routing (routers may change the value). By an
6971 oversight, this default was not being set up when an address was tested by
6972 -bt or -bv, which affected the outcome if any router or filter referred to
6975 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6976 log only when "message" or "log_message" are set. However, if one of the
6977 conditions was an address verification, or a call to a nested ACL, the
6978 messages generated by the underlying test were being passed through. This
6979 no longer happens. The underlying message is available in $acl_verify_
6980 message for both "message" and "log_message" expansions, so it can be
6981 passed through if needed.
6983 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6984 new expansion $bh_ to give the encoded byte string without charset
6985 translation. Translation happens only if iconv() is available; HAVE_ICONV
6986 indicates this at build time. HEADERS_CHARSET gives the charset to
6987 translate to; headers_charset can change it in the configuration, and
6988 "headers charset" can change it in an individual filter file.
6990 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6991 that creates RFC 2047 encoded "words" labels them as that charset instead
6992 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6993 expansion operator; (ii) when Exim creates a From: line for a local
6994 message; (iii) when a header line is rewritten to include a "phrase" part.
6996 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6997 buggy, causing it to skip the first lines of messages whose message ID
6998 ended in 'D'. This would not have bitten before Exim release 4.14, because
6999 message IDs were unlikely to end in 'D' before then. The effect was to have
7000 incorrect size information for certain domains.
7002 45. #include "config.h" was missing at the start of the crypt16.c module. This
7003 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7006 46. If there was a timeout during a "random" callout check, Exim treated it as
7007 a failure of the random address, and carried on sending RSET and the real
7008 address. If the delay was just some slowness somewhere, the response to the
7009 original RCPT would be taken as a response to RSET and so on, causing
7010 mayhem of various kinds.
7012 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7013 when I implemented it. It didn't allow for the fact that some option values
7014 may legitimately be negative (e.g. size_addition), and it didn't even do
7015 the right test for positive values.
7017 48. Domain names in DNS records are case-independent. Exim always looks them up
7018 in lower case. Some resolvers return domain names in exactly the case they
7019 appear in the zone file, that is, they may contain uppercase letters. Not
7020 all resolvers do this - some return always lower case. Exim was treating a
7021 change of case by a resolver as a change of domain, similar to a widening
7022 of a domain abbreviation. This triggered its re-routing code and so it was
7023 trying to route what was effectively the same domain again. This normally
7024 caused routing to fail (because the router wouldn't handle the domain
7025 twice). Now Exim checks for this case specially, and just changes the
7026 casing of the domain that it ultimately uses when it transmits the message
7029 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7032 50. If a filter generated a file delivery with a non-absolute name (possible if
7033 no home directory exists for the router), the forbid_file option was not
7036 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7037 the existing equality matching.
7039 52. Exim was using ints instead of ino_t variables in some places where it was
7040 dealing with inode numbers.
7042 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7043 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7044 called TMPDIR, and if it finds it is different, it changes its value.
7046 54. The smtp_printf() function is now made available to local_scan() so
7047 additional output lines can be written before returning. There is also an
7048 smtp_fflush() function to enable the detection of a dropped connection.
7049 The variables smtp_input and smtp_batched_input are exported to
7052 55. Changed the default runtime configuration: the message "Unknown user"
7053 has been removed from the ACL, and instead placed on the localuser router,
7054 using the cannot_route_message feature. This means that any verification
7055 failures that generate their own messages won't get overridden. Similarly,
7056 the "Unrouteable address" message that was in the ACL for unverifiable
7057 relay addresses has also been removed.
7059 56. Added hosts_avoid_esmtp to the smtp transport.
7061 57. The exicyclog script was not checking for the esoteric option
7062 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7063 will work only if exicyclog is run under the appropriate euid.
7065 58. Following a discussion on the list, the rules by which Exim recognises line
7066 endings on incoming messages have been changed. The -dropcr and drop_cr
7067 options are now no-ops, retained only for backwards compatibility. The
7068 following line terminators are recognized: LF CRLF CR. However, special
7069 processing applies to CR:
7071 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7072 nor a local message in the state where . is a terminator.
7074 (ii) If a bare CR is encountered in a header line, an extra space is added
7075 after the line terminator so as not to end the header. The reasoning
7076 behind this is that bare CRs in header lines are most likely either
7077 to be mistakes, or people trying to play silly games.
7079 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7080 was being incorrectly given as 18 bytes larger than it should have been.
7081 This is a VOB (very old bug).
7083 60. This may never have affected anything current, but just in case it has:
7084 When the local host is found other than at the start of a list of hosts,
7085 the local host, those with the same MX, and any that follow, are discarded.
7086 When the list in question was part of a longer list of hosts, the following
7087 hosts (not currently being processed) were also being discarded. This no
7088 longer happens. I'm not sure if this situation could ever has previously
7091 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7094 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7095 header. This is something that is recommended in a new Internet Draft, and
7096 is something that is documented as being done by Sendmail. There are two
7097 possible values. For messages generated by the autoreply transport, Exim
7100 Auto-Submitted: auto-replied
7102 whereas for all other generated messages (e.g. bounces) it adds
7104 Auto-Submitted: auto-generated
7106 63. The "personal" condition in filters now includes a test for the
7107 Auto-Submitted: header. If it contains the string "auto-" the message it
7108 not considered personal.
7110 64. Added rcpt_include_affixes as a generic transport option.
7112 65. Added queue_only_override (default true).
7114 66. Added the syslog_duplication option.
7116 67. If what should have been the first header line of a message consisted of
7117 a space followed by a colon, Exim was mis-interpreting it as a header line.
7118 It isn't of course - it is syntactically invalid and should therefore be
7119 treated as the start of the message body. The misbehaviour could have
7120 caused a number of strange effects, including loss of data in subsequent
7121 header lines, and spool format errors.
7123 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7124 client host had authenticated. This control can now be exercised by an ACL
7125 for more flexibility.
7127 69. By default, callouts do not happen when testing with -bh. There is now a
7128 variant, -bhc, which does actually run the callout code, including
7129 consulting and updating the callout cache.
7131 70. Added support for saslauthd authentication, courtesy of Alexander
7134 71. If statvfs() failed on the spool or log directories while checking their
7135 size for availability, Exim confusingly gave the error "space shortage".
7136 Furthermore, in debugging mode it crashed with a floating point exception.
7137 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7138 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7139 serious problem, Exim now writes to the main and panic logs when this
7140 happens, with details of the failure. It then refuses to accept the
7141 incoming message, giving the message "spool directory problem" or "log
7142 directory problem" with a 421 code for SMTP messages.
7144 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7145 0, 1, and 2 exist, because some OS complain for execs without them (see
7146 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7147 descriptors. However, the code omitted to check that the open succeeded,
7148 causing mysterious errors if for some reason the permissions on /dev/null
7149 got screwed. Now Exim writes a message to the main and panic logs, and
7150 bombs out if it can't open /dev/null.
7152 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7153 interact so that it is all more flexible. It is supposed to remain
7154 backwards compatible. Also added extra_local_interfaces.
7156 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7157 to bomb out with an assertion failure - to the client this appears as a
7158 connection drop. This problem occurs in the part of the code that was taken
7159 from the Samba project. Fortunately, the assertion is in a very simple
7160 function, so I have fixed this by reproducing the function inline in the
7161 one place where it is called, and arranging for authentication to fail
7162 instead of killing the process with assert().
7164 75. The SPA client code was not working when the server requested OEM rather
7165 than Unicode encoding.
7167 76. Added code to make require_files with a specific uid setting more usable in
7168 the case where statting the file as root fails - usually a non-root-mounted
7169 NFS file system. When this happens and the failure is EACCES, Exim now
7170 forks a subprocess and does the per-uid checking as the relevant uid.
7172 77. Added process_log_path.
7174 78. If log_file_path was not explicitly set, a setting of check_log_space or
7175 check_log_inodes was ignored.
7177 79. If a space check for the spool or log partitions fails, the incident is now
7178 logged. Of course, in the latter case the data may get lost...
7180 80. Added the %p formatting code to string_format() so that it can be used to
7181 print addresses in debug_print(). Adjusted all the address printing in the
7182 debugging in store.c to use %p rather than %d.
7184 81. There was a concern that a line of code in smtp_in.c could overflow a
7185 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7186 initially expressed, the concern was not well-founded, because trailing
7187 spaces are removed early. However, if the trailing spaces were followed by
7188 a NULL, they did not get removed, so the overflow was possible. Two fixes
7191 (a) I re-wrote the offending code in a cleaner fashion.
7192 (b) If an incoming SMTP command contains a NULL character, it is rejected
7195 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7196 runs initgroups(), so that if the Exim user is in any additional groups,
7197 they will be used during message reception.
7203 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob