1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe functions being used.
17 JH/03 Bug 2269: When presented with a received message having a stupidly large
18 number of DKIM-Signature headers, disable DKIM verification to avoid
19 a resource-consumption attack. The limit is set at twenty.
21 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
22 report of oldest_pass in ${authres } in consequence, and separate out
23 some descriptions of reasons for verification fail.
25 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
26 files in the spool were present and unlocked. A queue-runner could spot
27 them, resulting in a duplicate delivery. Fix that by doing the unlock
28 after the unlink. Investigation by Tim Stewart. Take the opportunity to
29 add more error-checking on spoolfile handling while that code is being
32 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
33 No known attacks, no CVE, this is defensive hardening.
35 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
36 a queue-runner could start a delivery while other operations were ongoing.
37 Cutthrough delivery was a common victim, resulting in duplicate delivery.
38 Found and investigated by Tim Stewart. Fix by using the open message data
39 file handle rather than opening another, and not locally closing it (which
40 releases a lock) for that case, while creating the temporary .eml format
41 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
43 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
44 $sender_verify_failure/$recipient_verify_failure to "random".
50 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
51 When redis_servers is set to a list of > 1 element, and the Redis servers
52 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
53 case of MOVED into a DEFER case instead, thus moving the query onto the
54 next server in the list. For a cluster of N elements, all N servers must
55 be defined in redis_servers.
57 GF/02 Catch and remove uninitialized value warning in exiqsumm
58 Check for existence of @ARGV before looking at $ARGV[0]
60 JH/01 Replace the store_release() internal interface with store_newblock(),
61 which internalises the check required to safely use the old one, plus
62 the allocate and data copy operations duplicated in both (!) of the
65 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
66 modifier. This matches the restriction on the commandline.
68 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
69 Previously only the last row was returned.
71 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
72 we assumed that tags in the header were well-formed, and parsed the
73 element content after inspecting only the first char of the tag.
74 Assumptions at that stage could crash the receive process on malformed
77 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
78 While running the DKIM ACL we operate on the Permanent memory pool so that
79 variables created with "set" persist to the DATA ACL. Also (at any time)
80 DNS lookups that fail create cache records using the Permanent pool. But
81 expansions release any allocations made on the current pool - so a dnsdb
82 lookup expansion done in the DKIM ACL releases the memory used for the
83 DNS negative-cache, and bad things result. Solution is to switch to the
84 Main pool for expansions.
85 While we're in that code, add checks on the DNS cache during store_reset,
86 active in the testsuite.
87 Problem spotted, and debugging aided, by Wolfgang Breyha.
89 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
90 When none of the hosts presented to a transport match an already-open
91 connection, close it and proceed with the list. Previously we would
92 queue the message. Spotted by Lena with Yahoo, probably involving
95 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
96 Previously a spurious "250 OK id=" response was appended to the proper
99 JH/08 The "support for" informational output now, which built with Content
100 Scanning support, has a line for the malware scanner interfaces compiled
101 in. Interface can be individually included or not at build time.
103 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
104 by the template makefile "src/EDITME". The "STREAM" support for an older
105 ClamAV interface method is removed.
107 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
108 rows affected is given instead).
110 JH/11 The runtime Berkeley DB library version is now additionally output by
111 "exim -d -bV". Previously only the compile-time version was shown.
113 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
114 SMTP connection. Previously, when one had more receipients than the
115 first, an abortive onward connection was made. Move to full support for
116 multiple onward connections in sequence, handling cutthrough connection
117 for all multi-message initiating connections.
119 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
120 routers. Previously, a multi-recipient message would fail to match the
121 onward-connection opened for the first recipient, and cause its closure.
123 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
124 a timeout on read on a GnuTLS initiating connection, resulting in the
125 initiating connection being dropped. This mattered most when the callout
126 was marked defer_ok. Fix to keep the two timeout-detection methods
129 JH/15 Relax results from ACL control request to enable cutthrough, in
130 unsupported situations, from error to silently (except under debug)
131 ignoring. This covers use with PRDR, frozen messages, queue-only and
134 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
136 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
137 metadata, resulting in a crash in free().
139 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
140 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
141 Broken also in d185889f4, with init system revamp.
143 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
144 Previously we abruptly closed the connection after reading a malware-
145 found indication; now we go on to read the "scan ok" response line,
148 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
149 ACL. Previously, a crash would result.
151 JH/19 Speed up macro lookups during configuration file read, by skipping non-
152 macro text after a replacement (previously it was only once per line) and
153 by skipping builtin macros when searching for an uppercase lead character.
155 JH/20 DANE support moved from Experimental to mainline. The Makefile control
156 for the build is renamed.
158 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
159 was allocated for every new TLS startup, meaning one per message. Fix
160 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
162 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
163 reported the original. Fix to report (as far as possible) the ACL
164 result replacing the original.
166 JH/23 Fix memory leak during multi-message connections using STARTTLS under
167 OpenSSL. Certificate information is loaded for every new TLS startup,
168 and the resources needed to be freed.
170 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
172 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
175 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
176 DATA response info to the (existing) per-recipient response info for
177 the "C=" log element. It can have useful tracking info from the
178 destination system. Patch from Simon Arlott.
180 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
181 length value. Previously this would segfault.
183 HS/02 Support Avast multiline protoocol, this allows passing flags to
184 newer versions of the scanner.
186 JH/28 Ensure that variables possibly set during message acceptance are marked
187 dead before release of memory in the daemon loop. This stops complaints
188 about them when the debug_store option is enabled. Discovered specifically
189 for sender_rate_period, but applies to a whole set of variables.
190 Do the same for the queue-runner and queue-list loops, for variables set
191 from spool message files. Do the same for the SMTP per-message loop, for
192 certain variables indirectly set in ACL operations.
194 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
195 as a multi-recipient message from a mailinglist manager). The coding had
196 an arbitrary cutoff number of characters while checking for more input;
197 enforced by writing a NUL into the buffer. This corrupted long / fast
198 input. The problem was exposed more widely when more pipelineing of SMTP
199 responses was introduced, and one Exim system was feeding another.
200 The symptom is log complaints of SMTP syntax error (NUL chars) on the
201 receiving system, and refused recipients seen by the sending system
202 (propating to people being dropped from mailing lists).
203 Discovered and pinpointed by David Carter.
205 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
206 replaced by the ${authresults } expansion.
208 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
210 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
211 allows proper process termination in container environments.
213 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
214 Previously the "final dot" had a newline after it; ensure it is CR,LF.
216 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
217 and "err_perm", deprecated since 4.83 when the RFC-defined words
218 "temperror" and "permerror" were introduced.
220 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
221 transport-filters or DKIM-signing. The restriction was lost in the
222 consolidation of verify-callout and delivery SMTP handling.
223 Extend the restriction to also cover ARC-signing.
225 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
226 in defer=pass mode supply a 450 to the initiator. Previously the message
229 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
230 tls_require_ciphers is used as before.
232 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
233 "pass_unscanned". Only tmpfails from the scanner are written to
234 the paniclog, as they may require admin intervention (permission
235 denied, license issues). Other scanner errors (like decompression
236 bombs) do not cause a paniclog entry.
238 JH/36 Fix reinitialisation of DKIM logging variable between messages.
239 Previously it was possible to log spurious information in receive log
242 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
243 triggered odd behaviour from Outlook Express clients.
245 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
248 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
249 since the IETF WG has not yet settled on that versus the original
250 "bare" representation.
252 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
253 Previously the millisecond value corrupted the output.
254 Fix also for syslog_pid=no and log_selector +pid, for which the pid
255 corrupted the output.
261 JH/01 Rework error string handling in TLS interface so that the caller in
262 more cases is responsible for logging. This permits library-sourced
263 string to be attached to addresses during delivery, and collapses
264 pairs of long lines into single ones.
266 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
267 during configuration. Wildcards are allowed and expanded.
269 JH/02 Rework error string handling in DKIM to pass more info back to callers.
270 This permits better logging.
272 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
273 do not close it down and have the child transport start it up again on
274 the passed-on TCP connection. Instead, proxy the child (and any
275 subsequent ones) for TLS via a unix-domain socket channel. Logging is
276 affected: the continued delivery log lines do not have any DNSSEC, TLS
277 Certificate or OCSP information. TLS cipher information is still logged.
279 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
280 identical IP addresses on different listening ports. Will also affect
283 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
284 add noisy ifdef guards to special-case this sillyness.
285 Patch from Bernd Kuhls.
287 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
288 than 255 are no longer allowed.
290 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
291 Disable the session-cache too, which might reduce our load. Since we
292 currrectly use a new context for every connection, both as server and
293 client, there is no benefit for these.
294 GnuTLS appears to not support tickets server-side by default (we don't
295 call gnutls_session_ticket_enable_server()) but client side is enabled
296 by default on recent versions (3.1.3 +) unless the PFS priority string
299 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
300 <https://reproducible-builds.org/specs/source-date-epoch/>.
302 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
303 the check for any unsuccessful recipients did not notice the limit, and
304 erroneously found still-pending ones.
306 JH/08 Pipeline CHUNKING command and data together, on kernels that support
307 MSG_MORE. Only in-clear (not on TLS connections).
309 JH/09 Avoid using a temporary file during transport using dkim. Unless a
310 transport-filter is involved we can buffer the headers in memory for
311 creating the signature, and read the spool data file once for the
312 signature and again for transmission.
314 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
315 4.77 as the kernel support then wasn't solid, having issues in 64bit
316 mode. Now, it's been long enough. Add support for FreeBSD also.
318 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
319 case where the routing stage had gathered several addresses to send to
320 a host before calling the transport for the first, we previously failed
321 to close down TLS in the old transport process before passing the TCP
322 connection to the new process. The new one sent a STARTTLS command
323 which naturally failed, giving a failed delivery and bloating the retry
324 database. Investigation and fix prototype from Wolfgang Breyha.
326 JH/12 Fix check on SMTP command input synchronisation. Previously there were
327 false-negatives in the check that the sender had not preempted a response
328 or prompt from Exim (running as a server), due to that code's lack of
329 awareness of the SMTP input buffering.
331 PP/04 Add commandline_checks_require_admin option.
332 Exim drops privileges sanely, various checks such as -be aren't a
333 security problem, as long as you trust local users with access to their
334 own account. When invoked by services which pass untrusted data to
335 Exim, this might be an issue. Set this option in main configuration
336 AND make fixes to the calling application, such as using `--` to stop
339 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
340 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
341 responses to those, into a single TLS record each way (this usually means
342 a single packet). As a side issue, smtp_enforce_sync now works on TLS
345 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
346 affects you only if you're dancing at the edge of the param size limits.
347 If you are, and this message makes sense to you, then: raise the
348 configured limit or use OpenSSL 1.1. Nothing we can do for older
351 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
352 cmdline element to get the documented default one. Previously it was
355 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
356 are unsuable for non-message-specific reasons - eg. connection timeout,
357 banner-time rejection.
359 JH/16 Fix logging of delivery remote port, when specified by router, under
362 PP/06 Repair manualroute's ability to take options in any order, even if one
363 is the name of a transport.
366 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
368 JH/17 Change the list-building routines interface to use the expanding-string
369 triplet model, for better allocation and copying behaviour.
371 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
372 Previously it was constructed the first time a possibly-matching string
373 was met in the configuration file input during startup; now it is done
376 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
377 compatible one, to avoid the (poorly documented) possibility of a config
378 file in the working directory redirecting the DB files, possibly correpting
379 some existing file. CVE-2017-10140 assigned for BDB.
381 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
382 cache-hot. Previously, although the result was properly cached, the
383 initial verify call returned a defer.
385 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
386 the main verify for receipient in uncached-mode.
388 JH/22 Retire historical build files to an "unsupported" subdir. These are
389 defined as "ones for which we have no current evidence of testing".
391 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
392 if present. Previously it was ignored.
394 JH/24 Start using specified-initialisers in C structure init coding. This is
395 a C99 feature (it's 2017, so now considered safe).
397 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
398 if was a fixed-sized field and bitmask ops via macros; it is now more
401 PP/07 GitHub PR 56: Apply MariaDB build fix.
402 Patch provided by Jaroslav Å karvada.
404 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
405 during Coverity cleanups [4.87 JH/47]
406 Diagnosis and fix provided by Michael Fischer v. Mollard.
408 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
409 the right size to place the terminating semicolon on its own folded
410 line, the header hash was calculated to an incorrect value thanks to
411 the (relaxed) space the fold became.
413 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
414 and confused the parent.
416 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
417 which could crash as a result. This could lead to undeliverable messages.
419 JH/28 Logging: "next input sent too soon" now shows where input was truncated
422 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
423 matters on fast-turnover and PID-randomising systems, which were getting
424 out-of-order delivery.
426 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
427 a possibly-overlapping copy. The symptom was that "Remote host closed
428 connection in response to HELO" was logged instead of the actual 4xx
431 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
432 Previously only that bufferd was discarded, resulting in SYMTP command
435 JH/32 DKIM: when a message has multiple signatures matching an identity given
436 in dkim_verify_signers, run the dkim acl once for each. Previously only
437 one run was done. Bug 2189.
439 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
440 "panic the current process" to "deliberately defer". The panic log is
441 still written with the problem list name; the mail and reject logs now
442 get a temp-reject line for the message that was being handled, saying
443 something like "domains check lookup or other defer". The SMTP 451
444 message is still "Temporary local problem".
446 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
447 A crafted sequence of BDAT commands could result in in-use memory beeing
448 freed. CVE-2017-16943.
450 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
451 from SMTP input. Previously it was always done; now only done for DATA
452 and not BDAT commands. CVE-2017-16944.
454 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
455 to the message (such as an overlong header line). Previously this was
456 not done and we did not exit BDAT mode. Followon from the previous item
457 though a different problem.
463 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
464 than -2003 did; needs libidn2 in addition to libidn.
466 JH/02 The path option on a pipe transport is now expanded before use.
468 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
469 Patch provided by "Björn", documentation fix added too.
471 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
472 missing a wire-to-host endian conversion.
474 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
475 close after a BDAT command line could be taken as a following command,
476 giving a synch failure. Fix by only checking for synch immediately
477 before acknowledging the chunk.
479 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
480 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
481 macro. Patches provided by Josh Soref.
483 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
484 Previously we did not; the RFC seems ambiguous and VRFY is not listed
485 by IANA as a service extension. However, John Klensin suggests that we
488 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
489 the dkim code may be unix-mode line endings rather than smtp wire-format
490 CRLF, so prepend a CR to any bare LF.
492 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
493 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
495 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
496 the first were themselves being wrongly included in the feed into dkim
497 processing; with most chunk sizes in use this resulted in an incorrect
498 body hash calculated value.
500 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
501 DKIM signature block, for verification. Although advised against by
502 standards it is specifically not ruled illegal.
504 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
506 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
507 missing a body hash (the bh= tag).
509 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
510 It seems that HAProxy sends the Proxy Protocol information in clear and
511 only then does a TLS startup, so do the same.
513 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
514 TCP connections (such as for Spamd) unless the daemon successfully set
515 Fast Open mode on its listening sockets. This fixes breakage seen on
516 too-old kernels or those not configured for Fast Open, at the cost of
517 requiring both directions being enabled for TFO, and TFO never being used
518 by non-daemon-related Exim processes.
520 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
521 endings, at least on the first header line. Try to canonify any that get
522 past that check, despite the cost.
524 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
525 now limited to an arbitrary five deep, while parsing addresses with the
526 strip_excess_angle_brackets option enabled.
528 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
529 instead leave the unprompted TLS handshake in socket buffer for the
530 TLS library to consume.
532 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
534 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
536 JH/16 Drop variables when they go out of scope. Memory management drops a whole
537 region in one operation, for speed, and this leaves assigned pointers
538 dangling. Add checks run only under the testsuite which checks all
539 variables at a store-reset and panics on a dangling pointer; add code
540 explicitly nulling out all the variables discovered. Fixes one known
541 bug: a transport crash, where a dangling pointer for $sending_ip_address
542 originally assigned in a verify callout, is re-used.
544 PP/06 Drop '.' from @INC in various Perl scripts.
546 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
548 PP/08 Reduce a number of compilation warnings under clang; building with
549 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
550 should be warning-free.
552 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
554 HS/01 Fix portability problems introduced by PP/08 for platforms where
555 realloc(NULL) is not equivalent to malloc() [SunOS et al].
557 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
558 chunk. This allows us to accept broken chunked messages. We need a more
559 general solution here.
561 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
562 already-broken messages in the queue.
564 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
566 JH/19 Fix reference counting bug in routing-generated-address tracking.
572 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
573 supports it and a size is available (ie. the sending peer gave us one).
575 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
576 years of being deprecated). The replacements are the ACLs
577 acl_smtp_mime and acl_not_smtp_mime.
579 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
580 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
581 or even an in-clear connection were permitted. Now, if the host lookup
582 was dnssec and dane was requested then the host is only used if the
583 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
584 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
585 if one fails this test.
586 This means that a poorly-configured remote DNS will make it incommunicado;
587 but it protects against a DNS-interception attack on it.
589 JH/04 Bug 1810: make continued-use of an open smtp transport connection
590 non-noisy when a race steals the message being considered.
592 JH/05 If main configuration option tls_certificate is unset, generate a
593 self-signed certificate for inbound TLS connections.
595 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
596 in rewrites and routers.
598 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
599 and logged a warning sing 4.83; now they are a configuration file error.
601 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
602 (lacking @domain). Apply the same qualification processing as RCPT.
604 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
606 JH/10 Support ${sha256:} applied to a string (as well as the previous
609 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
610 a cutthrough deliver is pending, as we always want to make a connection.
611 This also avoids re-routing the message when later placing the cutthrough
612 connection after a verify cache hit.
613 Do not update it with the verify result either.
615 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
616 when routing results in more than one destination address.
618 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
619 signing (which inhibits the cutthrough capability). Previously only
620 the presence of an option was tested; now an expansion evaluating as
621 empty is permissible (obviously it should depend only on data available
622 when the cutthrough connection is made).
624 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
625 the relevant preceding SMTP command did not note the pipelining mode.
627 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
628 Previously they were not counted.
630 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
631 as one having no matching records. Previously we deferred the message
632 that needed the lookup.
634 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
635 distinguished as "(=".
637 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
638 for missing MX records. Previously it only worked for missing A records.
640 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
642 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
643 after the data-go-ahead and data-ack. Patch from Jason Betts.
645 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
646 even for a "none" policy. Patch from Tony Meyer.
648 JH/22 Fix continued use of a connection for further deliveries. If a port was
649 specified by a router, it must also match for the delivery to be
652 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
653 When one of the recipients of a message was unsuitable for the connection
654 (has no matching addresses), we lost track of needing to mark it
655 deferred. As a result mail would be lost.
657 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
659 JH/25 Decoding ACL controls is now done using a binary search; the source code
660 takes up less space and should be simpler to maintain. Merge the ACL
661 condition decode tables also, with similar effect.
663 JH/26 Fix problem with one_time used on a redirect router which returned the
664 parent address unchanged. A retry would see the parent address marked as
665 delivered, so not attempt the (identical) child. As a result mail would
668 JH/27 Fix a possible security hole, wherein a process operating with the Exim
669 UID can gain a root shell. Credit to http://www.halfdog.net/ for
670 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
673 JH/28 Enable {spool,log} filesystem space and inode checks as default.
674 Main config options check_{log,spool}_{inodes,space} are now
675 100 inodes, 10MB unless set otherwise in the configuration.
677 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
678 Previously it only applied to the main-section connection policy
681 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
683 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
684 by me. Added RFC7919 DH primes as an alternative.
686 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
687 are not in the system include path.
689 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
690 GnuTLS, when a session startup failed (eg because the client disconnected)
691 Exim did stdio operations after fclose. This was exposed by a recent
692 change which nulled out the file handle after the fclose.
694 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
695 signed directly by the cert-signing cert, rather than an intermediate
696 OCSP-signing cert. This is the model used by LetsEncrypt.
698 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
700 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
701 an incoming connection.
703 HS/02 Bug 1802: Do not half-close the connection after sending a request
706 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
707 fallback to "prime256v1".
709 JH/34 SECURITY: Use proper copy of DATA command in error message.
710 Could leak key material. Remotely exploitable. CVE-2016-9963.
716 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
717 and 3.4.4 - once the server is enabled to respond to an OCSP request
718 it does even when not requested, resulting in a stapling non-aware
719 client dropping the TLS connection.
721 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
722 support variable-length bit vectors. No functional change.
724 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
725 The I= interface field on outgoing lines is now after the H= remote
726 host field, same as incoming lines. There is a separate
727 outgoing_interface log selector which allows you to disable the
730 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
731 If not running log_selector +smtp_connection the mainlog would be held
732 open indefinitely after a "too many connections" event, including to a
733 deleted file after a log rotate. Leave the per net connection logging
734 leaving it open for efficiency as that will be quickly detected by the
735 check on the next write.
737 HS/01 Bug 1671: Fix post transport crash.
738 Processing the wait-<transport> messages could crash the delivery
739 process if the message IDs didn't exist for some reason. When
740 using 'split_spool_directory=yes' the construction of the spool
741 file name failed already, exposing the same netto behaviour.
743 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
744 mime_regex ACL conditions.
746 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
747 to DSN fail messages (bounces): remote IP, remote greeting, remote response
748 to HELO, local diagnostic string.
750 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
751 log line to debug. Even when configured with a tls authenticator many
752 client connections are expected to not authenticate in this way, so
753 an authenticate fail is not an error.
755 HS/02 Add the Exim version string to the process info. This way exiwhat
756 gives some more detail about the running daemon.
758 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
759 matter for fast-change records such as DNSBLs.
761 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
762 retry record, even if constant. There may be multiple transports with
763 different interface settings and the retry behaviour needs to be kept
766 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
768 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
770 JH/10 Bug 840: fix log_defer_output option of pipe transport
772 JH/11 Bug 830: use same host for all RCPTS of a message, even under
773 hosts_randomize. This matters a lot when combined with mua_wrapper.
775 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
776 ${quote_pgsql:<string>} operator.
778 JH/13 Bug 1708: avoid misaligned access in cached lookup.
780 JH/14 Change header file name for freeradius-client. Relevant if compiling
781 with Radius support; from the Gentoo tree and checked under Fedora.
783 JH/15 Bug 1712: Introduce $prdr_requested flag variable
785 JH/16 Bug 1714: Permit an empty string as expansion result for transport
786 option transport_filter, meaning no filtering.
788 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
790 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
791 defaults to "*" (all hosts). The variable is now available when not built
792 with TLS, default unset, mainly to enable keeping the testsuite sane.
793 If a server certificate is not supplied (via tls_certificate) an error is
794 logged, and clients will find TLS connections fail on startup. Presumably
795 they will retry in-clear.
796 Packagers of Exim are strongly encouraged to create a server certificate
797 at installation time.
799 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
800 with the $config_file variable.
802 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
803 in transport context, after the attempt, and per-recipient. The latter type
804 is per host attempted. The event data is the error message, and the errno
805 information encodes the lookup type (A vs. MX) used for the (first) host,
806 and the trailing two digits of the smtp 4xx response.
808 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
809 to write to mainlog (or rejectlog, paniclog) in the window between file
810 creation and permissions/ownership being changed. Particularly affects
811 installations where exicyclog is run as root, rather than exim user;
812 result is that the running daemon panics and dies.
814 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
816 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
817 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
818 "pri" and "weight". Note that the previous implicit priority given by the
819 list order is no longer honoured.
821 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
824 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
825 by defining SUPPORT_SOCKS.
827 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
828 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
829 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
830 variables are renamed to proxy_{local,external}_{address,port}.
832 JH/27 Move Internationalisation support from Experimental to mainline, enabled
833 for a build by defining SUPPORT_I18N
835 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
836 of the query string, and make ${quote_redis:} do that quoting.
838 JH/29 Move Events support from Experimental to mainline, enabled by default
839 and removable for a build by defining DISABLE_EVENT.
841 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
843 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
844 cached by the daemon.
846 JH/32 Move Redis support from Experimental to mainline, enabled for a build
847 by defining LOOKUP_REDIS. The libhiredis library is required.
849 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
850 keys are given for lookup.
852 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
853 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
854 only supported when built with TLS support. The PolarSSL SHA routines
855 are still used when the TLS library is too old for convenient support.
857 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
858 openssl_options), for security. OpenSSL forces this from version 1.1.0
859 server-side so match that on older versions.
861 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
862 allocation for $value could be released as the expansion processing
863 concluded, but leaving the global pointer active for it.
865 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
866 and to use the domains and local_parts ACL conditions.
868 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
869 incorrectly not doubled on cutthrough transmission, hence seen as a
870 body-termination at the receiving system - resulting in truncated mails.
871 Commonly the sender saw a TCP-level error, and retransmitted the message
872 via the normal store-and-forward channel. This could result in duplicates
873 received - but deduplicating mailstores were liable to retain only the
874 initial truncated version.
876 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
878 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
880 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
881 we're in there, support oversigning also; bug 1309.
883 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
885 HS/04 Add support for keep_environment and add_environment options.
887 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
888 either intentional arithmetic overflow during PRNG, or testing config-
891 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
892 delivery resulted in actual delivery. Cancel cutthrough before DATA
895 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
896 rejects a recipient: pass the reject to the originator.
898 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
899 Many were false-positives and ignorable, but it's worth fixing the
902 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
903 for the new environment-manipulation done at startup. Move the routines
904 from being local to tls.c to being global via the os.c file.
906 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
907 an extract embedded as result-arg for a map, the first arg for extract
908 is unavailable so we cannot tell if this is a numbered or keyed
909 extraction. Accept either.
915 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
918 JH/02 The smtp transport option "multi_domain" is now expanded.
920 JH/03 The smtp transport now requests PRDR by default, if the server offers
923 JH/04 Certificate name checking on server certificates, when exim is a client,
924 is now done by default. The transport option tls_verify_cert_hostnames
925 can be used to disable this per-host. The build option
926 EXPERIMENTAL_CERTNAMES is withdrawn.
928 JH/05 The value of the tls_verify_certificates smtp transport and main options
929 default to the word "system" to access the system default CA bundle.
930 For GnuTLS, only version 3.0.20 or later.
932 JH/06 Verification of the server certificate for a TLS connection is now tried
933 (but not required) by default. The verification status is now logged by
934 default, for both outbound TLS and client-certificate supplying inbound
937 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
940 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
941 Status Notification (bounce) messages are now MIME format per RFC 3464.
942 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
943 under the control of the dsn_advertise_hosts option, and routers may
944 have a dsn_lasthop option.
946 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
947 default, modifiable by a malware= option. The list separator for
948 the options can now be changed in the usual way. Bug 68.
950 JH/10 The smtp_receive_timeout main option is now expanded before use.
952 JH/11 The incoming_interface log option now also enables logging of the
953 local interface on delivery outgoing connections.
955 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
956 if the interface and destination host and port all match.
958 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
961 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
962 Patch from Andrew Lewis.
964 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
965 now supports optional time-restrictions, weighting, and priority
966 modifiers per server. Patch originally by <rommer@active.by>.
968 JH/16 The spamd_address main option now supports a mixed list of local
969 and remote servers. Remote servers can be IPv6 addresses, and
970 specify a port-range.
972 JH/17 Bug 68: The spamd_address main option now supports an optional
973 timeout value per server.
975 JH/18 Bug 1581: Router and transport options headers_add/remove can
976 now have the list separator specified.
978 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
981 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
984 JH/21 Support for the A6 type of dns record is withdrawn.
986 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
987 rather than the verbs used.
989 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
990 from 255 to 1024 chars.
992 JH/24 Verification callouts now attempt to use TLS by default.
994 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
995 are generic router options now. The defaults didn't change.
997 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
998 Original patch from Alexander Shikoff, worked over by JH.
1000 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1001 files if ZCAT_COMMAND is not executable.
1003 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1005 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1007 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1008 Normally benign, it bites when the pair was led to by a CNAME;
1009 modern usage is to not canonicalize the domain to a CNAME target
1010 (and we were inconsistent anyway for A-only vs AAAA+A).
1012 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1014 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1015 when evaluating $sender_host_dnssec.
1017 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1018 $sender_helo_dnssec variable.
1020 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1022 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1024 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1026 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1027 documented as working, but never had. Support all but $spam_report.
1029 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1030 added for tls authenticator.
1032 HS/03 Add perl_taintmode main config option
1038 TL/01 When running the test suite, the README says that variables such as
1039 no_msglog_check are global and can be placed anywhere in a specific
1040 test's script, however it was observed that placement needed to be near
1041 the beginning for it to behave that way. Changed the runtest perl
1042 script to read through the entire script once to detect and set these
1043 variables, reset to the beginning of the script, and then run through
1044 the script parsing/test process like normal.
1046 TL/02 The BSD's have an arc4random API. One of the functions to induce
1047 adding randomness was arc4random_stir(), but it has been removed in
1048 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1049 function when detected.
1051 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1052 cause callback expansion.
1054 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1055 syntax errors in an expansion can be treated as a string instead of
1056 logging or causing an error, due to the internal use of bool_lax
1057 instead of bool when processing it.
1059 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1060 server certificates when making smtp deliveries.
1062 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1064 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1066 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1068 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1069 Merged patch from Sebastian Wiedenroth.
1071 JH/05 Fix results-pipe from transport process. Several recipients, combined
1072 with certificate use, exposed issues where response data items split
1073 over buffer boundaries were not parsed properly. This eventually
1074 resulted in duplicates being sent. This issue only became common enough
1075 to notice due to the introduction of connection certificate information,
1076 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1078 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1079 size buffer was used, resulting in syntax errors when an expansion
1082 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1083 version 3.3.6 or later.
1085 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1086 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1087 both become "event_action", the variables become $event_name, $event_data
1088 and $event_defer_errno. There is a new variable $verify_mode, usable in
1089 routers, transports and related events. The tls:cert event is now also
1090 raised for inbound connections, if the main configuration event_action
1093 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1094 early OCSP support, but no stapling (appears to be less than 1.0.0).
1096 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1097 server certificate names available under the smtp transport option
1098 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1101 JH/10 Time-related extraction expansions from certificates now use the main
1102 option "timezone" setting for output formatting, and are consistent
1103 between OpenSSL and GnuTLS compilations. Bug 1541.
1105 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1106 encoded parameter in the incoming message. Bug 1558.
1108 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1109 include certificate info, eximon was claiming there were spoolfile
1112 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1114 JH/14 Log delivery-related information more consistently, using the sequence
1115 "H=<name> [<ip>]" wherever possible.
1117 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1118 are problematic for Debian distribution, omit them from the release
1121 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1123 JH/16 Fix string representation of time values on 64bit time_t architectures.
1126 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1127 output list separator was used.
1132 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1133 checkers that were complaining about end of non-void function with no
1136 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1137 This was a regression introduced in 4.83 by another bugfix.
1139 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1141 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1142 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1148 TF/01 Correctly close the server side of TLS when forking for delivery.
1150 When a message was received over SMTP with TLS, Exim failed to clear up
1151 the incoming connection properly after forking off the child process to
1152 deliver the message. In some situations the subsequent outgoing
1153 delivery connection happened to have the same fd number as the incoming
1154 connection previously had. Exim would try to use TLS and fail, logging
1155 a "Bad file descriptor" error.
1157 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1158 utilities have not been installed.
1160 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1161 temporary space as the ACL may create new global variables.
1163 TL/01 LDAP support uses per connection or global context settings, depending
1164 upon the detected version of the libraries at build time.
1166 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1167 to extract and use the src ip:port in logging and expansions as if it
1168 were a direct connection from the outside internet. PPv2 support was
1169 updated based on HAProxy spec change in May 2014.
1171 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1173 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1174 Properly escape header and check for NULL return.
1176 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1179 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1181 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1182 characters in header names, implemented as a verify condition.
1183 Contributed by Michael Fischer v. Mollard.
1185 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1186 results permerror and temperror. Previous values are deprecated but
1187 still accepted. In a future release, err_perm and err_temp will be
1188 completely removed, which will be a backward incompatibility if the
1189 ACL tests for either of these two old results. Patch contributed by
1190 user bes-internal on the mailing list.
1192 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1194 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1195 selectors, in both main and reject logs.
1197 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1200 JH/07 Add malware type "sock" for talking to simple daemon.
1202 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1204 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1205 routers/transports under cutthrough routing.
1207 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1208 numbers. Touch up "bool" conditional to keep the same definition.
1210 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1212 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1214 JH/12 Expand items in router/transport headers_add or headers_remove lists
1215 individually rather than the list as a whole. Bug 1452.
1217 Required for reasonable handling of multiple headers_ options when
1218 they may be empty; requires that headers_remove items with embedded
1219 colons must have them doubled (or the list-separator changed).
1221 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1222 view the policy declared in the DMARC record. Currently, $dmarc_status
1223 is a combined value of both the record presence and the result of the
1226 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1228 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1229 dnslookup router and the smtp transport (applying to the forward
1232 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1233 of ldap servers used for a specific lookup. Patch provided by Heiko
1236 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1237 New variable $lookup_dnssec_authenticated for observability.
1239 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1240 Patch submitted by Lars Timman.
1242 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1244 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1245 Requires trusted mode and valid format message id, aborts otherwise.
1246 Patch contributed by Heiko Schlichting.
1248 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1249 certextract with support for various fields. Bug 1358.
1251 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1252 is requested by default, modifiable by smtp transport option
1255 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1256 operate on certificate variables to give certificate fingerprints
1257 Also new ${sha256:cert_variable}.
1259 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1261 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1262 Christian Aistleitner.
1264 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1266 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1267 file. Patch from Wolfgang Breyha.
1269 JH/25 Expand the coverage of the delivery $host and $host_address to
1270 client authenticators run in verify callout. Bug 1476.
1272 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1273 align with daemon_smtp_ports. Bug 72.
1275 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1276 support and error reporting did not work properly.
1278 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1279 and is readable. Patch from Andrew Colin Kissa.
1281 TL/14 Enhance documentation of ${run expansion and how it parses the
1282 commandline after expansion, particularly in the case when an
1283 unquoted variable expansion results in an empty value.
1285 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1287 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1288 by renaming away from it. Addresses GH issue 15, reported by
1291 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1293 TL/15 SECURITY: prevent double expansion in math comparison functions
1294 (can expand unsanitized data). Not remotely exploitable.
1301 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1303 PP/02 Make -n do something, by making it not do something.
1304 When combined with -bP, the name of an option is not output.
1306 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1309 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1310 $sender_host_name and config options to manage this, and basic check
1313 PP/05 DSCP support for outbound connections and control modifier for inbound.
1315 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1316 (Only plugin which currently uses this is kerberos4, which nobody should
1317 be using, but we should make it available and other future plugins might
1318 conceivably use it, even though it would break NAT; stuff *should* be
1319 using channel bindings instead).
1321 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1322 name; added for Sendmail compatibility; requires admin caller.
1323 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1324 just ignore it); requires trusted caller.
1325 Also parse but ignore: -Ac -Am -X<logfile>
1328 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1330 TL/02 Add +smtp_confirmation as a default logging option.
1332 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1333 Patch by Magnus Holmgren from 2007-02-20.
1335 TL/04 Bugzilla 1281 - Spec typo.
1336 Bugzilla 1283 - Spec typo.
1337 Bugzilla 1290 - Spec grammar fixes.
1339 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1341 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1343 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1344 mailing list by by Dmitry Isaikin.
1346 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1348 JH/02 Support "G" suffix to numbers in ${if comparisons.
1350 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1352 NM/01 Bugzilla 1197 - Spec typo
1353 Bugzilla 1196 - Spec examples corrections
1355 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1357 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1358 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1361 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1362 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1364 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1365 "acl {{name}{arg}...}", and optional args on acl condition
1368 JH/05 Permit multiple router/transport headers_add/remove lines.
1370 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1372 JH/07 Avoid using a waiting database for a single-message-only transport.
1373 Performance patch from Paul Fisher. Bugzilla 1262.
1375 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1378 JH/09 Add $headers_added variable, with content from use of ACL modifier
1379 add_header (but not yet added to the message). Bugzilla 199.
1381 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1382 Pulled from Bugzilla 817 by Wolfgang Breyha.
1384 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1386 (nb: this is the same fix as in Exim 4.80.1)
1388 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1391 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1394 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1396 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1397 advertise SMTP AUTH mechanism to us, instead of a generic
1398 protocol violation error. Also, make Exim more robust to bad
1399 data from the Dovecot auth socket.
1401 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1403 When a queue runner is handling a message, Exim first routes the
1404 recipient addresses, during which it prunes them based on the retry
1405 hints database. After that it attempts to deliver the message to
1406 any remaining recipients. It then updates the hints database using
1409 So if a recipient address works intermittently, it can get repeatedly
1410 deferred at routing time. The retry hints record remains fresh so the
1411 address never reaches the final cutoff time.
1413 This is a fairly common occurrence when a user is bumping up against
1414 their storage quota. Exim had some logic in its local delivery code
1415 to deal with this. However it did not apply to per-recipient defers
1416 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1418 This change adds a proper retry rule check during routing so that the
1419 final cutoff time is checked against the message's age. We only do
1420 this check if there is an address retry record and there is not a
1421 domain retry record; this implies that previous attempts to handle
1422 the address had the retry_use_local_parts option turned on. We use
1423 this as an approximation for the destination being like a local
1424 delivery, as in LMTP.
1426 I suspect this new check makes the old local delivery cutoff check
1427 redundant, but I have not verified this so I left the code in place.
1429 TF/02 Correct gecos expansion when From: is a prefix of the username.
1431 Test 0254 submits a message to Exim with the header
1435 When I ran the test suite under the user fanf2, Exim expanded
1436 the header to contain my full name, whereas it should have added
1437 a Resent-Sender: header. It erroneously treats any prefix of the
1438 username as equal to the username.
1440 This change corrects that bug.
1442 GF/01 DCC debug and logging tidyup
1443 Error conditions log to paniclog rather than rejectlog.
1444 Debug lines prefixed by "DCC: " to remove any ambiguity.
1446 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1448 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1449 Bug spotted by Jeremy Harris; was flawed since initial commit.
1450 Would have resulted in OCSP responses post-SNI triggering an Exim
1451 NULL dereference and crash.
1453 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1455 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1456 Bug detection, analysis and fix by Samuel Thibault.
1457 Bugzilla 1331, Debian bug #698092.
1459 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1461 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1462 Server implementation by Todd Lyons, client by JH.
1463 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1464 config variable "prdr_enable" controls whether the server
1465 advertises the facility. If the client requests PRDR a new
1466 acl_data_smtp_prdr ACL is called once for each recipient, after
1467 the body content is received and before the acl_smtp_data ACL.
1468 The client is controlled by both of: a hosts_try_prdr option
1469 on the smtp transport, and the server advertisement.
1470 Default client logging of deliveries and rejections involving
1471 PRDR are flagged with the string "PRDR".
1473 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1474 fclose(). Diagnosis by Todd Lyons.
1476 PP/17 Update configure.default to handle IPv6 localhost better.
1477 Patch by Alain Williams (plus minor tweaks).
1480 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1481 This is now consistent with GnuTLS, and is now documented: the
1482 previous undocumented portable approach to treating the option as
1483 unset was to force an expansion failure. That still works, and
1484 an empty string is now equivalent.
1486 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1487 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1488 not performing validation itself.
1490 PP/20 Added force_command boolean option to pipe transport.
1491 Patch from Nick Koston, of cPanel Inc.
1493 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1496 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1498 PP/21 Fix eximon continuous updating with timestamped log-files.
1499 Broken in a format-string cleanup in 4.80, missed when I repaired the
1500 other false fix of the same issue.
1501 Report and fix from Heiko Schlichting.
1504 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1505 Report from Prashanth Katuri.
1507 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1508 It's SecureTransport, so affects any MacOS clients which use the
1509 system-integrated TLS libraries, including email clients.
1511 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1512 using a MIME ACL for non-SMTP local injection.
1513 Report and assistance in diagnosis by Warren Baker.
1515 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1517 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1519 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1520 last id that failed so it may be referenced in subsequent ACL's.
1522 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1525 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1526 ldap library initialization, allowing self-signed CA's to be
1527 used. Also properly sets require_cert option later in code by
1528 using NULL (global ldap config) instead of ldap handle (per
1529 session). Bug diagnosis and testing by alxgomz.
1531 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1532 the src/util/ subdirectory.
1534 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1535 renamed to Transport Post Delivery Action by Jeremy Harris, as
1538 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1539 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1540 redis_servers = needs to be configured which will be used by the redis
1541 lookup. Patch from Warren Baker, of The Packet Hub.
1543 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1545 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1546 hostname or reverse DNS when processing a host list. Used suggestions
1547 from multiple comments on this bug.
1549 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1551 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1552 Missed a few lines, added it to make the runtest require no keyboard
1555 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1556 contains upper case chars. Make router use caseful_local_part.
1558 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1559 support when GnuTLS has been built with p11-kit.
1565 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1567 This, or similar/improved, will also be change PP/11 of 4.82.
1573 PP/01 Handle short writes when writing local log-files.
1574 In practice, only affects FreeBSD (8 onwards).
1575 Bugzilla 1053, with thanks to Dmitry Isaikin.
1577 NM/01 Bugzilla 949 - Documentation tweak
1579 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1582 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1584 PP/02 Implemented gsasl authenticator.
1586 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1588 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1589 `pkg-config foo` for cflags/libs.
1591 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1592 with rest of GSASL and with heimdal_gssapi.
1594 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1595 `pkg-config foo` for cflags/libs for the TLS implementation.
1597 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1598 properties get this fed in as external SSF. A number of robustness
1599 and debugging improvements to the cyrus_sasl authenticator.
1601 PP/08 cyrus_sasl server now expands the server_realm option.
1603 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1604 Patch by Jeremy Harris.
1606 PP/10 Added dbmjz lookup type.
1608 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1610 PP/12 MAIL args handles TAB as well as SP, for better interop with
1611 non-compliant senders.
1612 Analysis and variant patch by Todd Lyons.
1614 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1615 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1616 Patch from Dirk Mueller <dmueller@suse.com>
1618 PP/13 tls_peerdn now print-escaped for spool files.
1619 Observed some $tls_peerdn in wild which contained \n, which resulted
1620 in spool file corruption.
1622 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1623 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1624 or write after TLS renegotiation, which otherwise led to messages
1627 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1628 as a tracking header (ie: a signed header comes before the signature).
1629 Patch from Wolfgang Breyha.
1631 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1632 comma-sep list; embedded commas doubled.
1634 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1636 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1638 Report and patch from Dmitry Banschikov.
1640 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1641 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1642 failures. We appear to now support TLS1.1+ with Exim.
1644 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1645 lets Exim select keys and certificates based upon TLS SNI from client.
1646 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1647 before an outbound SMTP session. New log_selector, +tls_sni.
1649 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1650 NULL dereference. Report and patch from Alun Jones.
1652 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1653 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1654 Not seeing resolver debug output on NetBSD, but suspect this is a
1655 resolver implementation change.
1657 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1658 Left warnings. Added "eximon gdb" invocation mode.
1660 PP/21 Defaulting "accept_8bitmime" to true, not false.
1662 PP/22 Added -bw for inetd wait mode support.
1664 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1665 locate the relevant includes and libraries. Made this the default.
1667 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1668 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1670 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1671 This may cause build issues on older platforms.
1673 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1674 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1675 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1676 Added SNI support via GnuTLS too.
1677 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1679 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1681 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1682 Applied second patch from Janne, implementing suggestion to default
1683 multiple-strings-in-record handling to match SPF spec.
1685 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1687 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1688 read-only, out of scope).
1689 Patch from Wolfgang Breyha, report from Stuart Northfield.
1691 PP/29 Fix three issues highlighted by clang analyser static analysis.
1692 Only crash-plausible issue would require the Cambridge-specific
1693 iplookup router and a misconfiguration.
1694 Report from Marcin Mirosław.
1696 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1698 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1699 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1700 As part of this, removing so much warning spew let me fix some minor
1701 real issues in debug logging.
1703 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1704 assignment on my part. Fixed.
1706 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1707 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1708 Janne Snabb (who went above and beyond: thank you).
1710 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1711 string otherwise requires a connection and a bunch more work and it's
1712 relatively easy to get wrong. Should also expose TLS library linkage
1715 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1716 64-bit ${eval} (JH/03).
1718 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1719 GNU libc to support some of the 64-bit stuff, should not lead to
1720 conflicts. Defined before os.h is pulled in, so if a given platform
1721 needs to override this, it can.
1723 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1724 protection layer was required, which is not implemented.
1725 Bugzilla 1254, patch from Wolfgang Breyha.
1727 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1728 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1729 tls_dhparam take prime identifiers. Also unbreak combination of
1730 OpenSSL+DH_params+TLSSNI.
1732 PP/39 Disable SSLv2 by default in OpenSSL support.
1738 PP/01 Solaris build fix for Oracle's LDAP libraries.
1739 Bugzilla 1109, patch from Stephen Usher.
1741 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1743 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1746 TF/02 Fix a couple more cases where we did not log the error message
1747 when unlink() failed. See also change 4.74-TF/03.
1749 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1750 lock up or crash if it happened to be inside a call to libc when it
1751 got a SIGUSR1 from exiwhat.
1753 The SIGUSR1 handler appends the current process status to the process
1754 log which is later printed by exiwhat. It used to use the general
1755 purpose logging code to do this, but several functions it calls are
1756 not safe for signals.
1758 The new output code in the SIGUSR1 handler is specific to the process
1759 log, and simple enough that it's easy to inspect for signal safety.
1760 Removing some special cases also simplifies the general logging code.
1761 Removing the spurious timestamps from the process log simplifies
1764 TF/04 Improved ratelimit ACL condition.
1766 The /noupdate option has been deprecated in favour of /readonly which
1767 has clearer semantics. The /leaky, /strict, and /readonly update modes
1768 are mutually exclusive. The update mode is no longer included in the
1769 database key; it just determines when the database is updated. (This
1770 means that when you upgrade Exim will forget old rate measurements.)
1772 Exim now checks that the per_* options are used with an update mode that
1773 makes sense for the current ACL. For example, when Exim is processing a
1774 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1775 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1776 must specify per_mail/readonly. If you omit the update mode it defaults to
1777 /leaky where that makes sense (as before) or /readonly where required.
1779 The /noupdate option is now undocumented but still supported for
1780 backwards compatibility. It is equivalent to /readonly except that in
1781 ACLs where /readonly is required you may specify /leaky/noupdate or
1782 /strict/noupdate which are treated the same as /readonly.
1784 A useful new feature is the /count= option. This is a generalization
1785 of the per_byte option, so that you can measure the throughput of other
1786 aggregate values. For example, the per_byte option is now equivalent
1787 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1789 The per_rcpt option has been generalized using the /count= mechanism
1790 (though it's more complicated than the per_byte equivalence). When it is
1791 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1792 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1793 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1794 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1795 Note that using per_rcpt with a non-readonly update mode in more than
1796 one ACL will cause the recipients to be double-counted. (The per_mail
1797 and per_byte options don't have this problem.)
1799 The handling of very low rates has changed slightly. If the computed rate
1800 is less than the event's count (usually one) then this event is the first
1801 after a long gap. In this case the rate is set to the same as this event's
1802 count, so that the first message of a spam run is counted properly.
1804 The major new feature is a mechanism for counting the rate of unique
1805 events. The new per_addr option counts the number of different
1806 recipients that someone has sent messages to in the last time period. It
1807 behaves like per_rcpt if all the recipient addresses are different, but
1808 duplicate recipient addresses do not increase the measured rate. Like
1809 the /count= option this is a general mechanism, so the per_addr option
1810 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1811 example, measure the rate that a client uses different sender addresses
1812 with the options per_mail/unique=$sender_address. There are further
1813 details in the main documentation.
1815 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1817 TF/06 Removed a few PCRE remnants.
1819 TF/07 Automatically extract Exim's version number from tags in the git
1820 repository when doing development or release builds.
1822 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1823 Bugzilla 879. Patch from Paul Fisher.
1825 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1826 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1829 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1831 PP/05 Variable $av_failed, true if the AV scanner deferred.
1832 Bugzilla 1078. Patch from John Horne.
1834 PP/06 Stop make process more reliably on build failure.
1835 Bugzilla 1087. Patch from Heiko Schlittermann.
1837 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1838 Bugzilla 1089. Patch from Heiko Schlittermann.
1840 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1841 Bugzilla 1131. Patch from Holger Weiß.
1843 PP/09 Handle IPv6 addresses with SPF.
1844 Bugzilla 860. Patch from Wolfgang Breyha.
1846 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1848 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1851 PP/11 match_* no longer expand right-hand-side by default.
1852 New compile-time build option, EXPAND_LISTMATCH_RHS.
1853 New expansion conditions, "inlist", "inlisti".
1855 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1857 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1859 PP/14 fix log_write() format string regression from TF/03.
1860 Bugzilla 1152. Patch from Dmitry Isaikin.
1866 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1868 PP/02 Harmonised TLS library version reporting; only show if debugging.
1869 Layout now matches that introduced for other libraries in 4.74 PP/03.
1871 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1873 PP/04 New "dns_use_edns0" global option.
1875 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1878 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1879 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1881 TK/01 Updated PolarSSL code to 0.14.2.
1882 Bugzilla 1097. Patch from Andreas Metzler.
1884 PP/07 Catch divide-by-zero in ${eval:...}.
1885 Fixes bugzilla 1102.
1887 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1890 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1891 format-string attack -- SECURITY: remote arbitrary code execution.
1893 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1894 time unintentionally subject to list matching rules, letting the header
1895 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1896 arbitrary string expansion). This allowed for information disclosure.
1898 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1899 INT_MIN/-1 -- value coerced to INT_MAX.
1905 NM/01 Workaround for PCRE version dependency in version reporting
1908 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1909 This fixes portability to compilers other than gcc, notably
1910 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1912 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1913 makefiles for portability to HP-UX and POSIX correctness.
1915 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1916 Also via indented variable definition in the Makefile.
1917 (Debugging by Oliver Heesakkers).
1919 PP/02 Restore caching of spamd results with expanded spamd_address.
1920 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1922 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1923 Improves build reliability. Fix from: Frank Elsner
1925 NM/02 Fix wide character breakage in the rfc2047 coding
1926 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1928 NM/03 Allow underscore in dnslist lookups
1929 Fixes bug 1026. Patch from Graeme Fowler
1931 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1932 Code patches from Adam Ciarcinski of NetBSD.
1934 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1937 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1938 is logged, to avoid truncation. Patch from John Horne.
1940 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1941 Patch from Jakob Hirsch.
1943 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1944 SQL string expansion failure details.
1945 Patch from Andrey Oktyabrski.
1947 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1948 Patch from Simon Arlott.
1950 PP/09 New lookups functionality failed to compile on old gcc which rejects
1951 extern declarations in function scope.
1952 Patch from Oliver Fleischmann
1954 PP/10 Use sig_atomic_t for flags set from signal handlers.
1955 Check getgroups() return and improve debugging.
1956 Fixed developed for diagnosis in bug 927 (which turned out to be
1959 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1960 Patch from Mark Zealey.
1962 PP/12 Bugzilla 1056: Improved spamd server selection.
1963 Patch from Mark Zealey.
1965 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1966 Based on patch from Heiko Schlittermann.
1968 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1969 Patch from Uwe Doering, sign-off by Michael Haardt.
1971 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1972 variable declaration deep within a block. Bug and patch from
1975 PP/15 lookups-Makefile IRIX compatibility coercion.
1977 PP/16 Make DISABLE_DKIM build knob functional.
1979 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1980 Patch by Simon Arlott
1982 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1983 variable argument macros. Our copy now differs from upstream.
1989 TF/01 Failure to get a lock on a hints database can have serious
1990 consequences so log it to the panic log.
1992 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1993 controlled using the smtp_confirmation log selector.
1995 TF/03 Include the error message when we fail to unlink a spool file.
1997 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1998 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1999 for maintaining out-of-tree patches for some time.
2001 PP/01 Bugzilla 139: Documentation and portability issues.
2002 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2003 Handle per-OS dynamic-module compilation flags.
2005 PP/02 Let /dev/null have normal permissions.
2006 The 4.73 fixes were a little too stringent and complained about the
2007 permissions on /dev/null. Exempt it from some checks.
2008 Reported by Andreas M. Kirchwitz.
2010 PP/03 Report version information for many libraries, including
2011 Exim version information for dynamically loaded libraries. Created
2012 version.h, now support a version extension string for distributors
2013 who patch heavily. Dynamic module ABI change.
2015 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2016 privilege escalation vulnerability whereby the Exim run-time user
2017 can cause root to append content of the attacker's choosing to
2020 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2023 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2024 If dropping privileges for untrusted macros, we disabled normal logging
2025 on the basis that it would fail; for the Exim run-time user, this is not
2026 the case, and it resulted in successful deliveries going unlogged.
2027 Fixed. Reported by Andreas Metzler.
2033 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2034 only prepend for the Resent-* case. Fixes regression introduced in
2035 Exim 4.70 by NM/22 for Bugzilla 607.
2037 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2038 increasing numbers of administrators be bitten by this.
2040 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2042 PP/03 Bugzilla 994: added openssl_options main configuration option.
2044 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2046 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2048 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2050 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2051 without a peer certificate, leading to a segfault because of an
2052 assumption that peers always have certificates. Be a little more
2053 paranoid. Problem reported by Martin Tscholak.
2055 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2056 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2057 NB: ClamAV planning to remove STREAM in "middle of 2010".
2058 CL also introduces -bmalware, various -d+acl logging additions and
2059 more caution in buffer sizes.
2061 PP/09 Implemented reverse_ip expansion operator.
2063 PP/10 Bugzilla 937: provide a "debug" ACL control.
2065 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2067 PP/12 Bugzilla 973: Implement --version.
2069 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2071 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2073 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2075 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2076 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2077 ignore trailing whitespace.
2079 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2081 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2084 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2085 Notification from Dr Andrew Aitchison.
2087 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2088 ExtendedDetectionInfo response format.
2089 Notification from John Horne.
2091 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2094 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2095 XSL and documented dependency on system catalogs, with examples of how
2098 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2101 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2102 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2103 configuration file which is writeable by the Exim user or group.
2105 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2106 of configuration files to cover files specified with the -C option if
2107 they are going to be used with root privileges, not just the default
2110 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2111 option (effectively making it always true).
2113 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2114 files to be used while preserving root privileges.
2116 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2117 that rogue child processes cannot use them.
2119 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2120 run-time user, instead of root.
2122 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2123 Exim run-time user without dropping privileges.
2125 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2126 result string, instead of calling string_vformat() twice with the same
2129 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2130 for other users. Others should always drop root privileges if they use
2131 -C on the command line, even for a whitelisted configure file.
2133 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2135 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2141 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2142 $data_path, and $header_path variables; fixed documentation bugs and
2145 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2146 exipick to access non-standard spools, including the "frozen" queue
2149 NM/01 Bugzilla 965: Support mysql stored procedures.
2150 Patch from Alain Williams
2152 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2154 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2155 Patch from Andreas Metzler
2157 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2158 Patch from Kirill Miazine
2160 NM/05 Bugzilla 671: Added umask to procmail example.
2162 JJ/03 installed exipick 20100323.0, fixing doc bug
2164 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2165 directory. Notification and patch from Dan Rosenberg.
2167 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2169 TK/02 Improve log output when DKIM signing operation fails.
2171 MH/01 Treat the transport option dkim_domain as a colon separated
2172 list, not as a single string, and sign the message with each element,
2173 omitting multiple occurences of the same signer.
2175 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2176 Bugzilla 985, 986. Patch by Simon Arlott
2178 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2179 Patch by Simon Arlott
2181 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2182 MBX locking. Notification from Dan Rosenberg.
2188 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2190 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2192 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2194 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2196 NM/04 Bugzilla 915: Flag broken perl installation during build.
2202 TK/01 Added patch by Johannes Berg that expands the main option
2203 "spamd_address" if it starts with a dollar sign.
2205 TK/02 Write list of recipients to X-Envelope-Sender header when building
2206 the mbox-format spool file for content scanning (suggested by Jakob
2209 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2210 (http://www.dcc-servers.net/) support via dccifd. Activated by
2211 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2213 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2214 by Mark Daniel Reidel <mr@df.eu>.
2216 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2217 When building exim an external PCRE library is now needed -
2218 PCRE is a system library on the majority of modern systems.
2219 See entry on PCRE_LIBS in EDITME file.
2221 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2222 conversation. Added nologin parameter to request.
2223 Patch contributed by Kirill Miazine.
2225 TF/01 Do not log submission mode rewrites if they do not change the address.
2227 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2229 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2230 log files in place. Contributed by Roberto Lima.
2232 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2234 TF/03 Bugzilla 615: When checking the local_parts router precondition
2235 after a local_part_suffix or local_part_prefix option, Exim now
2236 does not use the address's named list lookup cache, since this
2237 contains cached lookups for the whole local part.
2239 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2240 Robert Millan. Documentation is in experimental-spec.txt.
2242 TF/04 Bugzilla 668: Fix parallel build (make -j).
2244 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2246 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2247 Patch provided by Jan Srzednicki.
2249 TF/05 Leading white space used to be stripped from $spam_report which
2250 wrecked the formatting. Now it is preserved.
2252 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2253 that they are available at delivery time.
2255 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2257 TF/08 TLS error reporting now respects the incoming_interface and
2258 incoming_port log selectors.
2260 TF/09 Produce a more useful error message if an SMTP transport's hosts
2261 setting expands to an empty string.
2263 NM/06 Bugzilla 744: EXPN did not work under TLS.
2264 Patch provided by Phil Pennock.
2266 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2267 Patch provided by Richard Godbee.
2269 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2270 acl_smtp_notquit, added index entry.
2272 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2273 Patch provided by Eugene Bujak.
2275 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2276 accept(). Patch provided by Maxim Dounin.
2278 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2279 Patch provided by Phil Pennock.
2281 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2283 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2284 Patch provided by Brad "anomie" Jorsch.
2286 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2287 Patch provided by Dean Brooks.
2289 TK/05 Add native DKIM support (does not depend on external libraries).
2291 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2292 Patch provided by Graeme Fowler.
2294 NM/16 Bugzilla 851: Documentation example syntax fix.
2296 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2298 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2301 NM/19 Bugzilla 745: TLS version reporting.
2302 Patch provided by Phil Pennock.
2304 NM/20 Bugzilla 167: bool: condition support.
2305 Patch provided by Phil Pennock.
2307 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2308 clients. Patch provided by Phil Pennock.
2310 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2311 Patch provided by Brad "anomie" Jorsch.
2313 NM/23 Bugzilla 687: Fix misparses in eximstats.
2314 Patch provided by Heiko Schlittermann.
2316 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2317 Patch provided by Heiko Schlittermann.
2319 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2320 plus update to original patch.
2322 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2324 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2325 Patch provided by David Brownlee.
2327 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2329 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2331 NM/30 Bugzilla 888: TLS documentation bugfixes.
2333 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2335 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2336 Unlike the original bugzilla I have changed all shell scripts in src tree.
2338 NM/33 Bugzilla 898: Transport filter timeout fix.
2339 Patch by Todd Rinaldo.
2341 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2342 Patch by Serge Demonchaux.
2344 NM/35 Bugzilla 39: Base64 decode bug fixes.
2345 Patch by Jakob Hirsch.
2347 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2349 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2351 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2353 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2359 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2360 ALT-N's libdkim that I have put here:
2361 http://duncanthrax.net/exim-experimental/
2363 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2364 were called 'true' and it seems that C99 defines that as a reserved
2365 keyword to be used with 'bool' variable types. That means you could
2366 not include C99-style headers which use bools without triggering
2367 build errors in sieve.c.
2369 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2370 as mailq or other aliases. Changed the --help handling significantly
2371 to do whats expected. exim_usage() emits usage/help information.
2373 SC/01 Added the -bylocaldomain option to eximstats.
2375 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2377 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2379 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2385 PH/01 Another patch from the Sieve maintainer.
2387 PH/02 When an IPv6 address is converted to a string for single-key lookup
2388 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2389 dots are used instead of colons so that keys in lsearch files need not
2390 contain colons. This was done some time before quoting was made available
2391 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2392 (notated using the quote facility) so as to distinguish them from IPv4
2393 keys. This meant that lookups for IP addresses in host lists did not work
2394 for iplsearch lookups.
2396 This has been fixed by arranging for IPv6 addresses to be expressed with
2397 colons if the lookup type is iplsearch. This is not incompatible, because
2398 previously such lookups could never work.
2400 The situation is now rather anomalous, since one *can* have colons in
2401 ordinary lsearch keys. However, making the change in all cases is
2402 incompatible and would probably break a number of configurations.
2404 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2407 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2408 conversion specification without a maximum field width, thereby enabling
2409 a rogue spamd server to cause a buffer overflow. While nobody in their
2410 right mind would setup Exim to query an untrusted spamd server, an
2411 attacker that gains access to a server running spamd could potentially
2412 exploit this vulnerability to run arbitrary code as the Exim user.
2414 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2415 $primary_hostname instead of what libspf2 thinks the hosts name is.
2417 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2418 a directory entry by the name of the lookup key. Previously, if a
2419 symlink pointed to a non-existing file or a file in a directory that
2420 Exim lacked permissions to read, a lookup for a key matching that
2421 symlink would fail. Now it is enough that a matching directory entry
2422 exists, symlink or not. (Bugzilla 503.)
2424 PH/03 The body_linecount and body_zerocount variables are now exported in the
2427 PH/04 Added the $dnslist_matched variable.
2429 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2430 This means they are set thereafter only if the connection becomes
2433 PH/06 Added the client_condition to authenticators so that some can be skipped
2434 by clients under certain conditions.
2436 PH/07 The error message for a badly-placed control=no_multiline_responses left
2437 "_responses" off the end of the name.
2439 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2441 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2442 (without spaces) instead of just copying the configuration text.
2444 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2446 PH/11 Added $max_received_linelength.
2448 PH/12 Added +ignore_defer and +include_defer to host lists.
2450 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2451 way in which PCRE > 7.0 is built.
2453 PH/14 Implemented queue_only_load_latch.
2455 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2456 MAIL command. The effect was to mangle the value on 64-bit systems.
2458 PH/16 Another patch from the Sieve maintainer.
2460 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2462 PH/18 If a system quota error occurred while trying to create the file for
2463 a maildir delivery, the message "Mailbox is full" was not appended to the
2464 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2465 only to a quota excession during the actual writing of the file.
2467 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2468 characters?) which causes problems in log lines. The DN values are now
2469 passed through string_printing() before being added to log lines.
2471 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2472 and InterBase are left for another time.)
2474 PH/21 Added message_body_newlines option.
2476 PH/22 Guard against possible overflow in moan_check_errorcopy().
2478 PH/23 POSIX allows open() to be a macro; guard against that.
2480 PH/24 If the recipient of an error message contained an @ in the local part
2481 (suitably quoted, of course), incorrect values were put in $domain and
2482 $local_part during the evaluation of errors_copy.
2488 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2489 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2492 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2493 issue a MAIL command.
2495 PH/02 In an ACL statement such as
2497 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2499 if a client was not listed at all, or was listed with a value other than
2500 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2501 the condition was not true (as it should be), so access was not denied.
2502 The bug was that the ! inversion was incorrectly passed on to the second
2503 item. This has been fixed.
2505 PH/03 Added additional dnslists conditions == and =& which are different from
2506 = and & when the dns lookup returns more than one IP address.
2508 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2509 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2511 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2512 FSYNC, which compiles an option called disable_fsync that allows for
2513 bypassing fsync(). The documentation is heavily laced with warnings.
2515 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2517 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2518 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2519 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2520 including adding "make clean"; (3) Added -fPIC when compiling the test
2521 dynamically loaded module, to get rid of a warning.
2523 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2524 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2525 = 0s. The bug is otherwise harmless.
2527 PH/07 There was a bug in the dovecot authenticator such that the value of
2528 $auth1 could be overwritten, and so not correctly preserved, after a
2529 successful authentication. This usually meant that the value preserved by
2530 the server_setid option was incorrect.
2532 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2534 PH/09 Installed PCRE release 7.0.
2536 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2537 run for batched SMTP input. It is now run at the start of every message
2538 in the batch. While fixing this I discovered that the process information
2539 (output by running exiwhat) was not always getting set for -bs and -bS
2540 input. This is fixed, and it now also says "batched" for BSMTP.
2542 PH/11 Added control=no_pipelining.
2544 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2545 patch, slightly modified), and move the expansion of helo_data till after
2546 the connection is made in the smtp transport (so it can use these
2549 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2551 PH/14 Added log_selector = +pid.
2553 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2555 PH/16 Add ${if forany and ${if forall.
2557 PH/17 Added dsn_from option to vary the From: line in DSNs.
2559 PH/18 Flush SMTP output before performing a callout, unless control =
2560 no_callout_flush is set.
2562 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2563 was true (the default) a successful delivery failed to delete the retry
2564 item, thus causing premature timeout of the address. The bug is now
2567 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2569 PH/21 Long custom messages for fakedefer and fakereject are now split up
2570 into multiline responses in the same way that messages for "deny" and
2571 other ACL rejections are.
2573 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2574 with slight modification.
2576 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2577 draft, changing the syntax and factoring some duplicate code".
2579 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2580 for deliveries of the second and subsequent messages over the same SMTP
2583 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2584 ${reduce, with only minor "tidies".
2586 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2588 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2589 expansion side effects.
2591 PH/27 When a message times out after an over-quota error from an Exim-imposed
2592 quota, the bounce message says "mailbox is full". This message was not
2593 being given when it was a system quota that was exceeded. It now should
2596 MH/03 Made $recipients available in local_scan(). local_scan() already has
2597 better access to the recipient list through recipients_list[], but
2598 $recipients can be useful in postmaster-provided expansion strings.
2600 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2601 in the case of a MAIL command with additional options following the
2602 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2603 were accidentally chopped off.
2605 PH/29 SMTP synchronization checks are implemented when a command is read -
2606 there is a check that no more input is waiting when there shouldn't be
2607 any. However, for some commands, a delay in an ACL can mean that it is
2608 some time before the response is written. In this time, more input might
2609 arrive, invalidly. So now there are extra checks after an ACL has run for
2610 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2611 pipelining has not been advertised.
2613 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2615 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2616 returned in the SMTP response when smtp_return_error_details was false.
2617 This has been fixed.
2619 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2620 instead of the C I/O that was originally supplied, because problems were
2621 reported on Solaris.
2623 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2624 Exim which did not show up earlier: it was assuming that a call to
2625 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2626 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2627 was a macro that became an assignment, so it seemed to work. This has
2628 changed to a proper function call with a void return, hence the compile
2629 error. Exim's code has been fixed.
2631 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2634 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2636 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2638 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2639 because it turns out to be too big or there is a timeout) there is no
2640 "Completed" line in the log. When some messages of this type were
2641 selected by exigrep, they were listed as "not completed". Others were
2642 picked up by some special patterns. I have improved the selection
2643 criteria to be more general.
2645 PH/38 The host_find_failed option in the manualroute router can now be set
2646 to "ignore", to completely ignore a host whose IP address cannot be
2647 found. If all hosts are ignored, the behaviour is controlled by the new
2648 host_all_ignored option.
2650 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2651 homing or because of multiple MX records with /mx) generated more than
2652 one IP address, and the following item turned out to be the local host,
2653 all the secondary addresses of the first item were incorrectly removed
2654 from the list, along with the local host and any following hosts (which
2655 is what is supposed to happen).
2657 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2658 whoever called Exim into the -H file. In the case of the daemon it was
2659 behaving confusingly. When first started, it used values for whoever
2660 started the daemon, but after a SIGHUP it used the Exim user (because it
2661 calls itself on a restart). I have changed the code so that it now always
2664 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2665 message are rejected with the same error (e.g. no authentication or bad
2666 sender address), and a DATA command is nevertheless sent (as can happen
2667 with PIPELINING or a stupid MUA), the error message that was given to the
2668 RCPT commands is included in the rejection of the DATA command. This is
2669 intended to be helpful for MUAs that show only the final error to their
2672 PH/42 Another patch from the Sieve maintainer.
2674 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2675 Eximstats - Fixed some broken HTML links and added missing column headers
2677 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2678 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2680 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2682 PH/43 Yet another patch from the Sieve maintainer.
2684 PH/44 I found a way to check for a TCP/IP connection going away before sending
2685 the response to the final '.' that terminates a message, but only in the
2686 case where the client has not sent further data following the '.'
2687 (unfortunately, this is allowed). However, in many cases there won't be
2688 any further data because there won't be any more messages to send. A call
2689 to select() can be used: if it shows that the input is "ready", there is
2690 either input waiting, or the socket has been closed. An attempt to read
2691 the next input character can distinguish the two cases. Previously, Exim
2692 would have sent an OK response which the client would never have see.
2693 This could lead to message repetition. This fix should cure that, at
2694 least in a lot of common cases.
2696 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2697 advertised in response to EHLO.
2703 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2704 fixed by 4.65/MH/01 (is this a record?) are fixed:
2706 (i) An empty string was always treated as zero by the numeric comparison
2707 operators. This behaviour has been restored.
2709 (ii) It is documented that the numeric comparison operators always treat
2710 their arguments as decimal numbers. This was broken in that numbers
2711 starting with 0 were being interpreted as octal.
2713 While fixing these problems I realized that there was another issue that
2714 hadn't been noticed. Values of message_size_limit (both the global option
2715 and the transport option) were treated as octal if they started with 0.
2716 The documentation was vague. These values are now always treated as
2717 decimal, and I will make that clear in the documentation.
2723 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2724 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2727 MH/01 Don't check that the operands of numeric comparison operators are
2728 integers when their expansion is in "skipping" mode (fixes bug
2729 introduced by 4.64-PH/07).
2731 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2732 child addresses, Exim now panics and dies. Previously, because the count
2733 is held in a short int, deliveries were likely to be lost. As such a
2734 large number of recipients for a single message is ridiculous
2735 (performance will be very, very poor), I have chosen to impose a limit
2736 rather than extend the field.
2742 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2743 leftover -K file (the existence of which was triggered by #402).
2744 While we were at it, introduced process PID as part of the -K
2745 filename. This should rule out race conditions when creating
2748 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2749 processing considerably. Previous code took too long for large mails,
2750 triggering a timeout which in turn triggers #401.
2752 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2753 in the DK code in transports.c. sendfile() is not really portable,
2754 hence the _LINUX specificness.
2756 TF/01 In the add_headers option to the mail command in an Exim filter,
2757 there was a bug that Exim would claim a syntax error in any
2758 header after the first one which had an odd number of characters
2761 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2762 callout verification, Exim cached a "reject" for the entire domain. This
2763 is correct for most verifications, but it is not correct for a recipient
2764 verification with use_sender or use_postmaster set, because in that case
2765 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2766 case of MAIL FROM:<> rejection from other early rejections (e.g.
2767 rejection of HELO). When verifying a recipient using a non-null MAIL
2768 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2769 Whatever the result of the callout, the value of the domain cache is
2770 left unchanged (for any other kind of callout, getting as far as trying
2771 RCPT means that the domain itself is ok).
2773 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2776 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2777 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2778 session, the daemon ignores ECONNECT errors and logs others; it now
2779 ignores EPIPE as well.
2781 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2782 (quoted-printable decoding).
2784 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2785 later the small subsequent patch to fix an introduced bug.
2787 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2789 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2791 PH/08 An error is now given if message_size_limit is specified negative.
2793 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2794 to be given (somewhat) arbitrary names.
2796 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2799 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2800 miscellaneous code fixes
2802 PH/10 Added the log_reject_target ACL modifier to specify where to log
2805 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2806 hostname. This is wrong, because it relates to the incoming message (and
2807 probably the interface on which it is arriving) and not to the outgoing
2808 callout (which could be using a different interface). This has been
2809 changed to use the value of the helo_data option from the smtp transport
2810 instead - this is what is used when a message is actually being sent. If
2811 there is no remote transport (possible with a router that sets up host
2812 addresses), $smtp_active_hostname is used.
2814 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2815 tweaks were necessary in order to get it to work (see also 21 below):
2816 (a) The code assumed that strncpy() returns a negative number on buffer
2817 overflow, which isn't the case. Replaced with Exim's string_format()
2819 (b) There were several signed/unsigned issues. I just did the minimum
2820 hacking in of casts. There is scope for a larger refactoring.
2821 (c) The code used strcasecmp() which is not a standard C function.
2822 Replaced with Exim's strcmpic() function.
2823 (d) The code set only $1; it now sets $auth1 as well.
2824 (e) A simple test gave the error "authentication client didn't specify
2825 service in request". It would seem that Dovecot has changed its
2826 interface. Fortunately there's a specification; I followed it and
2827 changed what the client sends and it appears to be working now.
2829 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2832 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2833 address A is aliased to B and C, where B exists and C does not. Without
2834 -v the output is "A verified" because verification stops after a
2835 successful redirection if more than one address is generated. However,
2836 with -v the child addresses are also verified. Exim was outputting "A
2837 failed to verify" and then showing the successful verification for C,
2838 with its parentage. It now outputs "B failed to verify", showing B's
2839 parentage before showing the successful verification of C.
2841 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2842 look up a TXT record in a specific list after matching in a combined
2845 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2846 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2847 they consult the DNS. I had assumed they would set it the way they
2848 wanted; and indeed my experiments on Linux seem to show that in some
2849 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2850 To be on the safe side, however, I have now made the interface to
2851 host_find_byname() similar to host_find_bydns(), with an argument
2852 containing the DNS resolver options. The host_find_byname() function now
2853 sets these options at its start, just as host_find_bydns() does. The smtp
2854 transport options dns_qualify_single and dns_search_parents are passed to
2855 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2856 of host_find_byname() use the default settings of RES_DEFNAMES
2857 (qualify_single) but not RES_DNSRCH (search_parents).
2859 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2860 spool_read_header() do less string testing, by means of a preliminary
2861 switch on the second character of optional "-foo" lines. (This is
2862 overdue, caused by the large number of possibilities that now exist.
2863 Originally there were few.) While I was there, I also converted the
2864 str(n)cmp tests so they don't re-test the leading "-" and the first
2865 character, in the hope this might squeeze out yet more improvement.
2867 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2868 flag allowing group syntax was set by the header_syntax check but not
2869 turned off, possible causing trouble later; (2) The flag was not being
2870 set at all for the header_verify test, causing "group"-style headers to
2871 be rejected. I have now set it in this case, and also caused header_
2872 verify to ignore an empty address taken from a group. While doing this, I
2873 came across some other cases where the code for allowing group syntax
2874 while scanning a header line wasn't quite right (mostly, not resetting
2875 the flag correctly in the right place). These bugs could have caused
2876 trouble for malformed header lines. I hope it is now all correct.
2878 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2879 with the "reply" argument non-NULL. The code, however (which originally
2880 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2881 but it didn't always do it. This confused somebody who was copying the
2882 code for some other use. I have removed all the tests.
2884 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2885 feature that was used to support insecure browsers during the U.S. crypto
2886 embargo. It requires special client support, and Exim is probably the
2887 only MTA that supported it -- and would never use it because real RSA is
2888 always available. This code has been removed, because it had the bad
2889 effect of slowing Exim down by computing (never used) parameters for the
2890 RSA_EXPORT functionality.
2892 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2893 authenticator to fail if there's a tab character in the incoming data
2894 (there should never be unless someone is messing about, as it's supposed
2895 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2896 if the connection is using TLS or if the remote IP is the same as the
2897 local IP, and the "valid-client-cert option" if a client certificate has
2900 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2901 authenticators. This can be used for authorization after authentication
2902 succeeds. (In the case of plaintext, it servers for both authentication
2905 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2906 if any retry times were supplied.
2908 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2909 connection, where there is no client IP address to check. In this
2910 situation, the verify now always succeeds.
2912 PH/25 Applied John Jetmore's -Mset patch.
2914 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2916 PH/27 In a string expansion for a processed (not raw) header when multiple
2917 headers of the same name were present, leading whitespace was being
2918 removed from all of them, but trailing whitespace was being removed only
2919 from the last one. Now trailing whitespace is removed from each header
2920 before concatenation. Completely empty headers in a concatenation (as
2921 before) are ignored.
2923 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2924 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2926 PH/29 [Removed. This was a change that I later backed out, and forgot to
2927 correct the ChangeLog entry (that I had efficiently created) before
2928 committing the later change.]
2930 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2931 address errors (4xx response to RCPT) over the same connection as other
2932 messages routed to the same hosts. Such deliveries are always "forced",
2933 so retry times are not inspected. This resulted in far too many retries
2934 for the affected addresses. The effect occurred only when there were more
2935 hosts than the hosts_max_try setting in the smtp transport when it had
2936 the 4xx errors. Those hosts that it had tried were not added to the list
2937 of hosts for which the message was waiting, so if all were tried, there
2938 was no problem. Two fixes have been applied:
2940 (i) If there are any address or message errors in an SMTP delivery, none
2941 of the hosts (tried or untried) are now added to the list of hosts
2942 for which the message is waiting, so the message should not be a
2943 candidate for sending over the same connection that was used for a
2944 successful delivery of some other message. This seems entirely
2945 reasonable: after all the message is NOT "waiting for some host".
2946 This is so "obvious" that I'm not sure why it wasn't done
2947 previously. Hope I haven't missed anything, but it can't do any
2948 harm, as the worst effect is to miss an optimization.
2950 (ii) If, despite (i), such a delivery is accidentally attempted, the
2951 routing retry time is respected, so at least it doesn't keep
2952 hammering the server.
2954 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2955 in ${readsocket because some servers need this prod.
2957 PH/32 Added some extra debug output when updating a wait-xxx database.
2959 PH/33 The hint "could be header name not terminated by colon", which has been
2960 given for certain expansion errors for a long time, was not being given
2961 for the ${if def:h_colon_omitted{... case.
2963 PH/34 The spec says: "With one important exception, whenever a domain list is
2964 being scanned, $domain contains the subject domain." There was at least
2965 one case where this was not true.
2967 PH/35 The error "getsockname() failed: connection reset by peer" was being
2968 written to the panic log as well as the main log, but it isn't really
2969 panic-worthy as it just means the connection died rather early on. I have
2970 removed the panic log writing for the ECONNRESET error when getsockname()
2973 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2974 runs only) independently of the message's sender address. This meant
2975 that, if the 4xx error was in fact related to the sender, a different
2976 message to the same recipient with a different sender could confuse
2977 things. In particular, this can happen when sending to a greylisting
2978 server, but other circumstances could also provoke similar problems.
2979 I have changed the default so that the retry time for these errors is now
2980 based a combination of the sender and recipient addresses. This change
2981 can be overridden by setting address_retry_include_sender=false in the
2984 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2985 remote server are returned as part of bounce messages. This was not
2986 happening for LMTP over a pipe (the lmtp transport), but now it is the
2987 same for both kinds of LMTP.
2989 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2990 in header lines that were in fact CNAMEs. This is no longer the case.
2992 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2993 and queue runs started by the daemon processed all messages. This has
2994 been fixed so that -R and -S can now usefully be given with -q<time>.
2996 PH/40 Import PCRE release 6.7 (fixes some bugs).
2998 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3000 PH/42 Give an error if -q is specified more than once.
3002 PH/43 Renamed the variables $interface_address and $interface_port as
3003 $received_ip_address and $received_port, to make it clear that these
3004 values apply to message reception, and not to the outgoing interface when
3005 a message is delivered. (The old names remain recognized, of course.)
3007 PH/44 There was no timeout on the connect() call when using a Unix domain
3008 socket in the ${readsocket expansion. There now is.
3010 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3011 be meaningful with "accept".
3013 SC/01 Eximstats V1.43
3014 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3016 SC/02 Eximstats V1.44
3017 Use a glob alias rather than an array ref in the generated
3018 parser. This improves both readability and performance.
3020 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3021 Collect SpamAssassin and rejection statistics.
3022 Don't display local sender or destination tables unless
3023 there is data to show.
3024 Added average volumes into the top table text output.
3026 SC/04 Eximstats V1.46
3027 Collect data on the number of addresses (recipients)
3028 as well as the number of messages.
3030 SC/05 Eximstats V1.47
3031 Added 'Message too big' to the list of mail rejection
3032 reasons (thanks to Marco Gaiarin).
3034 SC/06 Eximstats V1.48
3035 Mainlog lines which have GMT offsets and are too short to
3036 have a flag are now skipped.
3038 SC/07 Eximstats V1.49 (Alain Williams)
3039 Added the -emptyok flag.
3041 SC/08 Eximstats V1.50
3042 Fixes for obtaining the IP address from reject messages.
3044 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3045 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3046 whitespace changes from 4.64-PH/27
3048 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3051 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3052 are found, allow negative numbers in numeric criteria)
3054 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3056 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3057 to match changes made in 4.64-PH/43
3059 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3061 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3062 "domain" is actually an IP address, and force a failure. This locks out
3063 those revolvers/nameservers that support "A-for-A" lookups, in
3064 contravention of the specifications.
3066 PH/48 When a host name was looked up from an IP address, and the subsequent
3067 forward lookup of the name timed out, the host name was left in
3068 $sender_host_name, contrary to the specification.
3070 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3071 restricted to single-key lookups, Exim was not diagnosing an error if
3072 * or *@ was used with a query-style lookup.
3074 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3076 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3077 long ago, but noone interested enough thought of it. Let's just say that
3078 the "1.1" means that there are some new functions that weren't there at
3079 some point in the past.
3081 PH/51 Error processing for expansion failure of helo_data from an smtp
3082 transport during callout processing was broken.
3084 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3085 tested/used via the -bh/-bhc/-bs options.
3087 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3088 bug, fixed in subsequent PCRE releases).
3090 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3091 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3093 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3099 SC/01 Use a glob alias rather than an array ref in eximstats generated
3100 parser. This improves both readability and performance.
3102 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3103 Don't display local sender or destination tables in eximstats unless
3104 there is data to show.
3105 Added average volumes into the eximstats top table text output.
3107 SC/03 Collect data on the number of addresses (recipients) as well
3108 as the number of messages in eximstats.
3110 TF/01 Correct an error in the documentation for the redirect router. Exim
3111 does (usually) call initgroups() when daemonizing.
3113 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3114 with consistent privilege compared to when running as a daemon.
3116 TF/03 Note in the spec that $authenticated_id is not set for local
3117 submissions from trusted users.
3119 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3120 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3122 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3123 by adding some example configuration directives to the default
3124 configuration file. A little bit of work is required to uncomment the
3125 directives and define how usernames and passwords are checked, but
3126 there is now a framework to start from.
3128 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3129 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3130 without this. I don't know how relevant this is to other LDAP libraries.
3132 PH/02 Add the verb name to the "unknown ACL verb" error.
3134 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3136 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3138 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3139 directory not expanded when it should be if an expanded home directory
3140 was set for the address (which is overridden by the transport).
3142 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3145 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3146 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3147 because it is too late at that time, and has no effect.
3149 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3150 security issue with \' (bugzilla #107). I could not use the
3151 PQescapeStringConn() function, because it needs a PGconn value as one of
3154 PH/08 When testing addresses using -bt, indicate those final addresses that
3155 are duplicates that would not cause an additional delivery. At least one
3156 person was confused, thinking that -bt output corresponded to deliveries.
3157 (Suppressing duplicates isn't a good idea as you lose the information
3158 about possibly different redirections that led to the duplicates.)
3160 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3161 systems where poll() doesn't work, in particular OS X.
3163 PH/10 Added more information to debugging output for retry time not reached.
3165 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3166 operations in malware.c.
3168 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3171 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3172 syslog_duplication set false, log lines that would normally be written
3173 both the the main log and to the reject log were not written to syslog at
3176 PH/14 In the default configuration, change the use of "message" in ACL warn
3177 statements to "add_header".
3179 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3180 not followed by a command (e.g. "seen endif").
3182 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3183 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3186 PH/17 Added extra conditions to the default value of delay_warning_condition
3190 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3191 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3192 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3195 The Auto-Submitted: and various List- headers are standardised, whereas I
3196 don't think Precedence: ever was.
3198 PH/18 Refactored debugging code in route_finduser() to show more information,
3199 in particular, the error code if getpwnam() issues one.
3201 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3202 This is apparently needed in addition to the PH/07 change above to avoid
3203 any possible encoding problems.
3205 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3206 but not after initializing Perl.
3208 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3209 output them only if debugging. By default they are written stderr,
3210 apparently, which is not desirable.
3212 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3215 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3218 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3220 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3221 authenticated or an ident call has been made. Suppress the default
3222 values for $authenticated_id and $authenticated_sender (but permit -oMai
3223 and -oMas) when testing with -bh.
3225 PH/24 Re-jigged the order of the tests in the default configuration so that the
3226 tests for valid domains and recipients precede the DNS black list and CSA
3227 tests, on the grounds that those ones are more expensive.
3229 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3230 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3231 This bug exists in every version of Exim that I still have, right back to
3234 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3235 However, an attempt to turn on case-sensitivity in a regex key by
3236 including (?-i) didn't work because the subject string was already
3237 lowercased, and the effects were non-intuitive. It turns out that a
3238 one-line patch can be used to allow (?-i) to work as expected.
3244 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3245 other effects) broke the use of negated acl sub-conditions.
3247 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3250 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3251 "Deny" causes Exim to reject the incoming connection with a 554 error.
3252 Unfortunately, if there is a major crisis, such as a disk failure,
3253 tcp-wrappers gives "deny", whereas what one would like would be some
3254 kind of temporary error. A kludge has been added to help with this.
3255 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3256 554 error is used if errno is still zero or contains ENOENT (which occurs
3257 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3260 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3262 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3263 errors. Otherwise a message that provokes a temporary error (when other
3264 messages do not) can cause a whole host to time out.
3266 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3267 addresses were routed directly to files or pipes from a redirect router.
3268 File deliveries just didn't batch; pipe deliveries might have suffered
3271 PH/06 A failure to get a lock for a hints database would erroneously always say
3272 "Failed to get write lock", even when it was really a read lock.
3274 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3275 of 0600. This has been changed to use the value of the lockfile_mode
3276 option (which defaults to 0600).
3278 PH/08 Applied small patch from the Sieve maintainer.
3280 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3281 folder from quota calculations, a direct delivery into this folder messed
3282 up the contents of the maildirsize file. This was because the regex was
3283 used only to exclude .Trash (or whatever) when the size of the mailbox
3284 was calculated. There was no check that a delivery was happening into an
3285 excluded directory. This bug has been fixed by ignoring all quota
3286 processing for deliveries into excluded directories.
3288 PH/10 Added the maildirfolder_create_regex option to appendfile.
3294 PH/01 The code for finding all the local interface addresses on a FreeBSD
3295 system running IPv6 was broken. This may well have applied to all BSD
3296 systems, as well as to others that have similar system calls. The broken
3297 code found IPv4 interfaces correctly, but gave incorrect values for the
3298 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3299 that it would not match correctly against @[] and not recognize the IPv6
3302 PH/02 The ipliteral router was not recognizing addresses of the form user@
3303 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3305 PH/03 Added disable_ipv6.
3307 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3308 decoded form, because it is most often used to construct To: headers
3309 lines in autoreplies, and the decoded form may well be syntactically
3310 invalid. However, $reply_address has leading white space removed, and all
3311 newlines turned into spaces so that the autoreply transport does not
3314 PH/05 If group was specified without a user on a router, and no group or user
3315 was specified on a transport, the group from the router was ignored.
3317 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3318 for visible compile-time settings that can be used to change these
3319 numbers, for those that want even more. Backwards compatibility with old
3320 spool files has been maintained. However, going back to a previous Exim
3321 release will lost any variables that are in spool files.
3323 PH/07 Two small changes when running in the test harness: increase delay when
3324 passing a TCP/IP connection to a new process, in case the original
3325 process has to generate a bounce, and remove special handling of
3326 127.0.0.2 (sic), which is no longer necessary.
3328 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3329 be the same on different OS.
3331 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3334 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3335 whether --show-vars was specified or not
3337 JJ/02 exipick: Added support for new ACL variable spool format introduced
3340 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3341 syntactically invalid From: or Reply-to: line, and a filter used this to
3342 generate an autoreply, and therefore failed to obtain an address for the
3343 autoreply, Exim could try to deliver to a non-existent relative file
3344 name, causing unrelated and misleading errors. What now happens is that
3345 it logs this as a hard delivery error, but does not attempt to create a
3348 PH/11 The exinext utility has a -C option for testing purposes, but although
3349 the given file was scanned by exinext itself; it wasn't being passed on
3350 when Exim was called.
3352 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3353 an end-of-file indication when reading a command response.
3355 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3356 compiled. In many other places in Exim, IPv6 addresses are always
3357 recognized, so I have changed this. It also means that IPv4 domain
3358 literals of the form [IPV4:n.n.n.n] are now always recognized.
3360 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3361 used if the router is not running as root, for example, when verifying at
3362 ACL time, or when using -bh. The debugging output from this situation was
3363 non-existent - all you got was a failure to exec. I have made two
3366 (a) Failures to set uid/gid, the current directory, or a process leader
3367 in a subprocess such as that created by queryprogram now generate
3368 suitable debugging output when -d is set.
3370 (b) The queryprogram router detects when it is not running as root,
3371 outputs suitable debugging information if -d is set, and then runs
3372 the subprocess without attempting to change uid/gid.
3374 PH/15 Minor change to Makefile for building test_host (undocumented testing
3377 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3378 additional section of a DNS packet that returns MX or SRV records.
3379 Instead, it always explicitly searches for A/AAAA records. This avoids
3380 major problems that occur when a DNS server includes only records of one
3381 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3382 fixed another bug: if SRV records were looked up and the corresponding
3383 address records were *not* found in the additional section, the port
3384 values from the SRV records were lost.
3386 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3387 using the correct key (the original address) when searching the retry
3388 rules in order to find which one to use for generating the retry hint.
3390 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3391 adding the default one. Similarly, if it contains a Reply-To: header, the
3392 errors_reply_to option, if set, is not used.
3394 PH/19 When calculating a retry time, Exim used to measure the "time since
3395 failure" by looking at the "first failed" field in the retry record. Now
3396 it does not use this if it is later than than the arrival time of the
3397 message. Instead it uses the arrival time. This makes for better
3398 behaviour in cases where some deliveries succeed, thus re-setting the
3399 "first failed" field. An example is a quota failure for a huge message
3400 when small messages continue to be delivered. Without this change, the
3401 "time since failure" will always be short, possible causing more frequent
3402 delivery attempts for the huge message than are intended.
3403 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3405 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3406 $1, $2, $3) because the numerical variables can be reset during some
3407 expansion items (e.g. "match"), thereby losing the authentication data.
3409 PH/21 Make -bV show the size of off_t variables so that the test suite can
3410 decide whether to run tests for quotas > 2G.
3412 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3413 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3414 filecount value is greater than 2G or if a quota value is greater than 2G
3415 on a system where the size of off_t is not greater than 4, a panic error
3418 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3419 never match. The debug and -bh output now contains an explicit error
3420 message indicating a malformed IPv4 address or mask.
3422 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3423 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3424 PH/23 above applies.
3426 PH/25 Do not write to syslog when running in the test harness. The only
3427 occasion when this arises is a failure to open the main or panic logs
3428 (for which there is an explicit test).
3430 PH/26 Added the /no_tell option to "control=freeze".
3432 PH/27 If a host name lookup failed very early in a connection, for example, if
3433 the IP address matched host_lookup and the reverse lookup yielded a name
3434 that did not have a forward lookup, an error message of the form "no IP
3435 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3436 could be logged. Now it outputs the IP address instead of "NULL".
3438 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3439 allows the sender and the authenticated sender to be set when
3440 submitting a message from within Exim. Since child_open_exim() is
3441 documented for local_scan(), the new function should be too.
3443 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3444 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3445 results in an empty string is now treated as unset.
3447 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3449 PH/31 Added sender_verify_fail logging option.
3451 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3452 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3453 tidied the source and removed it altogether.
3455 PH/33 When a queue run was abandoned because the load average was too high, a
3456 log line was always written; now it is written only if the queue_run log
3457 selector is set. In addition, the log line for abandonment now contains
3458 information about the queue run such as the pid. This is always present
3459 in "start" and "stop" lines but was omitted from the "abandon" line.
3461 PH/34 Omit spaces between a header name and the colon in the error message that
3462 is given when verify = headers_syntax fails (if there are lots of them,
3463 the message gets confusing).
3465 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3466 names, as there are now some PTR records that contain slashes. This check
3467 is only to protect against broken name servers that fall over on strange
3468 characters, so the fact that it applies to all lookups doesn't matter.
3470 PH/36 Now that the new test suite is complete, we can remove some of the
3471 special code in Exim that was needed for the old test suite. For example,
3472 sorting DNS records because real resolvers return them in an arbitrary
3473 order. The new test suite's fake resolver always returns records in the
3476 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3477 bounces) except when queue_only is set, to avoid logging races between
3478 the different processes.
3480 PH/38 Panic-die if .include specifies a non-absolute path.
3482 PH/39 A tweak to the "H" retry rule from its user.
3484 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3485 a label. They prevented compilation on older perls.
3487 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3488 a warning to be raised on newish perls.
3490 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3491 on queue. Changes to match documented behaviour of showing count of
3492 messages matching specified criteria.
3494 PH/40 Changed the default ident timeout from 30s to 5s.
3496 PH/41 Added support for the use of login_cap features, on those BSD systems
3497 that have them, for controlling the resources used by pipe deliveries.
3499 PH/42 The content-scanning code uses fopen() to create files in which to put
3500 message data. Previously it was not paying any attention to the mode of
3501 the files. Exim runs with umask(0) because the rest of the code creates
3502 files with open(), and sets the required mode explicitly. Thus, these
3503 files were ending up world-writeable. This was not a big issue, because,
3504 being within the spool directory, they were not world-accessible. I have
3505 created a function called modefopen, which takes an additional mode
3506 argument. It sets umask(777), creates the file, chmods it to the required
3507 mode, then resets the umask. All the relevant calls to fopen() in the
3508 content scanning code have been changed to use this function.
3510 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3511 to 24 hours. This avoids potential overflow problems when processing G
3512 and H retry rules. I suspect nobody ever tinkers with this value.
3514 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3516 PH/45 When the plaintext authenticator is running as a client, the server's
3517 challenges are checked to ensure they are valid base64 strings. By
3518 default, the authentication attempt is cancelled if an invalid string is
3519 received. Setting client_ignore_invalid_base64 true ignores these errors.
3520 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3521 they are received. Thus, the responses can be made to depend on the
3522 challenges. If an invalid string is ignored, an empty string is placed in
3525 PH/46 Messages that are created by the autoreply transport now contains a
3526 References: header, in accordance with RFCs 2822 and 3834.
3528 PH/47 Added authenticated_sender_force to the smtp transport.
3530 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3532 PH/49 Installed latest patch from the Sieve maintainer.
3534 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3535 was also set, the appendfile transport was unnecessarily scanning a
3536 directory of message files (e.g. for maildir delivery) to find the count
3537 of files (along with the size), even though it did not need this
3538 information. It now does the scan only if it needs to find either the
3539 size of the count of files.
3541 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3543 PH/52 Two bugs concerned with error handling when the smtp transport is
3546 (i) Exim was not creating retry information for temporary errors given
3547 for individual recipients after the DATA command when the smtp transport
3548 was used in LMTP mode. This meant that they could be retried too
3549 frequently, and not timed out correctly.
3551 (ii) Exim was setting the flag that allows error details to be returned
3552 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3553 recipients that were returned after the DATA command.
3555 PH/53 This is related to PH/52, but is more general: for any failing address,
3556 when detailed error information was permitted to be returned to the
3557 sender, but the error was temporary, then after the final timeout, only
3558 "retry timeout exceeded" was returned. Now it returns the full error as
3559 well as "retry timeout exceeded".
3561 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3562 do this, and (what is worse) MTAs that accept it.
3564 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3565 will now be deprecated.
3567 PH/56 New os.c-cygwin from the Cygwin maintainer.
3569 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3570 formats (previously only available in exim formats via -bpr, -bpru,
3571 and -bpra. Now also available in native and exiqgrep formats)
3573 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3574 with very large, slow to parse queues
3576 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3578 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3580 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3581 responses to authentication challenges, though it was showing the
3582 challenges; (ii) I've removed the CR characters from the debug output for
3585 PH/58 Allow for the insertion of a newline as well as a space when a string
3586 is turned into more than one encoded-word during RFC 2047 encoding. The
3587 Sieve code now uses this.
3589 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3590 data_4xx, lost_connection, tls_required.
3592 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3593 message was being sent as an SMTP response.
3595 PH/61 Add -l and -k options to exicyclog.
3597 PH/62 When verifying, if an address was redirected to one new address, so that
3598 verification continued, and the new address failed or deferred after
3599 having set something in $address_data, the value of $address_data was not
3600 passed back to the ACL. This was different to the case when no
3601 redirection occurred. The value is now passed back in both cases.
3603 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3604 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3605 use login_cap.h, so on its own it isn't the distinguishing feature. The
3606 new name refers directly to the setclassresources() function.
3608 PH/65 Added configuration files for NetBSD3.
3610 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3612 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3613 is preferred over IPv4.
3615 PH/68 The bounce_return_message and bounce_return_body options were not being
3616 honoured for bounces generated during the reception of non-SMTP messages.
3617 In particular, this applied to messages rejected by the ACL. This bug has
3618 been fixed. However, if bounce_return_message is true and bounce_return_
3619 body is false, the headers that are returned for a non-SMTP message
3620 include only those that have been read before the error was detected.
3621 (In the case of an ACL rejection, they have all been read.)
3623 PH/69 The HTML version of the specification is now built in a directory called
3624 spec_html instead of spec.html, because the latter looks like a path with
3625 a MIME-type, and this confuses some software.
3627 PH/70 Catch two compiler warnings in sieve.c.
3629 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3630 function verify_get_ident() calls ip_connect() to connect a socket, but
3631 if the "connect()" function timed out, ip_connect() used to close the
3632 socket. However, verify_get_ident() also closes the socket later, and in
3633 between Exim writes to the log, which may get opened at this point. When
3634 the socket was closed in ip_connect(), the log could get the same file
3635 descriptor number as the socket. This naturally causes chaos. The fix is
3636 not to close the socket in ip_connect(); the socket should be closed by
3637 the function that creates it. There was only one place in the code where
3638 this was missing, in the iplookup router, which I don't think anybody now
3639 uses, but I've fixed it anyway.
3641 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3642 well as to direct DNS lookups. Otherwise the handling of names in host
3643 lists is inconsistent and therefore confusing.
3649 PH/01 Two changes to the default runtime configuration:
3651 (1) Move the checks for relay_from_hosts and authenticated clients from
3652 after to before the (commented out) DNS black list checks.
3654 (2) Add control=submission to the relay_from_hosts and authenticated
3655 clients checks, on the grounds that messages accepted by these
3656 statements are most likely to be submissions.
3658 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3660 (1) Generate an error if the third argument for the ${prvs expansion is
3663 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3666 (3) Reset the variables that are obtained from the first argument of
3667 ${prvscheck and used in the second argument before leaving the code,
3668 because their memory is reclaimed, so using them afterwards may do
3671 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3672 one (it's much easier than Tom thought :-).
3674 (5) Because of (4), we can now allow for the use of $prvscheck_result
3675 inside the third argument.
3677 PH/03 For some reason, the default setting of PATH when running a command from
3678 a pipe transport was just "/usr/bin". I have changed it to
3681 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3682 anything to be listed in the output from -bV.
3684 PH/05 When a filter generated an autoreply, the entire To: header line was
3685 quoted in the delivery log line, like this:
3687 => >A.N.Other <ano@some.domain> <original@ddress> ...
3689 This has been changed so that it extracts the operative address. There
3690 may be more than one such address. If so, they are comma-separated, like
3693 => >ano@some.domain,ona@other.domain <original@ddress> ...
3695 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3696 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3697 not being looked up in the rDNS to get a host name, Exim was showing the
3698 IP address twice in Received: lines, even though the IP addresses were
3699 identical. For example:
3701 Received: from [1.2.3.4] (helo=[1.2.3.4])
3703 However, if the real host name was known, it was omitting the HELO data
3704 if it matched the actual IP address. This has been tidied up so that it
3705 doesn't show the same IP address twice.
3707 PH/07 When both +timestamp and +memory debugging was on, the value given by
3708 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3709 called by the debug printing, thereby overwriting the timestamp buffer.
3710 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3712 PH/08 When the original message was included in an autoreply transport, it
3713 always said "this is a copy of the message, including all the headers",
3714 even if body_only or headers_only was set. It now gives an appropriate
3717 PH/09 Applied a patch from the Sieve maintainer which:
3719 o fixes some comments
3720 o adds the (disabled) notify extension core
3721 o adds some debug output for the result of if/elsif tests
3722 o points to the current vacation draft in the documentation
3723 and documents the missing references header update
3727 o fixes a bug in processing the envelope test (when testing
3728 multiple envelope elements, the last element determined the
3731 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3732 Electronic Mail") by including:
3734 Auto-submitted: auto-generated
3736 in the messages that it generates (bounce messages and others, such as
3737 warnings). In the case of bounce messages for non-SMTP messages, there was
3738 also a typo: it was using "Auto_submitted" (underscore instead of
3739 hyphen). Since every message generated by Exim is necessarily in response
3740 to another message, thes have all been changed to:
3742 Auto-Submitted: auto-replied
3744 in accordance with these statements in the RFC:
3746 The auto-replied keyword:
3748 - SHOULD be used on messages sent in direct response to another
3749 message by an automatic process,
3751 - MUST NOT be used on manually-generated messages,
3753 - MAY be used on Delivery Status Notifications (DSNs) and Message
3754 Disposition Notifications (MDNs),
3756 - MUST NOT be used on messages generated by automatic or periodic
3757 processes, except for messages which are automatic responses to
3760 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3761 to the default Received: header definition.
3763 PH/12 Added log selector acl_warn_skipped (default on).
3765 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3766 variables because (a) they are in the wrong storage pool and (b) even if
3767 they were copied, it wouldn't work properly because of the caching.
3769 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3770 checking when decoding. Apparently there are clients that generate
3771 overlong encoded strings. Why am I not surprised?
3773 PH/15 If the first argument of "${if match_address" was not empty, but did not
3774 contain an "@" character, Exim crashed. Now it writes a panic log message
3775 and treats the condition as false.
3777 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3779 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3780 extension "envelope-auth". The code is finished and in agreement with
3781 other implementations, but there is no documentation so far and in fact,
3782 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3783 not changing the active code.
3785 Print executed "if" and "elsif" statements when debugging is used. This
3786 helps a great deal to understand what a filter does.
3788 Document more things not specified clearly in RFC3028. I had all this
3789 sorted out, when out of a sudden new issues came to my mind. Oops."
3791 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3794 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3795 canonical form (as documented). However, after a host name lookup from
3796 the IP address, check_host() was doing a simple string comparison with
3797 addresses acquired from the DNS when checking that the found name did
3798 have the original IP as one of its addresses. Since any found IPv6
3799 addresses are likely to be in abbreviated form, the comparison could
3800 fail. Luckily, there already exists a function for doing the comparison
3801 by converting both addresses to binary, so now that is used instead of
3802 the text comparison.
3804 PH/20 There was another similar case to PH/19, when a complete host name was
3805 given in a host list; looking up its IP address could give an abbreviated
3806 form, whereas the current host's name might or might not be abbreviated.
3807 The same fix has been applied.
3813 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3814 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3817 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3818 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3820 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3822 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3823 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3824 Also, for "Auto-Submitted", treat anything other than "no" as
3825 non-personal, in accordance with RFC 3834. (Previously it treated
3826 anything starting "auto-" as non-personal.)
3828 TF/01 The control=submission/name=... option had a problem with syntax
3829 errors if the name included a slash character. The /name= option
3830 now slurps the rest of the string, so it can include any characters
3831 but it must come last in the list of options (after /sender_retain
3834 PH/05 Some modifications to the interface to the fake nameserver for the new
3842 TK/01 Added the "success_on_redirect" address verification option. See
3843 NewStuff for rationale and an example.
3845 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3847 PH/02 Patch to exigrep to allow it to work on syslog lines.
3849 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3850 fread() to skip over the body file's header line, because in Cygwin the
3851 header line is locked and is inaccessible.
3853 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3854 co-exist for some time) to make it clear that it is the Exim ID that is
3855 referenced, not the Message-ID: header line.
3857 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3858 string_format() function, because snprintf() does not exist on all
3861 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3862 ${stat: expansion item.
3864 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3865 protocol synchronization error", to keep the pedants happy.
3867 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3868 well as for IRIX systems, when gcc is being used. See the host.c source
3871 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3873 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3876 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3877 transport and to the smtp transport in LMTP mode.
3879 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3881 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3882 run a filter in a subprocess. This could lead to confusion in subsequent
3883 lookups in the parent process. There should also be a search_tidyup() at
3884 the end of the subprocess.
3886 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3887 only if the host matched helo_try_verify_hosts, which caused the
3888 verification to occur when the EHLO/HELO command was issued. The ACL just
3889 tested the remembered result. Now, if a previous verification attempt has
3890 not happened, "verify = helo" does it there and then.
3892 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3894 TK/03 Fix log output including CR from clamd.
3896 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3897 exist provoked a memory error which could cause a segfault.
3899 PH/15 Installed PCRE 6.2
3901 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3903 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3904 of the problem. Specifically, suggested +O2 rather than +O1 for the
3907 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3909 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3910 record, the port was not correctly logged when the outgoing_port log
3911 selector was set (it logged the transort's default port).
3913 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3914 fallback_hosts, and "hosts" in the smtp transport.
3916 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3917 host errors such as "Connection refused".
3919 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3920 authentication with radiusclient 0.4.9:
3922 - Error returned from rc_read_config was caught wrongly
3923 - Username/password not passed on to radius server due to wrong length.
3925 The presumption is that some radiusclient API changes for 4.51/PH/17
3926 were not taken care of correctly. The code is still untested by me (my
3927 Linux distribution still has 0.3.2 of radiusclient), but it was
3928 contributed by a Radius user.
3930 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3931 expanding the "port" option of the smtp transport.
3933 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3934 while reading a MIME header. Thanks to Tom Hughes for a patch.
3936 PH/24 Include config.h inside local_scan.h so that configuration settings are
3939 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3940 that in an ACL for RCPT (for example), you can examine exactly what was
3943 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3944 commands, but it was not correctly comparing the address with the actual
3945 client host address. Thus, it would show the EHLO address in Received:
3946 header lines when this was not necessary.
3948 PH/27 Added the % operator to ${eval:}.
3950 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3951 it should be ignoring failures (because with -C, for example, it has lost
3952 privilege). It wasn't ignoring creation failures other than "already
3955 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3958 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3959 because an input line was too long, either on its own, or by virtue of
3960 too many continuations, the temporary file was not being removed, and the
3961 return code was incorrect.
3963 PH/31 Missing "BOOL" in function definition in filtertest.c.
3965 PH/32 Applied Sieve patches from the maintainer.
3967 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3969 PH/33 Added "verify = not_blind".
3971 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3972 Local/Makefile (with some defaults set). These are used in built scripts
3973 such as exicyclog, but they have never been used in the exim_install
3974 script (though there are many overriding facilities there). I have
3975 arranged that the exim_install script now takes note of these two
3978 PH/35 Installed configuration files for Dragonfly.
3980 PH/36 When a locally submitted message by a trusted user did not contain a
3981 From: header, and the sender address was obtained from -f or from an SMTP
3982 MAIL command, and the trusted user did not use -F to supply a sender
3983 name, $originator_name was incorrectly used when constructing a From:
3984 header. Furthermore, $originator_name was used for submission mode
3985 messages from external hosts without From: headers in a similar way,
3986 which is clearly wrong.
3988 PH/37 Added control=suppress_local_fixups.
3990 PH/38 When log_selector = +received_sender was set, and the addition of the
3991 sender made the log line's construction buffer exactly full, or one byte
3992 less than full, an overflow happened when the terminating "\n" was
3995 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3996 when the result of a list match is failure because a DNS lookup failed.
3998 PH/40 RM_COMMAND is now used in the building process.
4000 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4001 the "build-* directories that it finds.
4003 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4004 address in a domain literal was a prefix of an interface address.
4006 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4007 when verifying a sender address, unless rewrite_headers is false.
4009 PH/44 Wrote a long comment about why errors_to addresses are verified as
4010 recipients, not senders.
4012 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4013 the ratelimit ACL was added.
4015 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4017 PH/46 Added extra information about PostgreSQL errors to the error string.
4019 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4020 avoiding the need to install special zones in a real server. This is
4021 backwards compatible; if it can't find the fake resolver, it drops back.
4022 Thus, both old and new test suites can be run.
4024 TF/02 Added util/ratelimit.pl
4026 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4027 clock is set back in time.
4029 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4030 Candler <B.Candler@pobox.com>.
4032 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4033 OK for addresses that are the result of redirections.
4035 PH/48 A number of further additions for the benefit of the new test suite,
4036 including a fake gethostbyname() that interfaces to the fake DNS resolver
4039 TF/06 The fix for widen_domains has also been applied to qualify_single and
4040 search_parents which are the other dnslookup options that can cause
4043 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4046 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4048 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4049 service" if it encountered an MX record with an empty target hostname.
4050 The message is now "an MX or SRV record indicated no SMTP service".
4052 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4053 if the DNS of the sending site is misconfigured. This is quite a
4054 common situation. This change restores the behaviour of treating a
4055 helo verification defer as a failure.
4057 PH/51 If self=fail was set on a router, the bounce message did not include the
4058 actual error message.
4064 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4066 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4067 command itself ended in error, the underlying message about the transport
4068 filter timeout was being overwritten with the pipe command error. Now the
4069 underlying error message should be appended to the second error message.
4071 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4073 PH/02 Reduce the amount of output that "make" produces by default. Full output
4074 can still be requested.
4076 PH/03 The warning log line about a condition test deferring for a "warn" verb
4077 was being output only once per connection, rather than after each
4078 occurrence (because it was using the same function as for successful
4079 "warn" verbs). This seems wrong, so I have changed it.
4081 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4082 it should not have, which might have caused a crash in the right
4083 circumstances, but probably never did.
4085 PH/04 Installed a modified version of Tony Finch's patch to make submission
4086 mode fix the return path as well as the Sender: header line, and to
4087 add a /name= option so that you can make the user's friendly name appear
4090 TF/03 Added the control = fakedefer ACL modifier.
4092 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4093 Mark Lowes for thorough testing.
4095 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4097 TK/03 Merged latest SRS patch from Miles Wilton.
4099 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4100 with the definition in sysexits.h (which is #included earlier).
4101 Fortunately, Exim does not actually use EX_OK. The code used to try to
4102 preserve the sysexits.h value, by assuming that macro definitions were
4103 scanned for macro replacements. I have been disabused of this notion,
4104 so now the code just undefines EX_OK before #including unistd.h.
4106 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4107 in the smtp transport. When a block could not be written in a single
4108 write() function, the timeout was being re-applied to each part-write.
4109 This seems wrong - if the receiver was accepting one byte at a time it
4110 would take for ever. The timeout is now adjusted when this happens. It
4111 doesn't have to be particularly precise.
4113 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4114 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4116 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4117 without a domain if the check to <postmaster@domain> fails.
4119 SC/01 Eximstats: added -xls and the ability to specify output files
4120 (patch written by Frank Heydlauf).
4122 SC/02 Eximstats: use FileHandles for outputting results.
4124 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4126 SC/04 Eximstats: fixed display of large numbers with -nvr option
4128 SC/05 Eximstats: fixed merging of reports with empty tables.
4130 SC/06 Eximstats: added the -include_original_destination flag
4132 SC/07 Eximstats: removed tabs and trailing whitespace.
4134 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4136 TK/06 MBOX spool code: Add real "From " MBOX separator line
4137 so the .eml file is really in mbox format (even though
4138 most programs do not really care). Patch from Alex Miller.
4140 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4141 The latter is generated from $received_to and is only set if the
4142 message has one envelope recipient. SA can use these headers,
4143 obviously out-of-the-box. Patch from Alex Miller.
4145 PH/08 The ${def test on a variable was returning false if the variable's
4146 value was "0", contrary to what the specification has always said!
4147 The result should be true unless the variable is empty.
4149 PH/09 The syntax error of a character other than { following "${if
4150 def:variable_name" (after optional whitespace) was not being diagnosed.
4151 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4152 accidental colon was present, for example, could give incorrect results.
4154 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4155 result is used (not including appendfile, where other changes are about
4158 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4159 This involved changing a lot of size variables from int to off_t. It
4160 should work with maildirs and everything.
4162 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4163 spamd dying while we are connected to it.
4165 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4168 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4169 function for BDB 4.3.
4171 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4173 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4174 the sources for PCRE have been re-arranged and more files are now
4177 PH/15 The code I had for printing potentially long long variables in PH/11
4178 above was not the best (it lost precision). The length of off_t variables
4179 is now inspected at build time, and an appropriate printing format (%ld
4180 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4181 to be "long long int" or "long int". This is needed for the internal
4182 formatting function string_vformat().
4184 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4185 the configuration file to be ":syslog", then the script "guesses" where
4186 the logs files are, rather than using the compiled in default. In our
4187 case the guess is not the same as the compiled default, so the script
4188 suddenly stopped working when I started to use syslog. The patch checks
4189 to see if log_file_path is "". If so, it attempts to read it from exim
4190 with no configuration file to get the compiled in version, before it
4191 falls back to the previous guessing code."
4193 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4194 implementing BATV in an Exim configuration. See NewStuff for the gory
4197 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4198 Makefile that are specific to HP-UX.
4200 PH/18 If the "use_postmaster" option was set for a recipient callout together
4201 with the "random" option, the postmaster address was used as the MAIL
4202 FROM address for the random test, but not for the subsequent recipient
4203 test. It is now used for both.
4205 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4206 patch removes a few documentation additions to RFC 3028, because the
4207 latest draft now contains them. It adds the new en;ascii-case comparator
4208 and a new error check for 8bit text in MIME parts. Comparator and
4209 require names are now matched exactly. I enabled the subaddress
4210 extension, but it is not well tested yet (read: it works for me)."
4212 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4213 rework some of the code of TK/09 above to avoid the hardwired use of
4214 "%lld" and "long long". Replaced the call to snprintf() with a call to
4217 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4218 records point to non-existent hosts", "retry timeout exceeded", and
4219 "retry time not reached for any host after a long failure period".
4221 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4222 experimental DomainKeys support:
4224 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4225 (2) On an error such as an illegally used "control", the wrong name for
4226 the control was given.
4228 These problems did NOT occur unless DomainKeys support was compiled.
4230 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4232 PH/24 Added ${if match_ip condition.
4234 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4235 databases so that it will be absolutely obvious if a crash occurs in the
4236 DB library. This is a regular occurrence (often caused by mis-matched
4239 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4240 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4241 picked up on a user's system that detects such things. There doesn't seem
4242 to be a gcc warning option for this - only an attribute that has to be
4243 put on the function's prototype. It seems that in Fedora Core 4 they have
4244 set this on a number of new functions. No doubt there will be more in due
4247 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4248 specify a transport. However, if an address that was verified by such a
4249 router was the subject of a callout, Exim crashed because it tried to
4250 read the rcpt_include_affixes from the non-existent transport. Now it
4251 just assumes that the setting of that option is false. This bug was
4252 introduced by 4.51/PH/31.
4254 PH/28 Changed -d+all to exclude +memory, because that information is very
4255 rarely of interest, but it makes the output a lot bigger. People tend to
4256 do -d+all out of habit.
4258 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4259 code in os-type was giving problems when libc.so lives in lib64, like on
4262 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4263 aren't the modern standard, and it seems that some systems' include files
4264 don't always have them. Exim was already checking for some of the newer
4265 ones like T_AAAA, and defining it itself. I've added checks for all the
4266 record types that Exim uses.
4268 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4269 not automatically generating a new one, as it is supposed to. This
4270 prevented TLS from working. If the file did exist, but contained invalid
4271 data, a new version was generated, as expected. It was only the case of a
4272 non-existent file that was broken.
4274 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4275 with a change in libdomainkeys > 0.64.
4277 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4278 from DNS. If the selector record carries the flag, it now has
4279 precedence over the domain-wide flag.
4281 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4283 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4284 the use of a port name that isn't defined in /etc/services) occurred, the
4285 message was deferred as in a normal delivery, and thus remained on the
4286 spool, instead of being failed because of the mua_wrapper setting. This
4287 is now fixed, and I tidied up some of the mua_wrapper messages at the
4290 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4291 the messages in a hash of arrays rather than using individual hashes.
4292 This is a bit cleaner and results in dramatic memory savings, albeit
4293 at a slight CPU cost.
4295 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4296 as requested by Marc Sherman.
4298 SC/10 Eximstats: added histograms for user specified patterns as requested
4301 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4303 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4304 fopen() in the content-scanning modules that did not already have it.
4310 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4311 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4313 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4315 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4317 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4318 to test Sieve filters that use "vacation".
4320 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4321 that changes the way the GnuTLS parameters are stored in the cache file.
4322 The new format can be generated externally. For backward compatibility,
4323 if the data in the cache doesn't make sense, Exim assumes it has read an
4324 old-format file, and it generates new data and writes a new file. This
4325 means that you can't go back to an older release without removing the
4328 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4329 work if there are any delivery delays because "one_time" forces the
4330 parent to be marked "delivered", so its unseen clone is never tried
4331 again. For this reason, Exim now forbids the simultaneous setting of
4334 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4335 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4336 entry if you want to know the details. The fix, however, neglected to
4337 consider the case when local delivery batching is involved. The test for
4338 "previously delivered" was not happening when checking to see if an
4339 address could be batched with a previous (undelivered) one; under
4340 certain circumstances this could lead to multiple deliveries to the same
4343 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4344 in its include files, and this causes problems building Exim.
4346 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4347 header_syntax) but Exim was just ignoring anything given after a slash.
4348 In particular, this caused confusion with an attempt to use "verify =
4349 reverse_host_lookup/defer_ok". An error is now given when options are
4350 supplied for verify items that do not have them. (Maybe reverse_host_
4351 lookup should have a defer_ok option, but that's a different point.)
4353 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4354 defined by RFC 821) to 2048, because there were problems with some AUTH
4355 commands, and RFC 1869 says the size should be increased for extended
4356 SMTP commands that take arguments.
4358 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4361 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4362 "unknown" error; now it says that the functionality isn't in the binary.
4364 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4365 an address' error message when a string expansion fails (syntax or
4366 whatever). Otherwise the password may appear in the log. Following change
4367 PH/42 below, there is no longer a chance of it appearing in a bounce
4370 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4372 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4373 of its addresses was ever tried. (Bugzilla bug #2.)
4375 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4376 the result incorrectly in the debug output. (It correctly added a newline
4377 to what was transported.)
4379 TF/01 Added $received_time.
4381 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4382 commented out examples of how to interface to a virus scanner and to
4383 SpamAssassin. Also added commented examples of av_scanner and
4384 spamd_address settings.
4386 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4387 and controls are allowed in which ACLs. There were a couple of minor
4388 errors. Some of the entries in the conditions table (which is a table of
4389 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4390 negation of where the condition IS allowed.
4392 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4394 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4395 header file does not have a version number, so I've had to invent a new
4396 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4397 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4398 radiusclient), but it was contributed by a Radius user.
4400 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4401 files or directories, for OpenSSL.
4403 PH/19 When an Exim process that is running as root has to create an Exim log
4404 file, it does so in a subprocess that runs as exim:exim so as to get the
4405 ownership right at creation (otherwise, other Exim processes might see
4406 the file with the wrong ownership). There was no test for failure of this
4407 fork() call, which would lead to the process getting stuck as it waited
4408 for a non-existent subprocess. Forks do occasionally fail when resources
4409 run out. I reviewed all the other calls to fork(); they all seem to check
4412 PH/20 When checking for unexpected SMTP input at connect time (before writing
4413 the banner), Exim was not dealing correctly with a non-positive return
4414 from the read() function. If the client had disconnected by this time,
4415 the result was a log entry for a synchronization error with an empty
4416 string after "input=" when read() returned zero. If read() returned -1
4417 (an event I could not check), uninitialized data bytes were printed.
4418 There were reports of junk text (parts of files, etc) appearing after
4421 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4423 PH/22 Added support for macro redefinition, and (re)definition in between
4424 driver and ACL definitions.
4426 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4427 forgetting to use the resulting value; it was using the unexpanded value.
4429 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4430 hadn't been configured. The fix is from Juergen Kreileder, who
4431 understands it better than I do:
4433 "Here's what I see happening with three configured cyrus_sasl
4434 authenticators configured (plain, login, cram-md5):
4436 On startup auth_cyrus_sasl_init() gets called for each of these.
4437 This means three calls to sasl_listmech() without a specified mech_list.
4438 => SASL tests which mechs of all available mechs actually work
4439 => three warnings about OTP not working
4440 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4442 With the patch, sasl_listmech() also gets called three times. But now
4443 SASL's mech_list option is set to the server_mech specified in the the
4444 authenticator. Or in other words, the answer from sasl_listmech()
4445 gets limited to just the mech you're testing for (which is different
4447 => the return list contains just 'plain' or 'login', 'cram-md5' or
4448 nothing depending on the value of ob->server_mech.
4450 I've just tested the patch: Authentication still works fine,
4451 unavailable mechs specified in the exim configuration are still
4452 caught, and the auth.log warnings about OTP are gone."
4454 PH/25 When debugging is enabled, the contents of the command line are added
4455 to the debugging output, even when log_selector=+arguments is not
4458 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4459 answer is "GNU", and only if the return is "GNU/something" is the answer
4462 PH/27 $acl_verify_message is now set immediately after the failure of a
4463 verification in an ACL, and so is available in subsequent modifiers. In
4464 particular, the message can be preserved by coding like this:
4466 warn !verify = sender
4467 set acl_m0 = $acl_verify_message
4469 Previously, $acl_verify_message was set only while expanding "message"
4470 and "log_message" when a very denied access.
4472 PH/28 Modified OS/os.c-Linux with
4474 -#ifndef OS_LOAD_AVERAGE
4475 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4477 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4478 nomenclature these days.)
4480 PH/29 Installed patch from the Sieve maintainer that adds the options
4481 sieve_useraddress and sieve_subaddress to the redirect router.
4483 PH/30 In these circumstances:
4484 . Two addresses routed to the same list of hosts;
4485 . First host does not offer TLS;
4486 . First host accepts first address;
4487 . First host gives temporary error to second address;
4488 . Second host offers TLS and a TLS session is established;
4489 . Second host accepts second address.
4490 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4491 and peerdn, if requested) that were in fact used only for the second
4494 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4495 attention to any local part prefix or suffix that was matched by the
4496 router that accepted the address. It now behaves in the same way as it
4497 does for delivery: the affixes are removed from the local part unless
4498 rcpt_include_affixes is set on the transport.
4500 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4501 timeout during the DATA phase of an incoming message.
4503 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4504 applied a patch sanctioned by the Sieve maintainer.
4506 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4507 the uid or gid is negative. A case of a negative gid caused this to be
4508 noticed. The fix allows for either to be negative.
4510 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4511 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4514 PH/36 The change PH/12 above was broken. Fixed it.
4516 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4517 the grounds that routing the same address twice would always produce the
4518 same answer. This might have been true once, but it is certainly no
4519 longer true now. Routing a child address may depend on the previous
4520 routing that produced that child. Some complicated redirection strategies
4521 went wrong when messages had multiple recipients, and made Exim's
4522 behaviour dependent on the order in which the addresses were given.
4524 I have moved the duplicate checking until after the routing is complete.
4525 Exim scans the addresses that are assigned to local and remote
4526 transports, and removes any duplicates. This means that more work will be
4527 done, as duplicates will always all be routed, but duplicates are
4528 presumably rare, so I don't expect this is of any significance.
4530 For deliveries to pipes, files, and autoreplies, the duplicate checking
4531 still happens during the routing process, since they are not going to be
4534 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4535 It corrects a timeout issue with spamd. This is Ian's comment: "The
4536 background is that sometimes spamd either never reads data from a
4537 connection it has accepted, or it never writes response data. The exiscan
4538 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4539 blindly assumes that writes won't block so it may never time out."
4541 PH/39 Allow G after quota size as well as K and M.
4543 PH/40 The value set for $authenticated_id in an authenticator may not contain
4544 binary zeroes or newlines because the value is written to log lines and
4545 to spool files. There was no check on this. Now the value is run through
4546 the string_printing() function so that such characters are converted to
4547 printable escape sequences.
4549 PH/41 $message_linecount is a new variable that contains the total number of
4550 lines in the message. Compare $body_linecount, which is the count for the
4553 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4554 bounce and delay warning messages, except in certain special cases, which
4557 (a) An SMTP error message from a remote host;
4558 (b) A message specified in a :fail: redirection;
4559 (c) A message specified in a "fail" command in a system filter;
4560 (d) A message specified in a FAIL return from the queryprogram router;
4561 (e) A message specified by the cannot_route_message router option.
4563 In these cases only, Exim does include the error details in bounce and
4564 warning messages. There are also a few cases where bland messages such
4565 as "unrouteable address" or "local delivery error" are given.
4567 PH/43 $value is now also set for the "else" part of a ${run expansion.
4569 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4570 being worked on, but at least Exim now implements the latest version to
4573 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4574 process to complete was treated as a delivery failure, a timeout while
4575 writing the message to the pipe was logged, but erroneously treated as a
4576 successful delivery. Such timeouts include transport filter timeouts. For
4577 consistency with the overall process timeout, these timeouts are now
4578 treated as errors, giving rise to delivery failures by default. However,
4579 there is now a new Boolean option for the pipe transport called
4580 timeout_defer, which, if set TRUE, converts the failures into defers for
4581 both kinds of timeout. A transport filter timeout is now identified in
4584 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4585 systems where "make" and "gmake" are different, calling "gmake" at top
4586 level broke things. I've arranged for the value of $(MAKE) to be passed
4587 from the Makefile to this script so that it can call the same version of
4591 A note about Exim versions 4.44 and 4.50
4592 ----------------------------------------
4594 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4595 changes of various kinds. As a consequence, a big documentation update was
4596 needed. This delayed the release for rather longer than seemed good, especially
4597 in the light of a couple of (minor) security issues. Therefore, the changes
4598 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4599 release. So 4.44 and 4.50 are in effect two different branches that both start
4602 I have left the 4.50 change log unchanged; it contains all the changes since
4603 4.43. The change log for 4.44 is below; many of its items are identical to
4604 those for 4.50. This seems to be the most sensible way to preserve the
4605 historical information.
4611 1. Minor wording change to the doc/README.SIEVE file.
4613 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4614 computation of the current number of files was incorrect.
4616 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4617 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4620 4. Give more explanation in the error message when the command for a transport
4621 filter fails to execute.
4623 5. There are several places where Exim runs a non-Exim command in a
4624 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4625 was being done only for the command run by the queryprogram router. It is
4626 now done for all such subprocesses. The other cases are: ${run, transport
4627 filters, and the commands run by the lmtp and pipe transports.
4629 6. Added CONFIGURE_GROUP build-time option.
4631 7. Some older OS have a limit of 256 on the maximum number of file
4632 descriptors. Exim was using setrlimit() to set 1000 as a large value
4633 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4634 systems. I've change it so that if it can't get 1000, it tries for 256.
4636 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4637 was an oversight, and furthermore, ever since the addition of extra
4638 controls (e.g. 4.43/32), the checks on when to allow different forms of
4639 "control" were broken. There should now be diagnostics for all cases when a
4640 control that does not make sense is encountered.
4642 9. Added the /retain_sender option to "control=submission".
4644 10. $recipients is now available in the predata ACL (oversight).
4646 11. Tidy the search cache before the fork to do a delivery from a message
4647 received from the command line. Otherwise the child will trigger a lookup
4648 failure and thereby defer the delivery if it tries to use (for example) a
4649 cached ldap connection that the parent has called unbind on.
4651 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4652 of $address_data from the recipient verification was clobbered by the
4653 sender verification.
4655 13. The value of address_data from a sender verification is now available in
4656 $sender_address_data in subsequent conditions in the ACL statement.
4658 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4660 15. Added a new option "connect=<time>" to callout options, to set a different
4663 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4664 was its contents. (It was OK if the option was not defined at all.)
4666 17. A "Completed" log line is now written for messages that are removed from
4667 the spool by the -Mrm option.
4669 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4670 information about exactly what failed.
4672 19. Added -dd to debug only the daemon process.
4674 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4675 handles timeouts, both on the server side and network timeouts. Renamed the
4676 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4678 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4679 It is now set to "smtps".
4681 22. $host_address is now set to the target address during the checking of
4682 ignore_target_hosts.
4684 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4685 being passed; this would have caused $sender_host_name to have been used if
4686 matching the list had actually called for a host name (not very likely,
4687 since this list is usually IP addresses). A host name is now passed as
4690 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4691 code that specifies a non-restarting handler (typically sigaction() in
4692 modern systems) in an attempt to fix a rare and obscure crash bug.
4694 25. Narrowed the window for a race in the daemon that could cause it to ignore
4695 SIGCHLD signals. This is not a major problem, because they are used only to
4696 wake it up if nothing else does.
4698 26. A malformed maildirsize file could cause Exim to calculate negative values
4699 for the mailbox size or file count. Odd effects could occur as a result.
4700 The maildirsize information is now recalculated if the size or filecount
4703 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4704 support for a long time. Removed HAVE_SYS_VFS_H.
4706 28. Installed the latest version of exipick from John Jetmore.
4708 29. In an address list, if the pattern was not a regular expression, an empty
4709 subject address (from a bounce message) matched only if the pattern was an
4710 empty string. Non-empty patterns were not even tested. This was the wrong
4711 because it is perfectly reasonable to use an empty address as part of a
4712 database query. An empty address is now tested by patterns that are
4713 lookups. However, all the other forms of pattern expect the subject to
4714 contain a local part and a domain, and therefore, for them, an empty
4715 address still always fails if the pattern is not itself empty.
4717 30. Exim went into a mad DNS loop when attempting to do a callout where the
4718 host was specified on an smtp transport, and looking it up yielded more
4719 than one IP address.
4721 31. Re-factored the code for checking spool and log partition space into a
4722 function that finds that data and another that does the check. The former
4723 is then used to implement four new variables: $spool_space, $log_space,
4724 $spool_inodes, and $log_inodes.
4726 32. The RFC2047 encoding function was originally intended for short strings
4727 such as real names; it was not keeping to the 75-character limit for
4728 encoded words that the RFC imposes. It now respects the limit, and
4729 generates multiple encoded words if necessary. To be on the safe side, I
4730 have increased the buffer size for the ${rfc2047: expansion operator from
4733 33. It is now permitted to omit both strings after an "if" condition; if the
4734 condition is true, the result is "true". As before, when the second string
4735 is omitted, a false condition yields an empty string. This makes it less
4736 cumbersome to write custom ACL and router conditions.
4738 34. Failure to deliver a bounce message always caused it to be frozen, even if
4739 there was an errors_to setting on the router. The errors_to setting is now
4742 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4743 canonical form (fully expanded) before being placed in
4744 $sender_host_address.
4746 36. The table in the code that translates DNS record types into text (T_A to
4747 "A" for instance) was missing entries for NS and CNAME. It is just possible
4748 that this could have caused confusion if both these types were looked up
4749 for the same domain, because the text type is used as part of Exim's
4750 per-process caching. But the chance of anyone hitting this buglet seems
4753 37. The dnsdb lookup has been extended in a number of ways.
4755 (1) There is a new type, "zns", which walks up the domain tree until it
4756 finds some nameserver records. It should be used with care.
4758 (2) There is a new type, "mxh", which is like "mx" except that it returns
4759 just the host names, not the priorities.
4761 (3) It is now possible to give a list of domains (or IP addresses) to be
4762 looked up. The behaviour when one of the lookups defers can be
4763 controlled by a keyword.
4765 (4) It is now possible to specify the separator character for use when
4766 multiple records are returned.
4768 38. The dnslists ACL condition has been extended: it is now possible to supply
4769 a list of IP addresses and/or domains to be looked up in a particular DNS
4772 39. Added log_selector=+queue_time_overall.
4774 40. When running the queue in the test harness, wait just a tad after forking a
4775 delivery process, to get repeatability of debugging output.
4777 41. Include certificate and key file names in error message when GnuTLS fails
4778 to set them up, because the GnuTLS error message doesn't include the name
4779 of the failing file when there is a problem reading it.
4781 42. Allow both -bf and -bF in the same test run.
4783 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4785 44. The "Exiscan patch" is now merged into the mainline Exim source.
4787 45. Sometimes the final signoff response after QUIT could fail to get
4788 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4789 < 0 before doing a fflush(). This bug looks as though it goes back to the
4790 introduction of TLS in release 3.20, but "sometimes" must have been rare
4791 because the tests only now provoked it.
4793 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4794 (this can affect the format of dates).
4796 47. exim_tidydb, when checking for the continued existence of a message for
4797 which it has found a message-specific retry record, was not finding
4798 messages that were in split spool directories. Consequently, it was
4799 deleting retry records that should have stayed in existence.
4801 48. Steve fixed some bugs in eximstats.
4803 49. The SPA authentication driver was not abandoning authentication and moving
4804 on to the next authenticator when an expansion was forced to fail,
4805 contradicting the general specification for all authenticators. Instead it
4806 was generating a temporary error. It now behaves as specified.
4808 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4809 (the order specifies the preference for clients). The order is now AES256,
4810 AES128, 3DES, ARCFOUR128.
4812 51. Small patch to Sieve code - explicitly set From: when generating an
4815 52. Exim crashed if a remote delivery caused a very long error message to be
4816 recorded - for instance if somebody sent an entire SpamAssassin report back
4817 as a large number of 550 error lines. This bug was coincidentally fixed by
4818 increasing the size of one of Exim's internal buffers (big_buffer) that
4819 happened as part of the Exiscan merge. However, to be on the safe side, I
4820 have made the code more robust (and fixed the comments that describe what
4823 53. Now that there can be additional text after "Completed" in log lines (if
4824 the queue_time_overall log selector is set), a one-byte patch to exigrep
4825 was needed to allow it to recognize "Completed" as not the last thing in
4828 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4829 patch that reportedly fixes this has been added. I am not expert enough to
4830 create a test for it. This is what the patch creator wrote:
4832 "I found a little strange behaviour of ldap code when working with
4833 Windows 2003 AD Domain, where users was placed in more than one
4834 Organization Units. When I tried to give exim partial DN, the exit code
4835 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4836 But simultaneously result of request was absolutely normal ldap result,
4837 so I produce this patch..."
4839 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4840 REFERENCE, so I have modified the code to exclude the patch when that macro
4843 55. Some experimental protocols are using DNS PTR records for new purposes. The
4844 keys for these records are domain names, not reversed IP addresses. The
4845 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4846 leaves it alone. Component reversal etc. now happens only for IP addresses.
4849 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4851 57. Double the size of the debug message buffer (to 2048) so that more of very
4852 long debug lines gets shown.
4854 58. The exicyclog utility now does better if the number of log files to keep
4855 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4857 59. Two changes related to the smtp_active_hostname option:
4859 (1) $smtp_active_hostname is now available as a variable.
4860 (2) The default for smtp_banner uses $smtp_active_hostname instead
4861 of $primary_hostname.
4863 60. The host_aton() function is supposed to be passed a string that is known
4864 to be a valid IP address. However, in the case of IPv6 addresses, it was
4865 not checking this. This is a hostage to fortune. Exim now panics and dies
4866 if the condition is not met. A case was found where this could be provoked
4867 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4868 components; fortuitously, this particular loophole had already been fixed
4869 by change 4.50/55 above.
4871 If there are any other similar loopholes, the new check in host_aton()
4872 itself should stop them being exploited. The report I received stated that
4873 data on the command line could provoke the exploit when Exim was running as
4874 exim, but did not say which command line option was involved. All I could
4875 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4876 running as the user.
4879 61. There was a buffer overflow vulnerability in the SPA authentication code
4880 (which came originally from the Samba project). I have added a test to the
4881 spa_base64_to_bits() function which I hope fixes it.
4884 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4885 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4887 63. The daemon start-up calls getloadavg() while still root for those OS that
4888 need the first call to be done as root, but it missed one case: when
4889 deliver_queue_load_max is set with deliver_drop_privilege. This is
4890 necessary for the benefit of the queue runner, because there is no re-exec
4891 when deliver_drop_privilege is set.
4893 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4894 This has been fixed.
4896 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4897 was in use, was not putting the data itself into the right store pool;
4898 consequently, it could be overwritten for a subsequent message in the same
4899 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4902 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4904 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4905 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4906 the function were treating the return as a boolean value, which happened to
4907 work because 0=false and not-0=true, but is not correct code.
4909 68. The host_aton() function was not handling scoped IPv6 addresses (those
4910 with, for example, "%eth0" on the end) correctly.
4912 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4913 negated items (that is, ~something) in unsigned ints. Some compilers
4914 apparently mutter when there is no cast.
4916 70. If an address verification called from an ACL failed, and did not produce a
4917 user-specific message (i.e. there was only a "system" message), nothing was
4918 put in $acl_verify_message. In this situation, it now puts the system
4921 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4922 session; change 4.31/43 added the unwanted input to the log line - except
4923 that it did not do this in the start of session case. It now does.
4925 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4926 This is wrong and can cause the other end to generate a synchronization
4927 error if it is another Exim or anything else that does the synchronization
4928 check. A QUIT command is no longer sent after a timeout.
4930 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4931 during host lookups.
4933 74. The defer_ok option of callout verification was not working if it was used
4934 when verifying addresses in header lines, that is, for this case:
4936 verify = header_sender/callout=defer_ok
4938 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4939 those file descriptors could be used for SMTP connections. If anything
4940 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4941 could be sent to the SMTP client, causing chaos. The daemon now opens
4942 stdin, stdout, and stderr to /dev/null when it puts itself into the
4945 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4946 log by default. The user can override this with suitable Perl magic.
4948 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4949 the log message when discard triggers, was not working for the DATA ACL or
4950 for the non-SMTP ACL.
4952 78. Error message wording change in sieve.c.
4954 79. If smtp_accept_max_per_host was set, the number of connections could be
4955 restricted to fewer than expected, because the daemon was trying to set up
4956 a new connection before checking whether the processes handling previous
4957 connections had finished. The check for completed processes is now done
4958 earlier. On busy systems, this bug wouldn't be noticed because something
4959 else would have woken the daemon, and it would have reaped the completed
4962 80. If a message was submitted locally by a user whose login name contained one
4963 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4964 It caused a spool format error. I have fixed the spool reading code. A
4965 related problem was that the "from" clause in the Received: line became
4966 illegal because of the space(s). It is now covered by ${quote_local_part.
4968 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4971 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4973 83. Patch from Sieve maintainer to fix unterminated string problem in
4974 "vacation" handling.
4976 84. Some minor changes to the Linux configuration files to help with other
4977 OS variants using glibc.
4979 85. One more patch for Sieve to update vacation handling to latest spec.
4982 ----------------------------------------------------
4983 See the note above about the 4.44 and 4.50 releases.
4984 ----------------------------------------------------
4990 1. Change 4.43/35 introduced a bug that caused file counts to be
4991 incorrectly computed when quota_filecount was set in an appendfile
4994 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4995 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4998 3. Give more explanation in the error message when the command for a transport
4999 filter fails to execute.
5001 4. There are several places where Exim runs a non-Exim command in a
5002 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5003 was being done only for the command run by the queryprogram router. It is
5004 now done for all such subprocesses. The other cases are: ${run, transport
5005 filters, and the commands run by the lmtp and pipe transports.
5007 5. Some older OS have a limit of 256 on the maximum number of file
5008 descriptors. Exim was using setrlimit() to set 1000 as a large value
5009 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5010 systems. I've change it so that if it can't get 1000, it tries for 256.
5012 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5013 was an oversight, and furthermore, ever since the addition of extra
5014 controls (e.g. 4.43/32), the checks on when to allow different forms of
5015 "control" were broken. There should now be diagnostics for all cases when a
5016 control that does not make sense is encountered.
5018 7. $recipients is now available in the predata ACL (oversight).
5020 8. Tidy the search cache before the fork to do a delivery from a message
5021 received from the command line. Otherwise the child will trigger a lookup
5022 failure and thereby defer the delivery if it tries to use (for example) a
5023 cached ldap connection that the parent has called unbind on.
5025 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5026 of $address_data from the recipient verification was clobbered by the
5027 sender verification.
5029 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5030 was its contents. (It was OK if the option was not defined at all.)
5032 11. A "Completed" log line is now written for messages that are removed from
5033 the spool by the -Mrm option.
5035 12. $host_address is now set to the target address during the checking of
5036 ignore_target_hosts.
5038 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5039 being passed; this would have caused $sender_host_name to have been used if
5040 matching the list had actually called for a host name (not very likely,
5041 since this list is usually IP addresses). A host name is now passed as
5044 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5045 code that specifies a non-restarting handler (typically sigaction() in
5046 modern systems) in an attempt to fix a rare and obscure crash bug.
5048 15. Narrowed the window for a race in the daemon that could cause it to ignore
5049 SIGCHLD signals. This is not a major problem, because they are used only to
5050 wake it up if nothing else does.
5052 16. A malformed maildirsize file could cause Exim to calculate negative values
5053 for the mailbox size or file count. Odd effects could occur as a result.
5054 The maildirsize information is now recalculated if the size or filecount
5057 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5058 support for a long time. Removed HAVE_SYS_VFS_H.
5060 18. Updated exipick to current release from John Jetmore.
5062 19. Allow an empty sender to be matched against a lookup in an address list.
5063 Previously the only cases considered were a regular expression, or an
5066 20. Exim went into a mad DNS lookup loop when doing a callout where the
5067 host was specified on the transport, if the DNS lookup yielded more than
5070 21. The RFC2047 encoding function was originally intended for short strings
5071 such as real names; it was not keeping to the 75-character limit for
5072 encoded words that the RFC imposes. It now respects the limit, and
5073 generates multiple encoded words if necessary. To be on the safe side, I
5074 have increased the buffer size for the ${rfc2047: expansion operator from
5077 22. Failure to deliver a bounce message always caused it to be frozen, even if
5078 there was an errors_to setting on the router. The errors_to setting is now
5081 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5082 canonical form (fully expanded) before being placed in
5083 $sender_host_address.
5085 24. Updated eximstats to version 1.33
5087 25. Include certificate and key file names in error message when GnuTLS fails
5088 to set them up, because the GnuTLS error message doesn't include the name
5089 of the failing file when there is a problem reading it.
5091 26. Expand error message when OpenSSL has problems setting up cert/key files.
5094 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5095 (this can affect the format of dates).
5097 28. exim_tidydb, when checking for the continued existence of a message for
5098 which it has found a message-specific retry record, was not finding
5099 messages that were in split spool directories. Consequently, it was
5100 deleting retry records that should have stayed in existence.
5102 29. eximstats updated to version 1.35
5103 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5104 1.35 - bugfix such that pie charts by volume are generated correctly
5106 30. The SPA authentication driver was not abandoning authentication and moving
5107 on to the next authenticator when an expansion was forced to fail,
5108 contradicting the general specification for all authenticators. Instead it
5109 was generating a temporary error. It now behaves as specified.
5111 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5112 (the order specifies the preference for clients). The order is now AES256,
5113 AES128, 3DES, ARCFOUR128.
5115 31. Small patch to Sieve code - explicitly set From: when generating an
5118 32. Exim crashed if a remote delivery caused a very long error message to be
5119 recorded - for instance if somebody sent an entire SpamAssassin report back
5120 as a large number of 550 error lines. This bug was coincidentally fixed by
5121 increasing the size of one of Exim's internal buffers (big_buffer) that
5122 happened as part of the Exiscan merge. However, to be on the safe side, I
5123 have made the code more robust (and fixed the comments that describe what
5126 33. Some experimental protocols are using DNS PTR records for new purposes. The
5127 keys for these records are domain names, not reversed IP addresses. The
5128 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5129 leaves it alone. Component reversal etc. now happens only for IP addresses.
5132 34. The host_aton() function is supposed to be passed a string that is known
5133 to be a valid IP address. However, in the case of IPv6 addresses, it was
5134 not checking this. This is a hostage to fortune. Exim now panics and dies
5135 if the condition is not met. A case was found where this could be provoked
5136 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5137 components; fortuitously, this particular loophole had already been fixed
5138 by change 4.50/55 or 4.44/33 above.
5140 If there are any other similar loopholes, the new check in host_aton()
5141 itself should stop them being exploited. The report I received stated that
5142 data on the command line could provoke the exploit when Exim was running as
5143 exim, but did not say which command line option was involved. All I could
5144 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5145 running as the user.
5148 35. There was a buffer overflow vulnerability in the SPA authentication code
5149 (which came originally from the Samba project). I have added a test to the
5150 spa_base64_to_bits() function which I hope fixes it.
5153 36. The daemon start-up calls getloadavg() while still root for those OS that
5154 need the first call to be done as root, but it missed one case: when
5155 deliver_queue_load_max is set with deliver_drop_privilege. This is
5156 necessary for the benefit of the queue runner, because there is no re-exec
5157 when deliver_drop_privilege is set.
5159 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5160 was in use, was not putting the data itself into the right store pool;
5161 consequently, it could be overwritten for a subsequent message in the same
5162 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5165 38. Sometimes the final signoff response after QUIT could fail to get
5166 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5167 < 0 before doing a fflush(). This bug looks as though it goes back to the
5168 introduction of TLS in release 3.20, but "sometimes" must have been rare
5169 because the tests only now provoked it.
5175 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5176 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5177 Now it returns an integer. A number of calls were still expecting a T/F
5178 return. Fortuitously, in all cases, the tests worked in OK situations,
5179 which is the norm. However, things would have gone wrong on any write
5180 failures on the smtp file descriptor. This function is used when sending
5181 messages over SMTP and also when doing verify callouts.
5183 2. When Exim is called to do synchronous delivery of a locally submitted
5184 message (the -odf or -odi options), it no longer closes stderr before doing
5187 3. Implemented the mua_wrapper option.
5189 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5191 5. Implemented the functions header_remove(), header_testname(),
5192 header_add_at_position(), and receive_remove_recipient(), and exported them
5195 6. If an ACL "warn" statement specified the addition of headers, Exim already
5196 inserted X-ACL-Warn: at the start if there was no header name. However, it
5197 was not making this test for the second and subsequent header lines if
5198 there were newlines in the string. This meant that an invalid header could
5199 be inserted if Exim was badly configured.
5201 7. Allow an ACL "warn" statement to add header lines at the start or after all
5202 the Received: headers, as well as at the end.
5204 8. Added the rcpt_4xx retry error code.
5206 9. Added postmaster_mailfrom=xxx to callout verification option.
5208 10. Added mailfrom=xxxx to the callout verification option, for verify=
5211 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5212 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5214 12. Inserted some casts to stop certain compilers warning when using pointer
5215 differences as field lengths or precisions in printf-type calls (mostly
5216 affecting debugging statements).
5218 13. Added optional readline() support for -be (dynamically loaded).
5220 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5221 same clock tick as a message's arrival, so that its received time was the
5222 same as the "first fail" time on the retry record, and that message
5223 remained on the queue past the ultimate address timeout, every queue runner
5224 would try a delivery (because it was past the ultimate address timeout) but
5225 after another failure, the ultimate address timeout, which should have then
5226 bounced the address, did not kick in. This was a "< instead of <=" error;
5227 in most cases the first failure would have been in the next clock tick
5228 after the received time, and all would be well.
5230 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5231 being recognized when the domain list was tested by the match_domain
5232 condition in an expansion string.
5234 16. Added the ${str2b64: operator.
5236 17. Exim was always calling setrlimit() to set a large limit for the number of
5237 processes, without checking whether the existing limit was already
5238 adequate. (It did check for the limit on file descriptors.) Furthermore,
5239 errors from getrlimit() and setrlimit() were being ignored. Now they are
5240 logged to the main and panic logs, but Exim does carry on, to try to do its
5241 job under whatever limits there are.
5243 18. Imported PCRE 5.0.
5245 19. Trivial typo in log message " temporarily refused connection" (the leading
5248 20. If the log selector return_path_on_delivery was set and an address was
5249 redirected to /dev/null, the delivery process crashed because it assumed
5250 that a return path would always be set for a "successful" delivery. In this
5251 case, the whole delivery is bypassed as an optimization, and therefore no
5254 21. Internal re-arrangement: the function for sending a challenge and reading
5255 a response while authentication was assuming a zero-terminated challenge
5256 string. It's now changed to take a pointer and a length, to allow for
5257 binary data in such strings.
5259 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5261 23. Exim was not respecting finduser_retries when seeking the login of the
5262 uid under which it was called; it was always trying 10 times. (The default
5263 setting of finduser_retries is zero.) Also, it was sleeping after the final
5264 failure, which is pointless.
5266 24. Implemented tls_on_connect_ports.
5268 25. Implemented acl_smtp_predata.
5270 26. If the domain in control=submission is set empty, Exim assumes that the
5271 authenticated id is a complete email address when it generates From: or
5272 Sender: header lines.
5274 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5275 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5276 chown and chgrp in /bin and hostname in /usr/bin.
5278 28. Exim was keeping the "process log" file open after each use, just as it
5279 does for the main log. This opens the possibility of it remaining open for
5280 long periods when the USR1 signal hits a daemon. Occasional processlog
5281 errors were reported, that could have been caused by this. Anyway, it seems
5282 much more sensible not to leave this file open at all, so that is what now
5285 29. The long-running daemon process does not normally write to the log once it
5286 has entered its main loop, and it closes the log before doing so. This is
5287 so that log files can straightforwardly be renamed and moved. However,
5288 there are a couple of unusual error situations where the daemon does write
5289 log entries, and I had neglected to close the log afterwards.
5291 30. The text of an SMTP error response that was received during a remote
5292 delivery was being truncated at 512 bytes. This is too short for some of
5293 the long messages that one sometimes sees. I've increased the limit to
5296 31. It is now possible to make retry rules that apply only when a message has a
5297 specific sender, in particular, an empty sender.
5299 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5300 it possible to be selective about when SMTP synchronization is enforced.
5302 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5304 32. Implemented hosts_connection_nolog.
5306 33. Added an ACL for QUIT.
5308 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5311 35. Added mailbox_size and mailbox_filecount to appendfile.
5313 36. Added control = no_multiline_responses to ACLs.
5315 37. There was a bug in the logic of the code that waits for the clock to tick
5316 in the case where the clock went backwards by a substantial amount such
5317 that the microsecond fraction of "now" was more than the microsecond
5318 fraction of "then" (but the whole seconds number was less).
5320 38. Added support for the libradius Radius client library this is found on
5321 FreeBSD (previously only the radiusclient library was supported).
5327 1. When certain lookups returned multiple values in the form name=value, the
5328 quoting of the values was not always being done properly. Specifically:
5329 (a) If the value started with a double quote, but contained no whitespace,
5331 (b) If the value contained whitespace other than a space character (i.e.
5332 tabs or newlines or carriage returns) it was not quoted.
5333 This fix has been applied to the mysql and pgsql lookups by writing a
5334 separate quoting function and calling it from the lookup code. The fix
5335 should probably also be applied to nisplus, ibase and oracle lookups, but
5336 since I cannot test any of those, I have not disturbed their existing code.
5338 2. A hit in the callout cache for a specific address caused a log line with no
5339 reason for rejecting RCPT. Now it says "Previous (cached) callout
5340 verification failure".
5342 3. There was an off-by-one bug in the queryprogram router. An over-long
5343 return line was truncated at 256 instead of 255 characters, thereby
5344 overflowing its buffer with the terminating zero. As well as fixing this, I
5345 have increased the buffer size to 1024 (and made a note to document this).
5347 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5348 when Exim is waiting for an SMTP response from a remote server, Exim
5349 restarts its select() call on the socket, thereby resetting its timeout.
5350 This is not a problem when such interrupts are rare. Somebody set up a cron
5351 job to run exiwhat every 2 minutes, which is less than the normal select()
5352 timeout (5 or 10 minutes). This meant that the select() timeout never
5353 kicked in because it was always reset. I have fixed this by comparing the
5354 time when an interrupt arrives with the time at the start of the first call
5355 to select(). If more time than the timeout has elapsed, the interrupt is
5356 treated as a timeout.
5358 5. Some internal re-factoring in preparation for the addition of Sieve
5359 extensions (by MH). In particular, the "personal" test is moved to a
5360 separate function, and given an option for scanning Cc: and Bcc: (which is
5361 not set for Exim filters).
5363 6. When Exim created an email address using the login of the caller as the
5364 local part (e.g. when creating a From: or Sender: header line), it was not
5365 quoting the local part when it contained special characters such as @.
5367 7. Installed new OpenBSD configuration files.
5369 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5370 try to make them clearer.
5372 9. Callout options, other than the timeout value, were being ignored when
5373 verifying sender addresses in header lines. For example, when using
5375 verify = header_sender/callout=no_cache
5377 the cache was (incorrectly) being used.
5379 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5380 only the Cygwin environment.
5382 11. When return_path_on_delivery was set as a log selector, if different remote
5383 addresses in the same message used different return paths and parallel
5384 remote delivery occurred, the wrong values would sometimes be logged.
5385 (Whenever a remote delivery process finished, the return path value from
5386 the most recently started remote delivery process was logged.)
5388 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5389 lines when AUTH and/or TLS are in use. This is the "received protocol"
5390 field. Exim used to use "asmtp" for authenticated SMTP, without any
5391 indication (in the protocol name) for TLS use. Now it follows the RFC and
5392 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5393 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5394 names appear in log lines as well as in Received: header lines.
5396 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5397 extensions, and comparison tests, and to fix some bugs.
5399 14. Changes to the "personal" filter test:
5401 (1) The test was buggy in that it was just doing the equivalent of
5402 "contains" tests on header lines. For example, if a user's address was
5403 anne@some.where, the "personal" test would incorrectly be true for
5405 To: susanne@some.where
5407 This test is now done by extracting each address from the header in turn,
5408 and checking the entire address. Other tests that are part of "personal"
5409 are now done using regular expressions (for example, to check local parts
5410 of addresses in From: header lines).
5412 (2) The list of non-personal local parts in From: addresses has been
5413 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5414 taken from the Sieve specification recommendations.
5416 (3) If the message contains any header line starting with "List-" it is
5417 treated as non-personal.
5419 (4) The test for "circular" in the Subject: header line has been removed
5420 because it now seems ill-conceived.
5422 15. Minor typos in src/EDITME comments corrected.
5424 16. Installed latest exipick from John Jetmore.
5426 17. If headers_add on a router specified a text string that was too long for
5427 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5428 of string_sprintf() is now avoided.
5430 18. $message_body_size was not set (it was always zero) when running the DATA
5431 ACL and the local_scan() function.
5433 19. For the "mail" command in an Exim filter, no default was being set for
5434 the once_repeat time, causing a random time value to be used if "once" was
5435 specified. (If the value happened to be <= 0, no repeat happened.) The
5436 default is now 0s, meaning "never repeat". The "vacation" command was OK
5437 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5438 (I found it when inspecting the code).
5440 20. There is now an overall timeout for performing a callout verification. It
5441 defaults to 4 times the callout timeout, which applies to individual SMTP
5442 commands during the callout. The overall timeout applies when there is more
5443 than one host that can be tried. The timeout is checked before trying the
5444 next host. This prevents very long delays if there are a large number of
5445 hosts and all are timing out (e.g. when the network connections are timing
5446 out). The value of the overall timeout can be changed by specifying an
5447 additional sub-option for "callout", called "maxwait". For example:
5449 verify = sender/callout=5s,maxwait=20s
5451 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5452 to the end before writing, but this should make it even safer).
5454 22. Exim was forgetting that it had advertised PIPELINING for the second and
5455 subsequent messages on an SMTP connection. It was also not resetting its
5456 memory on STARTTLS and an internal HELO.
5458 23. When Exim logs an SMTP synchronization error within a session, it now
5459 records whether PIPELINING has been advertised or not.
5461 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5462 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5463 rather than long int.
5465 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5467 26. Added the never_mail option to autoreply.
5473 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5474 crash if the getsockname() call failed; this can happen if a connection is
5475 closed very soon after it is established. The problem was simply in the
5476 order in which certain operations were done, causing Exim to try to write
5477 to the SMTP stream before it had set up the file descriptor. The bug has
5478 been fixed by making things happen in the correct order.
5484 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5485 before the connection was closed, thus losing the rejection response.
5487 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5488 some early Solaris releases, but causes trouble in current releases where
5489 socklen_t is defined.
5491 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5494 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5497 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5498 that releasing the top of it at the end releases what was used for sub-
5499 expansions (unless the block got too big). However, discard this block if
5500 the first thing is a variable or header, so that we can use its block when
5501 it is dynamic (useful for very large $message_headers, for example).
5503 6. Lookups now cache *every* query, not just the most recent. A new, separate
5504 store pool is used for this. It can be recovered when all lookup caches are
5505 flushed. Lookups now release memory at the end of their result strings.
5506 This has involved some general refactoring of the lookup sources.
5508 7. Some code has been added to the store_xxx() functions to reduce the amount
5509 of flapping under certain conditions.
5511 8. log_incoming_interface used to affect only the <= reception log lines. Now
5512 it causes the local interface and port to be added to several more SMTP log
5513 lines, for example "SMTP connection from", and rejection lines.
5515 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5517 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5519 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5520 could be overwritten at the end of the current message (or the start of a
5521 new message if it was set in a HELO ACL). The value is now preserved for
5522 the duration of the SMTP connection.
5524 12. If a transport had a headers_rewrite setting, and a matching header line
5525 contained an unqualified address, that address was qualified, even if it
5526 did not match any rewriting rules. The underlying bug was that the values
5527 of the flags that permit the existence of unqualified sender and recipient
5528 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5529 non-local messages, and by -bnq for local messages) were not being
5530 preserved with the message after it was received.
5532 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5533 "next input=" as part of the text comprising the host identity instead of
5534 the correct text. The code was using the same buffer for two different
5535 strings. However, depending on which order the printing function evaluated
5536 its arguments, the bug did not always show up. Under Linux, for example, my
5537 test suite worked just fine.
5539 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5540 doesn't work with some older versions of Perl. It has been changed to "my",
5541 which in any case is probably the better facility to use.
5543 15. A really picky compiler found some instances of statements for creating
5544 error messages that either had too many or two few arguments for the format
5547 16. The size of the buffer for calls to the DNS resolver has been increased
5548 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5549 for addresses that have a lot of PTR records. This alleviates a problem; it
5550 does not fully solve it.
5552 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5553 buffer now truncates the list and logs the incident, which is the same
5554 action as happens when Exim is looking up a host name and its aliases.
5555 Previously in this situation something unpredictable would happen;
5556 sometimes it was "internal error: store_reset failed".
5558 18. If a server dropped the connection unexpectedly when an Exim client was
5559 using GnuTLS and trying to read a response, the client delivery process
5560 crashed while trying to generate an error log message.
5562 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5563 string, for example:
5565 warn message = H1: something\nH2: something
5567 the text was added as a single header line from Exim's point of view
5568 though it ended up OK in the delivered message. However, searching for the
5569 second and subsequent header lines using $h_h2: did not work. This has been
5570 fixed. Similarly, if a system filter added multiple headers in this way,
5571 the routers could not see them.
5573 20. Expanded the error message when iplsearch is called with an invalid key to
5574 suggest using net-iplsearch in a host list.
5576 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5577 ACLs are no longer actually imposed (and a message to that effect is
5580 22. If a "gecos" field in a passwd entry contained escaped characters, in
5581 particular, if it contained a \" sequence, Exim got it wrong when building
5582 a From: or a Sender: header from that name. A second bug also caused
5583 incorrect handling when an unquoted " was present following a character
5584 that needed quoting.
5586 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5587 was not being matched caselessly.
5589 24. Arranged for all hyphens in the exim.8 source to be escaped with
5592 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5593 properly. Recipient callout cache records were still being keyed to include
5594 the sender, even when use_sender was set false. This led to far more
5595 callouts that were necessary. The sender is no longer included in the key
5596 when use_sender is false.
5598 26. Added "control = submission" modifier to ACLs.
5600 27. Added the ${base62d: operator to decode base 62 numbers.
5602 28. dnsdb lookups can now access SRV records.
5604 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5605 the configuration file.
5607 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5608 (-v) mode. This makes the output for a verbose queue run more intelligible.
5610 31. Added a use_postmaster feature to recipient callouts.
5612 32. Added the $body_zerocount variable, containing the number of binary zero
5613 bytes in the message body.
5615 33. The time of last modification of the "new" subdirectory is now used as the
5616 "mailbox time last read" when there is a quota error for a maildir
5619 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5621 35. Added +ignore_unknown as a special item in host lists.
5623 36. Code for decoding IPv6 addresses in host lists is now included, even if
5624 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5625 address was recognized as an IP address, but was then not correctly decoded
5626 into binary, causing unexpected and incorrect effects when compared with
5633 1. Very minor rewording of debugging text in manualroute to say "list of
5634 hosts" instead of "hostlist".
5636 2. If verify=header_syntax was set, and a header line with an unqualified
5637 address (no domain) and a large number of spaces between the end of the
5638 name and the colon was received, the reception process suffered a buffer
5639 overflow, and (when I tested it) crashed. This was caused by some obsolete
5640 code that should have been removed. The fix is to remove it!
5642 3. When running in the test harness, delay a bit after writing a bounce
5643 message to get a bit more predictability in the log output.
5645 4. Added a call to search_tidyup() just before forking a reception process. In
5646 theory, someone could use a lookup in the expansion of smtp_accept_max_
5647 per_host which, without the tidyup, could leave open a database connection.
5649 5. Added the variables $recipient_data and $sender_data which get set from a
5650 lookup success in an ACL "recipients" or "senders" condition, or a router
5651 "senders" option, similar to $domain_data and $local_part_data.
5653 6. Moved the writing of debug_print from before to after the "senders" test
5656 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5657 problems for message scanning, either using a data ACL, or using
5658 local_scan() because the Received: header was not generated till after they
5659 were called (in order to set the time as the time of reception completion).
5660 I have revised the way this works. The header is now generated after the
5661 body is received, but before the ACL or local_scan() are called. After they
5662 are run, the timestamp in the header is updated.
5668 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5669 before starting a queue runner without re-exec. This happened only when
5670 deliver_drop_privilege was set or when the Exim user was set to root. The
5671 effect of the bug was that timeouts during subsequent deliveries caused
5672 crashes instead of being properly handled. The handler is now left at its
5673 default (and expected) setting.
5675 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5676 message, again when deliver_drop_privilege is set or Exim is run as root.
5677 The bug described in (1) was not present in this case, but the tidying up
5678 of the other signals was missing. I have made the two cases consistent.
5680 3. The ignore_target_hosts setting on a manualroute router was being ignored
5681 for hosts that were looked up using the /MX notation.
5683 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5686 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5687 operated on the sender address. After changing the $sender_address to <>
5688 for the sender address verify, Exim was re-instated it as the original
5689 (before rewriting) address, but remembering that it had rewritten it, so it
5690 wasn't rewriting it again. This bug also had the effect of breaking the
5691 sender address verification caching when the sender address was rewritten.
5693 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5694 This has been changed so that if the ip literal address matches
5695 ignore_target_hosts, the router declines.
5697 7. Added expansion conditions match_domain, match_address, and match_local_
5698 part (NOT match_host).
5700 8. The placeholder for the Received: header didn't have a length field set.
5702 9. Added code to Exim itself and to exim_lock to test for a specific race
5703 condition that could lead to file corruption when using MBX delivery. The
5704 issue is with the lockfile that is created in /tmp. If this file is removed
5705 after a process has opened it but before that process has acquired a lock,
5706 there is the potential for a second process to recreate the file and also
5707 acquire a lock. This could lead to two Exim processes writing to the file
5708 at the same time. The added code performs the same test as UW imapd; it
5709 checks after acquiring the lock that its file descriptor still refers to
5710 the same named file.
5712 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5713 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5716 11. Added the smtp_active_hostname option. If used, this will typically be made
5717 to depend on the incoming interface address. Because $interface_address is
5718 not set up until the daemon has forked a reception process, error responses
5719 that can happen earlier (such as "too many connections") no longer contain
5722 12. If an expansion in a condition on a "warn" statement fails because a lookup
5723 defers, the "warn" statement is abandoned, and the next ACL statement is
5724 processed. Previously this caused the whole ACL to be aborted.
5726 13. Added the iplsearch lookup type.
5728 14. Added ident_timeout as a log selector.
5730 15. Added tls_certificate_verified as a log selector.
5732 16. Added a global option tls_require_ciphers (compare the smtp transport
5733 option of the same name). This controls incoming TLS connections.
5735 17. I finally figured out how to make tls_require_ciphers do a similar thing
5736 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5737 before starting the TLS session.
5739 18. Tabs are now shown as \t in -bP output.
5741 19. If the log selector return_path_on_delivery was set, Exim crashed when
5742 bouncing a message because it had too many Received: header lines.
5744 20. If two routers both had headers_remove settings, and the first one included
5745 a superfluous trailing colon, the final name in the first list and the
5746 first name in the second list were incorrectly joined into one item (with a
5747 colon in the middle).
5753 1. Added -C and -D options to the exinext utility, mainly to make it easier
5754 to include in the automated testing, but these could be helpful when
5755 multiple configurations are in use.
5757 2. The exinext utility was not formatting the output nicely when there was
5758 an alternate port involved in the retry record key, nor when there was a
5759 message id as well (for retries that were specific to a specific message
5760 and a specific host). It was also confused by IPv6 addresses, because of
5761 the additional colons they contain. I have fixed the IPv4 problem, and
5762 patched it up to do a reasonable job for IPv6.
5764 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5765 delivery, the log line now contains "pipelined" if PIPELINING was used.
5767 4. An SMTP transport process used to panic and die if the bind() call to set
5768 an explicit outgoing interface failed. This has been changed; it is now
5769 treated in the same way as a connect() failure.
5771 5. A reference to $sender_host_name in the part of a conditional expansion
5772 that was being skipped was still causing a DNS lookup. This no longer
5775 6. The def: expansion condition was not recognizing references to header lines
5776 that used bh_ and bheader_.
5778 7. Added the _cache feature to named lists.
5780 8. The code for checking quota_filecount in the appendfile transport was
5781 allowing one more file than it should have been.
5783 9. For compatibility with Sendmail, the command line option
5791 and sets the incoming protocol and host name (for trusted callers). The
5792 host name and its colon can be omitted when only the protocol is to be set.
5793 Note the Exim already has two private options, -pd and -ps, that refer to
5794 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5795 "s", but I don't think that's a major issue.
5797 10. A number of refactoring changes to the code, none of which should affect
5800 (a) The number of logging options was getting close to filling up the
5801 32-bit word that was used as a bit map. I have split them into two classes:
5802 those that are passed in the argument to log_write(), and those that are
5803 only ever tested independently outside of that function. These are now in
5804 separate 32-bit words, so there is plenty of room for expansion again.
5805 There is no change in the user interface or the logging behaviour.
5807 (b) When building, for example, log lines, the code previously used a
5808 macro that called string_cat() twice, in order to add two strings. This is
5809 not really sufficiently general. Furthermore, there was one instance where
5810 it was actually wrong because one of the argument was used twice, and in
5811 one call a function was used. (As it happened, calling the function twice
5812 did not affect the overall behaviour.) The macro has been replaced by a
5813 function that can join an arbitrary number of extra strings onto a growing
5816 (c) The code for expansion conditions now uses a table and a binary chop
5817 instead of a serial search (which was left over from when there were very
5818 few conditions). Also, it now recognizes conditions like "pam" even when
5819 the relevant support is not compiled in: a suitably worded error message is
5820 given if an attempt is made to use such a condition.
5822 11. Added ${time_interval:xxxxx}.
5824 12. A bug was causing one of the ddress fields not to be passed back correctly
5825 from remote delivery subprocesses. The field in question was not being
5826 subsequently used, so this caused to problems in practice.
5828 13. Added new log selectors queue_time and deliver_time.
5830 14. Might have fixed a bug in maildirsizefile handling that threw up
5831 "unexpected character" debug warnings, and recalculated the data
5832 unnecessarily. In any case, I expanded the warning message to give more
5835 15. Added the message "Restricted characters in address" to the statements in
5836 the default ACL that block characters like @ and % in local parts.
5838 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5839 Three changes have been made:
5841 (a) There was a serious bug; a negative response to MAIL caused the whole
5842 recipient domain to be cached as invalid, thereby blocking all messages
5843 to all local parts at the same domain, from all senders. This bug has
5844 been fixed. The domain is no longer cached after a negative response to
5845 MAIL if the sender used is not empty.
5847 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5850 (c) A new callout option, "use_sender" has been added for people who want
5851 the modified behaviour.
5857 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5860 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5861 indeed breaks things for older releases.
5863 3. Added additional logging to the case where there is a problem reading data
5864 from a filter that is running in a subprocess using a pipe, in order to
5865 try to track down a specific problem.
5867 4. Testing facility fudge: when running in the test harness and attempting
5868 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5869 getting "No route to host". Convert this to a timeout.
5871 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5874 6. Some OS don't have socklen_t but use size_t instead. This affects the
5875 fifth argument of getsockopt() amongst other things. This is now
5876 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5877 can be set for individual OS. I have set it for SunOS5, OSF1, and
5878 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5879 some earlier ones do not.
5881 7. Change 4.30/15 was not doing the test caselessly.
5883 8. The standard form for an IPv6 address literal was being rejected by address
5884 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5885 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5886 this, as well as the form without the "IPv6" on the front (but only when
5887 address literals are enabled, of course).
5889 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5891 10. Exim crashed if a message with an empty sender address specified by -f
5892 encountered a router with an errors_to setting. This could be provoked only
5893 by a command such as
5897 where an empty string was supplied; "<>" did not hit this bug.
5899 11. Installed PCRE release 4.5.
5901 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5902 remained set. It is now erased.
5904 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5905 times from message ids (which are base 36 rather than the normal 62).
5907 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5908 were being counted as actual protocol errors, and logged if the log
5909 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5910 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5911 and DATA following a set of rejected RCPTs do not count as protocol errors.
5912 In other words, Exim assumes they were pipelined, though this may not
5913 actually be the case. Of course, in all cases the client gets an
5914 appropriate error code.
5916 15. If a lookup fails in an ACL condition, a message about the failure may
5917 be available; it is used if testing the ACL cannot continue, because most
5918 such messages specify what the cause of the deferral is. However, some
5919 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5920 that caused an old message to be retained and used if a later statement
5921 caused a defer, replacing the real cause of the deferral.
5923 16. If an IP address had so many PTR records that the DNS lookup buffer
5924 was not large enough to hold them, Exim could crash while trying to process
5925 the truncated data. It now detects and logs this case.
5927 17. Further to 4.21/58, another change has been made: if (and only if) the
5928 first line of a message (the first header line) ends with CRLF, a bare LF
5929 in a subsequent header line has a space inserted after it, so as not to
5930 terminate the header.
5932 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5933 unnecessarily, used atoi() instead of strtol(), and didn't check the
5934 termination when getting file sizes from file names by regex.
5936 19. Completely re-implemented the support for maildirsize files, in the light
5937 of a number of problems with the previous contributed implementation
5938 (4.30/29). In particular:
5940 . If the quota is zero, the maildirsize file is maintained, but no quota is
5943 . If the maildir directory does not exist, it is created before any attempt
5944 to write a maildirsize file.
5946 . The quota value in the file is just a cache; if the quota is changed in
5947 the transport, the new value overrides.
5949 . A regular expression is available for excluding directories from the
5952 20. The autoreply transport checks the characters in options that define the
5953 message's headers; it allows continued headers, but it was checking with
5954 isspace() after an embedded newline instead of explicitly looking for a
5957 21. If all the "regular" hosts to which an address was routed had passed their
5958 expiry times, and had not reached their retry times, the address was
5959 bounced, even if fallback hosts were defined. Now Exim should go on to try
5962 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5963 equivalent code in the SMTP transport. Some hosts send humungous responses
5964 to HELO/EHLO, more than 1024 it seems.
5966 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5967 gives compiler warnings in some environments. I've now done it "properly",
5970 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5971 (because of problems with the built-in one) was declared to return uschar *
5972 instead of char *, causing compiler failure.
5974 25. Fixed a file descriptor leak when processing alias/forward files.
5976 26. Fixed a minor format string issue in dbfn.c.
5978 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5980 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5981 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5984 29. When a sender address is verified, it is cached, to save repeating the test
5985 when there is more than one recipient in a message. However, when the
5986 verification involves a callout, it is possible for different callout
5987 options to be set for different recipients. It is too complicated to keep
5988 track of this in the cache, so now Exim always runs a verification when a
5989 callout is required, relying on the callout cache for the optimization.
5990 The overhead is duplication of the address routing, but this should not be
5993 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5994 to be verified with callout=postmaster, and the main callout worked but the
5995 postmaster check failed, the verification correctly failed. However, if a
5996 subsequent RCPT command asked for sender verification *without* the
5997 postmaster check, incorrect caching caused this verification also to fail,
6000 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6001 it was not caching the DNS options (qualify_single, search_parents) that
6002 were used when the lookup failed. A subsequent lookup with different
6003 options therefore always gave the same answer, though there were cases
6004 where it should not have. (Example: a "domains = !$mx_any" option on a
6005 dnslookup router: the "domains" option is always processed without any
6006 widening, but the router might have qualify_single set.) Now Exim uses the
6007 cached value only when the same options are set.
6009 32. Added John Jetmore's "exipick" utility to the distribution.
6011 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6012 than a timeout (e.g. a certificate is required, and is not provided), an
6013 Exim server now closes the connection immediately. Previously it waited for
6014 the client to close - but if the client is SSL, it seems that they each
6015 wait for each other, leading to a delay before one of them times out.
6017 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6018 maintained 0.8.x compatibility because I don't think many are using it, and
6019 it is clearly obsolete.
6021 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6024 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6025 client certificate was expired. A simple patch fixes this, though I don't
6026 understand the full logic of why the verify callback is called multiple
6029 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6030 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6031 which causes problems with some clients (such as the Certicom SSL Plus
6032 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6033 disables the coutermeasure allowing Eudora to connect."
6035 38. Exim was not checking that a write() to a log file succeeded. This could
6036 lead to Bad Things if a log got too big, in particular if it hit a file
6037 size limit. Exim now panics and dies if it cannot write to a log file, just
6038 as it does if it cannot open a log file.
6040 39. Modified OS/Makefile-Linux so that it now contains
6042 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6044 The two -D definitions ensure that Exim is compiled with large file
6045 support, which makes it possible to handle log files that are bigger than
6048 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6049 instance) a domain was checked against a named list that involved a lookup,
6050 causing $domain_data to be set, then another domain was checked against the
6051 same list, then the first domain was re-checked, the value of $domain_data
6052 after the final check could be wrong. In particular, if the second check
6053 failed, it could be set empty. This bug probably also applied to
6056 41. The strip_trailing_dot option was not being applied to the address given
6057 with the -f command-line option.
6059 42. The code for reading a message's header from the spool was incrementing
6060 $received_count, but never initializing it. This meant that the value was
6061 incorrect (doubled) while delivering a message in the same process in which
6062 it was received. In the most common configuration of Exim, this never
6063 happens - a fresh exec is done - but it can happen when
6064 deliver_drop_privilege is set.
6066 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6067 it now includes up to 150 characters of the unexpected data in the log
6070 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6071 and building data strings. The size of both of these buffers was 10 000
6072 bytes - far larger than anybody would *ever* want, thought I. Needless to
6073 say, somebody hit the limit. I have increased the maximum line length to
6074 20 000 and the maximum data length of concatenated lines to 100 000. I have
6075 also fixed two bugs, because there was no checking on these buffers. Tsk,
6076 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6077 buffer is too small.
6079 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6080 lsearch lookups. Now it does.
6082 46. When parsing a route_list item in a manualroute router, a fixed-length
6083 buffer was used for the list of hosts. I made this 1024 bytes long,
6084 thinking that nobody would ever have a list of hosts that long. Wrong.
6085 Somebody had a whole pile of complicated expansion conditions, and the
6086 string was silently truncated, leading to an expansion error. It turns out
6087 that it is easier to change to an unlimited length (owing to other changes
6088 that have happened since this code was originally written) than to build
6089 structure for giving a limitation error. The length of the item that
6090 expands into the list of hosts is now unlimited.
6092 47. The lsearch lookup could not handle data where the length of text line was
6093 more than 4095 characters. Such lines were truncated, leading to shortened
6094 data being returned. It should now handle lines of any length.
6096 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6097 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6100 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6101 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6102 no longer happens in comment lines. A list of replacements is now placed
6103 at the head of all of the source files, except those whose only change is
6104 to replace PERL_COMMAND in the very first #! line.
6106 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6107 messages on the queue, with a bottom-up merge sort, using code contributed
6108 by Michael Haardt. This should make operations like -bp somewhat faster on
6109 large queues. It won't affect queue runners, except when queue_run_in_order
6112 51. Installed eximstats 1.31 in the distribution.
6114 52. Added support for SRV lookups to the dnslookup router.
6116 53. If an ACL referred to $message_body or $message_body_end, the value was not
6117 reset for any messages that followed in the same SMTP session.
6119 54. The store-handling optimization for building very long strings was not
6120 differentiating between the different store pools. I don't think this
6121 actually made any difference in practice, but I've tidied it.
6123 55. While running the routers to verify a sender address, $sender_address
6124 was still set to the sender address. This is wrong, because when routing to
6125 send a bounce to the sender, it would be empty. Therefore, I have changed
6126 it so that, while verifying a sender address, $sender_address is set to <>.
6127 (There is no change to what happens when verifying a recipient address.)
6129 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6130 target A or AAAA records (if not already returned) without resetting the
6131 qualify_single or search_parents options of the DNS resolver. These are
6132 inappropriate in this case because the targets of MX and SRV records must
6133 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6134 target that, when qualified, matched something in the local domain. These
6135 two options are now turned off when doing these lookups.
6137 57. It seems that at least some releases of Reiserfs (which does not have the
6138 concept of a fixed number of inodes) returns zero and not -1 for the
6139 number of available inodes. This interacted badly with check_spool_inodes,
6140 which assumed that -1 was the "no such thing" setting. What I have done is
6141 to check that the total number of inodes is greater than zero before doing
6142 the test of how many are available.
6144 58. When a "warn" ACL statement has a log_message modifier, the message is
6145 remembered, and not repeated. This is to avoid a lot of repetition when a
6146 message has many recipients that cause the same warning to be written.
6147 However, Exim was preserving the list of already written lines for an
6148 entire SMTP session, which doesn't seem right. The memory is now reset if a
6149 new message is started.
6151 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6152 debugging output unless log_rewrite was also set.
6154 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6155 of (int)(handle) when we know that handle contains (void *)(-1).
6157 61. The Exim daemon panic-logs an error return when it closes the incoming
6158 connection. However "connection reset by peer" seems to be common, and
6159 isn't really an error worthy of noting specially, so that particular error
6162 62. When Exim is trying to find all the local interfaces, it used to panic and
6163 die if the ioctl to get the interface flags failed. However, it seems that
6164 on at least one OS (Solaris 9) it is possible to have an interface that is
6165 included in the list of interfaces, but for which you get a failure error
6166 for this call. This happens when the interface is not "plumbed" into a
6167 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6168 failure of the "get flags" call assumes that the interface is down.
6170 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6171 makes life easier for people who are doing arithmetic on fields extracted
6172 from dates, where you often get leading zeros that should not be
6173 interpreted as octal.
6175 64. Added qualify_domain to the redirect router, to override the global
6178 65. If a pathologically long header line contained very many addresses (the
6179 report of this problem mentioned 10 000) and each of them was rewritten,
6180 Exim could use up a very large amount of memory. (It kept on making new
6181 copies of the header line as it rewrote, and never released the old ones.)
6182 At the expense of a bit more processing, the header rewriting function has
6183 been changed so that it no longer eats memory in this way.
6185 66. The generation of the Received: header has been moved from the time that a
6186 message starts to be received, to the time that it finishes. The timestamp
6187 in the Received: header should now be very close to that of the <= log
6188 line. There are two side-effects of this change:
6190 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6191 logged header lines no longer include the local Received: line, because
6192 it has not yet been created. The same applies to a copy of the message
6193 that is returned to a non-SMTP sender when a message is rejected.
6195 (b) When a filter file is tested using -bf, no additional Received: header
6196 is added to the test message. After some thought, I decided that this
6199 This change does not affect the value of $received_for. It is still set
6200 after address rewriting, but before local_scan() is called.
6202 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6204 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6205 gave an unhelpful panic error message, and a defer error. I have managed to
6206 change this behaviour so that it now rejects any supplied certificate,
6207 which seems right, as the list of acceptable certificates is empty.
6209 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6210 gave an unhelpful defer error. I have not managed to make this reject any
6211 supplied certificates, but the error message it gives is "no certificate
6212 supplied", which is not helpful.
6214 70. exigrep's output now also includes lines that are not associated with any
6215 message, but which match the given pattern. Implemented by a patch from
6216 Martin Sluka, which also tidied up the Perl a bit.
6218 71. Recipient callout verification, like sender verification, was using <> in
6219 the MAIL FROM command. This isn't really the right thing, since the actual
6220 sender may affect whether the remote host accepts the recipient or not. I
6221 have changed it to use the actual sender in the callout; this means that
6222 the cache record is now keyed on a recipient/sender pair, not just the
6223 recipient address. There doesn't seem to be a real danger of callout loops,
6224 since a callout by the remote host to check the sender would use <>.
6225 [SEE ABOVE: changed after hitting problems.]
6227 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6228 temporary errors. However, in the case of such a code being given after
6229 the end of a data transmission (i.e. after ".") Exim was failing to write
6230 a retry record for the message. (Yes, there was some broken host that was
6231 actually sending 8xx at this point.)
6233 73. An unknown lookup type in a host list could cause Exim to panic-die when
6234 the list was checked. (An example that provoked this was putting <; in the
6235 middle of a list instead of at the start.) If this happened during a DATA
6236 ACL check, a -D file could be left lying around. This kind of configuration
6237 error no longer causes Exim to die; instead it causes a defer error. The
6238 incident is still logged to the main and panic logs.
6240 74. Buglet left over from Exim 3 conversion. The message "too many messages
6241 in one connection" was written to the rejectlog but not the mainlog, except
6242 when address rewriting (yes!) was being logged.
6244 75. Added write_rejectlog option.
6246 76. When a system filter was run not as root (that is, when system_filter_user
6247 was set), the values of the $n variables were not being returned to the
6248 main process; thus, they were not subsequently available in the $sn
6251 77. Added +return_path_on_delivery log selector.
6253 78. A connection timeout was being treated differently from recipients deferred
6254 when testing hosts_max_try with a message that was older than the host's
6255 retry timeout. (The host should not be counted, thus allowing all hosts to
6256 be tried at least once before bouncing.) This may have been the cause of an
6257 occasionally reported bug whereby a message would remain on the queue
6258 longer than the retry timeout, but would be bounced if a delivery was
6259 forced. I say "may" because I never totally pinned down the problem;
6260 setting up timeout/retry tests is difficult. See also the next item.
6262 79. The ultimate address timeout was not being applied to errors that involved
6263 a combination of host plus message (for example, a timeout on a MAIL
6264 command). When an address resolved to a number of possible hosts, and they
6265 were not all tried for each delivery (e.g. because of hosts_max_try), a
6266 message could remain on the queue longer than the retry timeout.
6268 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6271 81. Fixed an obscure SMTP outgoing bug which required at least the following
6272 conditions: (a) there was another message waiting for the same server;
6273 (b) the server returned 5xx to all RCPT commands in the first message so
6274 that the message was not completed; (c) the server dropped the connection
6275 or gave a negative response to the RSET that Exim sends to abort the
6276 transaction. The observed case was a dropped connection after DATA that had
6277 been sent in pipelining mode. That is, the server had advertised PIPELINING
6278 but was not implementing it correctly. The effect of the bug was incorrect
6279 behaviour, such as trying another host, and this could lead to a crash.
6285 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6286 and daemon.c were passed as pointers to ints; they should have been
6287 pointers to socklen_t variables (which are typically unsigned ints).
6289 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6292 3. Fixed a really odd bug that affected only the testing scheme; patching a
6293 certain fixed string in the binary changed the value of another string that
6294 happened to be identical to the end of the original first string.
6296 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6297 name", it returns that address as the IP address. On some operating
6298 systems (e.g. Solaris), it also passes back the IP address string as the
6299 "host name". However, on others (e.g. Linux), it passes back an empty
6300 string. Exim wasn't checking for this, and was changing the host name to an
6301 empty string, assuming it had been canonicalized.
6303 5. Although rare, it is permitted to have more than one PTR record for a given
6304 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6305 all the names associated with an address, because they do in Solaris.
6306 However, it seems that they do not in Linux for data that comes from the
6307 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6308 I found this out when I moved to a new Linux workstation and tried to run
6309 the Exim test suite.
6311 To get round this problem I have changed the code so that it now does its
6312 own call to the DNS to look up PTR records when searching for a host name.
6313 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6314 addresses that are only in /etc/hosts are still found.
6316 This behaviour is, however, controlled by an option called host_lookup_
6317 order, which defaults to "bydns:byaddr". If people want to use the other
6318 order, or indeed, just use one or the other means of lookup, they can
6319 specify it in this variable.
6321 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6322 some operating systems, this comes back from gethostbyaddr() as an empty
6323 string, and this is what Exim used to test for. However, it seems that in
6324 other systems, "." is yielded. Exim now tests for this case too.
6326 7. The values of check_spool_space and check_log_space are now held internally
6327 as a number of kilobytes instead of an absolute number of bytes. If a
6328 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6329 kilobyte. This means that much larger values can be stored.
6331 8. Exim monitor: an attempt to get the action menu when not actually pointing
6332 at a message produces an empty menu entitled "No message selected". This
6333 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6334 no entries in it ("Shell widget menu has zero width and/or height"). So I
6335 have added a single, blank menu entry in this case.
6337 9. Added ${quote_local_part.
6339 10. MIME decoding is now applied to the contents of Subject: header lines when
6342 11. Now that a reference to $sender_host_address automatically causes a reverse
6343 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6344 host lookup before query-style lookups in lists that might use this
6345 variable. This has therefore been abolished, and the "net-" prefix is no
6346 longer necessary for query-style lookups.
6348 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6349 have been a typo for LFLAGS, so it has been changed.
6351 13. The install script calls Exim with "-C /dev/null" in order to find the
6352 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6353 to be output. However, since Exim outputs its version number before the
6354 error, it didn't break the script. It just looked ugly. I fixed this by
6355 always allowing "-C /dev/null" if the caller is root.
6357 14. Ignore overlarge ACL variable number when reading spool file - insurance
6358 against a later release with more variables having written the file.
6360 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6361 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6362 this, as well as the form without the "IPv6" on the front.
6364 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6365 OS/Makefile-Darwin file.
6367 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6368 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6370 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6371 variable. However, some OS use ints for inodes. I've added cast to long int
6372 to get rid of the compiler warning.
6374 19. I had forgotten to lock out "/../" in configuration file names when
6375 ALT_CONFIG_PREFIX was set.
6377 20. Routers used for verification do not need to specify transports. However,
6378 if such a router generated a host list, and callout was configured, Exim
6379 crashed, because it could not find a port number from the (non-existent)
6380 transport. It now assumes port 25 in this circumstance.
6382 21. Added the -t option to exigrep.
6384 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6385 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6386 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6387 them is set, LOOKUP_LSEARCH is forced.
6389 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6391 24. Added sender and host information to the "rejected by local_scan()" log
6392 line; previously there was no indication of these.
6394 25. Added .include_if_exists.
6396 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6397 Linux. It turns out that not all file systems support this. Apparently some
6398 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6399 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6400 not supported on the file descriptor, is now ignored when Exim is trying to
6401 sync a directory. This applies only to Linux.
6403 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6405 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6406 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6409 29. Added support for maildirsize files from supplied patch (modified a bit).
6411 30. The use of :fail: followed by an empty string could lead Exim to respond to
6412 sender verification failures with (e.g.):
6414 550 Verification failed for <xxx>
6415 550 Sender verify failed
6417 where the first response line was missing the '-' that indicates it is not
6418 the final line of the response.
6420 31. The loop for finding the name of the user that called Exim had a hardwired
6421 limit of 10; it now uses the value of finduser_retries, which is used for
6422 all other user lookups.
6424 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6427 33. Exim was neglecting to zero errno before one call of strtol() when
6428 expanding a string and expecting an integer value. On some systems this
6429 resulted in spurious "integer overflow" errors. Also, it was casting the
6430 result into an int without checking.
6432 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6433 did not work. The code looks as if it has *never* worked, though it appears
6434 to have been documented since at least release 1.62. I have made it work.
6436 35. The "timeout_DNS" error in retry rules, also documented since at least
6437 1.62, also never worked. As it isn't clear exactly what this means, and
6438 clearly it isn't a major issue, I have abolished the feature by treating it
6439 as "timeout", and writing a warning to the main and panic logs.
6441 36. The display of retry rules for -brt wasn't always showing the error code
6444 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6445 timeout_connect_A, timeout_connect_MX.
6447 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6448 to the empty sender.
6450 39. The daemon was not analysing the content of -oX till after it had closed
6451 stderr and disconnected from the controlling terminal. This meant that any
6452 syntax errors were only noted on the panic log, and the return code from
6453 the command was 0. By re-arranging the code a little, I've made the
6454 decoding happen first, so such errors now appear on stderr, and the return
6455 code is 1. However, the actual setting up of the sockets still happens in
6456 the disconnected process, so errors there are still only recorded on the
6459 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6460 connections (as happens on some IP stacks) was logged at start up time as
6461 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6462 it from "IPv6 and IPv4", which means that two separate sockets are being
6465 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6466 says whether AF_INET or AF_INET6 was passed as an argument.
6468 42. Exiwhat output was messed up when time zones were included in log
6471 43. Exiwhat now gives more information about the daemon's listening ports,
6472 and whether -tls-on-connect was used.
6474 44. The "port" option of the smtp transport is now expanded.
6476 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6477 silently ignored. Now an error message is written to the main and panic
6480 46. There's a new ACL modifier called "logwrite" which writes to a log file
6481 as soon as it is encountered.
6483 47. Added $local_user_uid and $local_user_gid at routing time.
6485 48. Exim crashed when trying to verify a sender address that was being
6488 49. Exim was recognizing only a space character after ".include". It now also
6489 recognizes a tab character.
6491 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6492 extracting the relevant information from the specification. The man page no
6493 longer contains scrambled data for the -d option, and I've added a section
6494 at the front about calling Exim under different names.
6496 51. Added "extra_headers" argument to the "mail" command in filter files.
6498 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6501 53. Installed eximstats 1.29.
6503 54. Added transport_filter_timeout as a generic transport option.
6505 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6506 Cc: header lines. This was required by RFC 822, but it not required by RFC
6509 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6510 incoming messages that did not have them. Now it does so only if the
6511 message originates locally, that is, if there is no associated remote host
6512 address. When Resent- header lines are present, this applies to the Resent-
6513 lines rather than the non-Resent- lines.
6515 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6516 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6518 58. Messages for configuration errors now include the name of the main
6519 configuration file - useful now that there may be more than one file in a
6520 list (.included file names were always shown).
6522 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6523 for those rare installations that do not start the daemon as root or run it
6524 setuid root. I've cut out the call to initgroups() if the daemon is not
6527 60. The Exim user and group can now be bound into the binary as text strings
6528 that are looked up at the start of Exim's processing.
6530 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6532 62. Added $mailstore_basename variable.
6534 63. Installed patch to sieve.c from Michael Haardt.
6536 64. When Exim failed to open the panic log after failing to open the main log,
6537 the original message it was trying to log was written to stderr and debug
6538 output, but if they were not available (the usual case in production), it
6539 was lost. Now it is written to syslog before the two lines that record the
6540 failures to open the logs.
6542 65. Users' Exim filters run in subprocesses under the user's uid. It is
6543 possible for a "deliver" command or an alias in a "personal" command to
6544 provoke an address rewrite. If logging of address rewriting is configured,
6545 this fails because the process is not running as root or exim. There may be
6546 a better way of dealing with this, but for the moment (because 4.30 needs
6547 to be released), I have disabled address rewrite logging when running a
6548 filter in a non-root, non-exim process.
6554 1. The buildconfig auxiliary program wasn't quoting the value set for
6555 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6556 not defined. This bug was masked in 4.22 by the effect that was fixed in
6559 2. Some messages that were rejected after a message id was allocated were
6560 shown as "incomplete" by exigrep. It no longer does this for messages that
6561 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6563 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6564 have allow_domain_literals set, the ID did not get logged in the <= line.
6565 Domain literals are now always recognized in Message-ID: header lines.
6567 4. The first argument for a ${extract expansion item is the key name or field
6568 number. Leading and trailing spaces in this item were not being ignored,
6569 causing some misleading effects.
6571 5. When deliver_drop_privilege was set, single queue runner processes started
6572 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6573 same command in the process it spins off) were not dropping privilege.
6575 6. When the daemon running as "exim" started a queue runner, it always
6576 re-executed Exim in the spun-off process. This is a waste of effort when
6577 deliver_drop_privilege is set. The new process now just calls the
6578 queue-runner function directly.
6584 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6587 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6588 ignored. Though the use of -oP was forcing the writing of a pid file, it
6589 was always written to the default place.
6591 3. If the message "no IP address found for host xxxx" is generated during
6592 incoming verification, it is now followed by identification of the incoming
6593 connection (so you can more easily find what provoked it).
6595 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6597 5. Added some features to "harden" Exim a bit more against certain attacks:
6599 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6600 be put in Local/Makefile. This is like the never_users runtime option,
6601 but it cannot be overridden. The default setting is "root".
6603 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6604 prefix string with which any file named in a -C command line option
6607 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6608 is retained for -C and -D only if the caller of Exim is root. Without
6609 it, the exim user may also use -C and -D and retain privilege.
6611 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6612 command line option is disabled.
6614 6. Macro names set by the -D option must start with an upper case letter, just
6615 like macro names defined in the configuration file.
6617 7. Added "dereference=" facility to LDAP.
6619 8. Two instances of the typo "uknown" in the source files are fixed.
6621 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6622 the Configure-Makefile script screwed up while processing it.
6624 10. Incorporated PCRE 4.4.
6626 11. The SMTP synchronization check was not operating right at the start of an
6627 SMTP session. For example, it could not catch a HELO sent before the client
6628 waited for the greeting. There is now a check for outstanding input at the
6629 point when the greeting is written. Because of the duplex, asynchronous
6630 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6631 way, but not yet received, when the check is performed.
6633 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6634 on TCP/IP sockets, because this apparently causes some broken clients to
6637 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6638 unchanged) from the Cygwin maintainer.
6640 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6641 was supported instead of testing for mbx. Effectively a typo.
6643 15. The spa authenticator server code was not checking that the input it
6644 received was valid base64.
6646 16. The debug output line for the "set" modifier in ACLs was not showing the
6647 name of the variable that was being set.
6649 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6651 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6652 lookup on its own. Something else was needed to trigger the lookup. For
6653 example, a match in host_lookup or the need for a host name in a host list.
6654 Now, if $sender_host_name is referenced and the host name has not yet been
6655 looked up, a lookup is performed. If the lookup fails, the variable remains
6656 empty, and $host_lookup_failed is set to "1".
6658 19. Added "eqi" as a case-independent comparison operator.
6660 20. The saslauthd authentication condition could segfault if neither service
6661 nor realm was specified.
6663 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6664 error message that was logged was misleading, and incoming SMTP
6665 connections were dropped. The message is now more accurate, and temporary
6666 errors are given to SMTP connections.
6668 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6669 (except RSET) with a 421 error, until QUIT is received. However, it was
6670 failing to send a response to QUIT.
6672 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6673 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6674 ACL is now run after the helo_try_verify_hosts code.
6676 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6677 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6678 case-independent, so other case variants are also recognized). Apparently
6679 some systems use these upper case variants.
6681 25. If more than two messages were waiting for the same host, and a transport
6682 filter was specified for the transport, Exim sent two messages over the
6683 same TCP/IP connection, and then failed with "socket operation on non-
6684 socket" when it tried to send the third.
6686 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6688 27. The extern definition of crypt16() in expand.c was not being excluded when
6689 the OS had its own crypt16() function.
6691 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6692 as a preferred synonym for return_size_limit, both as an option and as an
6695 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6697 30. Changed the default configuration ACL to relax the local part checking rule
6698 for addresses that are not in any local domains. For these addresses,
6699 slashes and pipe symbols are allowed within local parts, but the sequence
6700 /../ is explicitly forbidden.
6702 31. SPA server authentication was not clearing the challenge buffer before
6705 32. log_message in a "warn" ACL statement was writing to the reject log as
6706 well as to the main log, which contradicts the documentation and doesn't
6707 seem right (because no rejection is happening). So I have stopped it.
6709 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6710 However, I am unable to do any testing of this.
6712 34. Fixed an infelicity in the appendfile transport. When checking directories
6713 for a mailbox, to see if any needed to be created, it was accidentally
6714 using path names with one or more superfluous leading slashes; tracing
6715 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6717 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6718 contents are added to the log line that is written for every discarded
6719 recipient. (Previously a log_message setting was ignored.)
6721 36. The ${quote: operator now quotes the string if it is empty.
6723 37. The install script runs exim in order to find its version number. If for
6724 some reason other than non-existence or emptiness, which it checks, it
6725 could not run './exim', it was installing it with an empty version number,
6726 i.e. as "exim-". This error state is now caught, and the installation is
6729 38. An argument was missing from the function that creates an error message
6730 when Exim fails to connect to the socket for saslauthd authentication.
6731 This could cause Exim to crash, or give a corrupted message.
6733 39. Added isip, isip4, and isip6 to ${if conditions.
6735 40. The ACL variables $acl_xx are now saved with the message, and can be
6736 accessed later in routers, transports, and filters.
6738 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6739 strings in the file are not string-expanded.
6741 42. If a MAIL command specified a SIZE value that was too large to fit into an
6742 int variable, the check against message_size_limit failed. Such values are
6743 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6744 day this will have to be increased, but I don't think I want to be around
6745 when emails are that large.
6752 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6753 iconv() is not standard in FreeBSD.
6755 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6756 IPv6 enabled. The observed symptom was a segmentation fault on return from
6757 the function os_common_find_running_interfaces() in src/os.c.
6759 3. In the check_special_case() function in daemon.c I had used "errno" as an
6760 argument name, which causes warnings on some systems. This was basically a
6761 typo, since it was named "eno" in the comments!
6763 4. The code that waits for the clock to tick (at a resolution of some fraction
6764 of a second) so as to ensure message-id uniqueness was always waiting for
6765 at least one whole tick, when it could have waited for less. [This is
6766 almost certainly not relevant at current processor speeds, where it is
6767 unlikely to ever wait at all. But we try to future-proof.]
6769 5. The function that sleeps for a time interval that includes fractions of a
6770 second contained a race. It did not block SIGALRM between setting the
6771 timer, and suspending (a couple of lines later). If the interval was short
6772 and the sigsuspend() was delayed until after it had expired, the suspension
6773 never ended. On busy systems this could lead to processes getting stuck for
6776 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6777 process, before it forks any delivery processes. The open lookup caching
6778 mechanism meant that the open file or database connection was passed into
6779 the delivery process. The problem was that delivery processes always tidy
6780 up cached lookup data. This could cause a problem for the next delivery
6781 process started by the queue runner, because the external queue runner
6782 process does not know about the closure. So the next delivery process
6783 still has data in the lookup cache. In the case of a file lookup, there was
6784 no problem because closing a file descriptor in a subprocess doesn't affect
6785 the parent. However, if the lookup was caching a connection to a database,
6786 the connection was closed, and the second delivery process was likely to
6787 see errors such as "PGSQL: query failed: server closed the connection
6788 unexpectedly". The problem has been fixed by closing all cached lookups
6789 in a queue runner before running a delivery process.
6791 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6792 seem to have the "const" qualifier which it has on other OS. I've
6795 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6796 *of the same type* (client or server) with the same public name that an
6797 error should be diagnosed.
6799 9. When Exim looked up a host name for an IP address, but failed to find the
6800 original IP address when looking up the host name (a safety check), it
6801 output the message "<ip address> does not match any IP for NULL", which was
6802 confusing, to say the least. The bug was that the host name should have
6803 appeared instead of "NULL".
6805 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6806 user that is built into the binary, and the -C or -D options is used, root
6807 privilege is dropped before the configuration file is read. In addition,
6808 logging is switched to stderr instead of the normal log files. If the
6809 configuration then re-defines the Exim user, the unprivileged environment
6810 is probably not what is expected, so Exim logs a panic warning message (but
6813 However, if deliver_drop_privilege is set, the unprivileged state may well
6814 be exactly what is intended, so the warning has been cut out in that case,
6815 and Exim is allowed to try to write to its normal log files.
6821 1. smtp_return_error_details was not giving details for temporary sender
6822 or receiver verification errors.
6824 2. Diagnose a configuration error if two authenticators have the same public
6827 3. Exim used not to create the message log file for a message until the first
6828 delivery attempt. This could be confusing when incoming messages were held
6829 for policy or load reasons. The message log file is now created at the time
6830 the message is received, and an initial "Received" line is written to it.
6832 4. The automatically generated man page for command line options had a minor
6833 bug that caused no ill effects; however, a more serious problem was that
6834 the procedure for building the man page automatically didn't always
6835 operate. Consequently, release 4.20 contains an out-of-date version. This
6836 shouldn't happen again.
6838 5. When building Exim with embedded Perl support, the script that builds the
6839 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6840 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6842 6. The freeze_tell option was not being used for messages that were frozen on
6843 arrival, either by an ACL or by local_scan().
6845 7. Added the smtp_incomplete_transaction log selector.
6847 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6848 was accepting AUTH without a new EHLO.
6850 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6851 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6854 10. Logging of TCP/IP connections (when configured) now happens in the main
6855 daemon process instead of the child process, so that the TCP/IP connection
6856 count is more accurate (but it can never be perfect).
6858 11. The use of "drop" in a nested ACL was not being handled correctly in the
6859 outer ACL. Now, if condition failure induced by the nested "drop" causes
6860 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6861 or "require"), the connection is dropped.
6863 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6864 that yield "discard" can now be used with an "accept" or a "discard" verb,
6865 but an error is generated for any others (because I can't see a useful way
6866 to define what should happen).
6868 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6869 are now processed in the same way as lines in the Exim configuration file.
6870 In particular, continuation lines are supported.
6872 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6874 15. Added -ti meaning -t -i.
6876 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6877 domains, and warn by logging if others are found.
6879 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6880 returned from GIFCONF when Exim is trying to find the list of interfaces on
6881 a host. The code in os.c has been modified to copy each ifreq to an aligned
6882 structure in all cases.
6884 Also, in some cases, the returned ifreq's were being copied to a 'struct
6885 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6886 means the last couple of bytes of an IPv6 address could be chopped if the
6887 ifreq contained only a normal sockaddr (14 bytes storage).
6889 18. Named domain lists were not supported in the hosts_treat_as_local option.
6890 An entry such as +xxxx was not recognized, and was treated as a literal
6893 19. Ensure that header lines added by a DATA ACL are included in the reject log
6894 if the ACL subsequently rejects the message.
6896 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6897 MD5 (which is deprecated).
6899 21. When testing a filter file using -bf, Exim was writing a message when it
6900 took the sender from a "From " line in the message, but it was not doing so
6901 when it took $return_path from a Return-Path: header line. It now does.
6903 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6904 with a valid header line field name (a series of printing characters
6905 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6907 23. Changed "disc" in the source to "disk" to conform to the documentation and
6908 the book and for uniformity.
6910 24. Ignore Sendmail's -Ooption=value command line item.
6912 25. When execve() failed while trying to run a command in a pipe transport,
6913 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6914 could be confused with a return value of 69 from the command itself. This
6915 has been changed to 127, the value the shell returns if it is asked to run
6916 a non-existent command. The wording for the related log line suggests a
6917 non-existent command as the problem.
6919 26. If received_header_text expands to an empty string, do not add a Received:
6920 header line to the message. (Well, it adds a token one on the spool, but
6921 marks it "old" so that it doesn't get used or transmitted.)
6923 27. Installed eximstats 1.28 (addition of -nt option).
6925 28. There was no check for failure on the call to getsockname() in the daemon
6926 code. This can fail if there is a shortage of resources on the system, with
6927 ENOMEM, for example. A temporary error is now given on failure.
6929 29. Contrary to the C standard, it seems that in some environments, the
6930 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6931 program. Exim now does this explicitly; it affects the formatting of
6932 timestamps using strftime().
6934 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6935 complaints. I've now initialized all the variables, to avoid this.
6937 32. Header lines added by a system filter were not being "seen" during
6938 transport-time rewrites.
6940 33. The info_callback() function passed to OpenSSL is set up with type void
6941 (*)(SSL *, int, int), as described somewhere. However, when calling the
6942 function (actually a macro) that sets it up, the type void(*)() is
6943 expected. I've put in a cast to prevent warnings from picky compilers.
6945 34. If a DNS black list lookup found a CNAME record, but there were no A
6946 records associated with the domain it pointed at, Exim crashed.
6948 35. If a DNS black list lookup returned more than one A record, Exim ignored
6949 all but the first. It now scans all returned addresses if a particular IP
6950 value is being sought. In this situation, the contents of the
6951 $dnslist_value variable are a list of all the addresses, separated by a
6954 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6955 to accept a host name and all its aliases if the forward lookup for any of
6956 them yielded the IP address of the incoming connection. Now it accepts only
6957 those names whose forward lookup yields the correct IP address. Any other
6958 names are discarded. This closes a loophole whereby a rogue DNS
6959 administrator could create reverse DNS records to break through a
6960 wildcarded host restriction in an ACL.
6962 37. If a user filter or a system filter that ran in a subprocess used any of
6963 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6964 the wrong values were passed to the pipe command ($thisaddress had the
6965 value of $0, $0 had the value of $1, etc). This bug was introduced by
6966 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6968 38. Improved the line breaking for long SMTP error messages from ACLs.
6969 Previously, if there was no break point between 40 and 75 characters, Exim
6970 left the rest of the message alone. Two changes have been made: (a) I've
6971 reduced the minimum length to 35 characters; (b) if it can't find a break
6972 point between 35 and 75 characters, it looks ahead and uses the first one
6973 that it finds. This may give the occasional overlong line, but at least the
6974 remaining text gets split now.
6976 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6977 file descriptors might be low, and that setting 1000 would always raise it.
6978 It turns out that in some environments, the limit is already over 1000 and
6979 that lowering it causes trouble. So now Exim takes care not to decrease it.
6981 40. When delivering a message, the value of $return_path is set to $sender_
6982 address at the start of routing (routers may change the value). By an
6983 oversight, this default was not being set up when an address was tested by
6984 -bt or -bv, which affected the outcome if any router or filter referred to
6987 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6988 log only when "message" or "log_message" are set. However, if one of the
6989 conditions was an address verification, or a call to a nested ACL, the
6990 messages generated by the underlying test were being passed through. This
6991 no longer happens. The underlying message is available in $acl_verify_
6992 message for both "message" and "log_message" expansions, so it can be
6993 passed through if needed.
6995 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6996 new expansion $bh_ to give the encoded byte string without charset
6997 translation. Translation happens only if iconv() is available; HAVE_ICONV
6998 indicates this at build time. HEADERS_CHARSET gives the charset to
6999 translate to; headers_charset can change it in the configuration, and
7000 "headers charset" can change it in an individual filter file.
7002 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7003 that creates RFC 2047 encoded "words" labels them as that charset instead
7004 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7005 expansion operator; (ii) when Exim creates a From: line for a local
7006 message; (iii) when a header line is rewritten to include a "phrase" part.
7008 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7009 buggy, causing it to skip the first lines of messages whose message ID
7010 ended in 'D'. This would not have bitten before Exim release 4.14, because
7011 message IDs were unlikely to end in 'D' before then. The effect was to have
7012 incorrect size information for certain domains.
7014 45. #include "config.h" was missing at the start of the crypt16.c module. This
7015 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7018 46. If there was a timeout during a "random" callout check, Exim treated it as
7019 a failure of the random address, and carried on sending RSET and the real
7020 address. If the delay was just some slowness somewhere, the response to the
7021 original RCPT would be taken as a response to RSET and so on, causing
7022 mayhem of various kinds.
7024 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7025 when I implemented it. It didn't allow for the fact that some option values
7026 may legitimately be negative (e.g. size_addition), and it didn't even do
7027 the right test for positive values.
7029 48. Domain names in DNS records are case-independent. Exim always looks them up
7030 in lower case. Some resolvers return domain names in exactly the case they
7031 appear in the zone file, that is, they may contain uppercase letters. Not
7032 all resolvers do this - some return always lower case. Exim was treating a
7033 change of case by a resolver as a change of domain, similar to a widening
7034 of a domain abbreviation. This triggered its re-routing code and so it was
7035 trying to route what was effectively the same domain again. This normally
7036 caused routing to fail (because the router wouldn't handle the domain
7037 twice). Now Exim checks for this case specially, and just changes the
7038 casing of the domain that it ultimately uses when it transmits the message
7041 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7044 50. If a filter generated a file delivery with a non-absolute name (possible if
7045 no home directory exists for the router), the forbid_file option was not
7048 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7049 the existing equality matching.
7051 52. Exim was using ints instead of ino_t variables in some places where it was
7052 dealing with inode numbers.
7054 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7055 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7056 called TMPDIR, and if it finds it is different, it changes its value.
7058 54. The smtp_printf() function is now made available to local_scan() so
7059 additional output lines can be written before returning. There is also an
7060 smtp_fflush() function to enable the detection of a dropped connection.
7061 The variables smtp_input and smtp_batched_input are exported to
7064 55. Changed the default runtime configuration: the message "Unknown user"
7065 has been removed from the ACL, and instead placed on the localuser router,
7066 using the cannot_route_message feature. This means that any verification
7067 failures that generate their own messages won't get overridden. Similarly,
7068 the "Unrouteable address" message that was in the ACL for unverifiable
7069 relay addresses has also been removed.
7071 56. Added hosts_avoid_esmtp to the smtp transport.
7073 57. The exicyclog script was not checking for the esoteric option
7074 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7075 will work only if exicyclog is run under the appropriate euid.
7077 58. Following a discussion on the list, the rules by which Exim recognises line
7078 endings on incoming messages have been changed. The -dropcr and drop_cr
7079 options are now no-ops, retained only for backwards compatibility. The
7080 following line terminators are recognized: LF CRLF CR. However, special
7081 processing applies to CR:
7083 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7084 nor a local message in the state where . is a terminator.
7086 (ii) If a bare CR is encountered in a header line, an extra space is added
7087 after the line terminator so as not to end the header. The reasoning
7088 behind this is that bare CRs in header lines are most likely either
7089 to be mistakes, or people trying to play silly games.
7091 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7092 was being incorrectly given as 18 bytes larger than it should have been.
7093 This is a VOB (very old bug).
7095 60. This may never have affected anything current, but just in case it has:
7096 When the local host is found other than at the start of a list of hosts,
7097 the local host, those with the same MX, and any that follow, are discarded.
7098 When the list in question was part of a longer list of hosts, the following
7099 hosts (not currently being processed) were also being discarded. This no
7100 longer happens. I'm not sure if this situation could ever has previously
7103 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7106 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7107 header. This is something that is recommended in a new Internet Draft, and
7108 is something that is documented as being done by Sendmail. There are two
7109 possible values. For messages generated by the autoreply transport, Exim
7112 Auto-Submitted: auto-replied
7114 whereas for all other generated messages (e.g. bounces) it adds
7116 Auto-Submitted: auto-generated
7118 63. The "personal" condition in filters now includes a test for the
7119 Auto-Submitted: header. If it contains the string "auto-" the message it
7120 not considered personal.
7122 64. Added rcpt_include_affixes as a generic transport option.
7124 65. Added queue_only_override (default true).
7126 66. Added the syslog_duplication option.
7128 67. If what should have been the first header line of a message consisted of
7129 a space followed by a colon, Exim was mis-interpreting it as a header line.
7130 It isn't of course - it is syntactically invalid and should therefore be
7131 treated as the start of the message body. The misbehaviour could have
7132 caused a number of strange effects, including loss of data in subsequent
7133 header lines, and spool format errors.
7135 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7136 client host had authenticated. This control can now be exercised by an ACL
7137 for more flexibility.
7139 69. By default, callouts do not happen when testing with -bh. There is now a
7140 variant, -bhc, which does actually run the callout code, including
7141 consulting and updating the callout cache.
7143 70. Added support for saslauthd authentication, courtesy of Alexander
7146 71. If statvfs() failed on the spool or log directories while checking their
7147 size for availability, Exim confusingly gave the error "space shortage".
7148 Furthermore, in debugging mode it crashed with a floating point exception.
7149 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7150 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7151 serious problem, Exim now writes to the main and panic logs when this
7152 happens, with details of the failure. It then refuses to accept the
7153 incoming message, giving the message "spool directory problem" or "log
7154 directory problem" with a 421 code for SMTP messages.
7156 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7157 0, 1, and 2 exist, because some OS complain for execs without them (see
7158 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7159 descriptors. However, the code omitted to check that the open succeeded,
7160 causing mysterious errors if for some reason the permissions on /dev/null
7161 got screwed. Now Exim writes a message to the main and panic logs, and
7162 bombs out if it can't open /dev/null.
7164 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7165 interact so that it is all more flexible. It is supposed to remain
7166 backwards compatible. Also added extra_local_interfaces.
7168 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7169 to bomb out with an assertion failure - to the client this appears as a
7170 connection drop. This problem occurs in the part of the code that was taken
7171 from the Samba project. Fortunately, the assertion is in a very simple
7172 function, so I have fixed this by reproducing the function inline in the
7173 one place where it is called, and arranging for authentication to fail
7174 instead of killing the process with assert().
7176 75. The SPA client code was not working when the server requested OEM rather
7177 than Unicode encoding.
7179 76. Added code to make require_files with a specific uid setting more usable in
7180 the case where statting the file as root fails - usually a non-root-mounted
7181 NFS file system. When this happens and the failure is EACCES, Exim now
7182 forks a subprocess and does the per-uid checking as the relevant uid.
7184 77. Added process_log_path.
7186 78. If log_file_path was not explicitly set, a setting of check_log_space or
7187 check_log_inodes was ignored.
7189 79. If a space check for the spool or log partitions fails, the incident is now
7190 logged. Of course, in the latter case the data may get lost...
7192 80. Added the %p formatting code to string_format() so that it can be used to
7193 print addresses in debug_print(). Adjusted all the address printing in the
7194 debugging in store.c to use %p rather than %d.
7196 81. There was a concern that a line of code in smtp_in.c could overflow a
7197 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7198 initially expressed, the concern was not well-founded, because trailing
7199 spaces are removed early. However, if the trailing spaces were followed by
7200 a NULL, they did not get removed, so the overflow was possible. Two fixes
7203 (a) I re-wrote the offending code in a cleaner fashion.
7204 (b) If an incoming SMTP command contains a NULL character, it is rejected
7207 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7208 runs initgroups(), so that if the Exim user is in any additional groups,
7209 they will be used during message reception.
7215 The change log for 4.20 and earlier releases has been archived.