1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
12 non-signal-safe functions being used.
14 JH/03 Bug 2269: When presented with a received message having a stupidly large
15 number of DKIM-Signature headers, disable DKIM verification to avoid
16 a resource-consumption attack. The limit is set at twenty.
18 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
19 files in the spool were present and unlocked. A queue-runner could spot
20 them, resulting in a duplicate delivery. Fix that by doing the unlock
21 after the unlink. Investigation by Tim Stewart.
23 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
24 a queue-runner could start a delivery while other operations were ongoing.
25 Cutthrough delivery was a common victim, resulting in duplicate delivery.
26 Found and investigated by Tim Stewart. Fix by using the open message data
27 file handle rather than opening another, and not locally closing it (which
28 releases a lock) for that case, while creating the temporary .eml format
29 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
35 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
36 When redis_servers is set to a list of > 1 element, and the Redis servers
37 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
38 case of MOVED into a DEFER case instead, thus moving the query onto the
39 next server in the list. For a cluster of N elements, all N servers must
40 be defined in redis_servers.
42 GF/02 Catch and remove uninitialized value warning in exiqsumm
43 Check for existence of @ARGV before looking at $ARGV[0]
45 JH/01 Replace the store_release() internal interface with store_newblock(),
46 which internalises the check required to safely use the old one, plus
47 the allocate and data copy operations duplicated in both (!) of the
50 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
51 modifier. This matches the restriction on the commandline.
53 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
54 Previously only the last row was returned.
56 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
57 we assumed that tags in the header were well-formed, and parsed the
58 element content after inspecting only the first char of the tag.
59 Assumptions at that stage could crash the receive process on malformed
62 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
63 While running the DKIM ACL we operate on the Permanent memory pool so that
64 variables created with "set" persist to the DATA ACL. Also (at any time)
65 DNS lookups that fail create cache records using the Permanent pool. But
66 expansions release any allocations made on the current pool - so a dnsdb
67 lookup expansion done in the DKIM ACL releases the memory used for the
68 DNS negative-cache, and bad things result. Solution is to switch to the
69 Main pool for expansions.
70 While we're in that code, add checks on the DNS cache during store_reset,
71 active in the testsuite.
72 Problem spotted, and debugging aided, by Wolfgang Breyha.
74 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
75 When none of the hosts presented to a transport match an already-open
76 connection, close it and proceed with the list. Previously we would
77 queue the message. Spotted by Lena with Yahoo, probably involving
80 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
81 Previously a spurious "250 OK id=" response was appended to the proper
84 JH/08 The "support for" informational output now, which built with Content
85 Scanning support, has a line for the malware scanner interfaces compiled
86 in. Interface can be individually included or not at build time.
88 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
89 by the template makefile "src/EDITME". The "STREAM" support for an older
90 ClamAV interface method is removed.
92 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
93 rows affected is given instead).
95 JH/11 The runtime Berkeley DB library version is now additionally output by
96 "exim -d -bV". Previously only the compile-time version was shown.
98 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
99 SMTP connection. Previously, when one had more receipients than the
100 first, an abortive onward connection was made. Move to full support for
101 multiple onward connections in sequence, handling cutthrough connection
102 for all multi-message initiating connections.
104 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
105 routers. Previously, a multi-recipient message would fail to match the
106 onward-connection opened for the first recipient, and cause its closure.
108 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
109 a timeout on read on a GnuTLS initiating connection, resulting in the
110 initiating connection being dropped. This mattered most when the callout
111 was marked defer_ok. Fix to keep the two timeout-detection methods
114 JH/15 Relax results from ACL control request to enable cutthrough, in
115 unsupported situations, from error to silently (except under debug)
116 ignoring. This covers use with PRDR, frozen messages, queue-only and
119 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
121 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
122 metadata, resulting in a crash in free().
124 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
125 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
126 Broken also in d185889f4, with init system revamp.
128 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
129 Previously we abruptly closed the connection after reading a malware-
130 found indication; now we go on to read the "scan ok" response line,
133 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
134 ACL. Previously, a crash would result.
136 JH/19 Speed up macro lookups during configuration file read, by skipping non-
137 macro text after a replacement (previously it was only once per line) and
138 by skipping builtin macros when searching for an uppercase lead character.
140 JH/20 DANE support moved from Experimental to mainline. The Makefile control
141 for the build is renamed.
143 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
144 was allocated for every new TLS startup, meaning one per message. Fix
145 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
147 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
148 reported the original. Fix to report (as far as possible) the ACL
149 result replacing the original.
151 JH/23 Fix memory leak during multi-message connections using STARTTLS under
152 OpenSSL. Certificate information is loaded for every new TLS startup,
153 and the resources needed to be freed.
155 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
157 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
160 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
161 DATA response info to the (existing) per-recipient response info for
162 the "C=" log element. It can have useful tracking info from the
163 destination system. Patch from Simon Arlott.
165 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
166 length value. Previously this would segfault.
168 HS/02 Support Avast multiline protoocol, this allows passing flags to
169 newer versions of the scanner.
171 JH/28 Ensure that variables possibly set during message acceptance are marked
172 dead before release of memory in the daemon loop. This stops complaints
173 about them when the debug_store option is enabled. Discovered specifically
174 for sender_rate_period, but applies to a whole set of variables.
175 Do the same for the queue-runner and queue-list loops, for variables set
176 from spool message files. Do the same for the SMTP per-message loop, for
177 certain variables indirectly set in ACL operations.
179 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
180 as a multi-recipient message from a mailinglist manager). The coding had
181 an arbitrary cutoff number of characters while checking for more input;
182 enforced by writing a NUL into the buffer. This corrupted long / fast
183 input. The problem was exposed more widely when more pipelineing of SMTP
184 responses was introduced, and one Exim system was feeding another.
185 The symptom is log complaints of SMTP syntax error (NUL chars) on the
186 receiving system, and refused recipients seen by the sending system
187 (propating to people being dropped from mailing lists).
188 Discovered and pinpointed by David Carter.
190 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
191 replaced by the ${authresults } expansion.
193 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
195 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
196 allows proper process termination in container environments.
198 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
199 Previously the "final dot" had a newline after it; ensure it is CR,LF.
201 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
202 and "err_perm", deprecated since 4.83 when the RFC-defined words
203 "temperror" and "permerror" were introduced.
205 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
206 transport-filters or DKIM-signing. The restriction was lost in the
207 consolidation of verify-callout and delivery SMTP handling.
208 Extend the restriction to also cover ARC-signing.
210 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
211 in defer=pass mode supply a 450 to the initiator. Previously the message
214 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
215 tls_require_ciphers is used as before.
217 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
218 "pass_unscanned". Only tmpfails from the scanner are written to
219 the paniclog, as they may require admin intervention (permission
220 denied, license issues). Other scanner errors (like decompression
221 bombs) do not cause a paniclog entry.
223 JH/36 Fix reinitialisation of DKIM logging variable between messages.
224 Previously it was possible to log spurious information in receive log
227 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
228 triggered odd behaviour from Outlook Express clients.
230 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
233 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
234 since the IETF WG has not yet settled on that versus the original
235 "bare" representation.
237 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
238 Previously the millisecond value corrupted the output.
239 Fix also for syslog_pid=no and log_selector +pid, for which the pid
240 corrupted the output.
246 JH/01 Rework error string handling in TLS interface so that the caller in
247 more cases is responsible for logging. This permits library-sourced
248 string to be attached to addresses during delivery, and collapses
249 pairs of long lines into single ones.
251 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
252 during configuration. Wildcards are allowed and expanded.
254 JH/02 Rework error string handling in DKIM to pass more info back to callers.
255 This permits better logging.
257 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
258 do not close it down and have the child transport start it up again on
259 the passed-on TCP connection. Instead, proxy the child (and any
260 subsequent ones) for TLS via a unix-domain socket channel. Logging is
261 affected: the continued delivery log lines do not have any DNSSEC, TLS
262 Certificate or OCSP information. TLS cipher information is still logged.
264 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
265 identical IP addresses on different listening ports. Will also affect
268 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
269 add noisy ifdef guards to special-case this sillyness.
270 Patch from Bernd Kuhls.
272 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
273 than 255 are no longer allowed.
275 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
276 Disable the session-cache too, which might reduce our load. Since we
277 currrectly use a new context for every connection, both as server and
278 client, there is no benefit for these.
279 GnuTLS appears to not support tickets server-side by default (we don't
280 call gnutls_session_ticket_enable_server()) but client side is enabled
281 by default on recent versions (3.1.3 +) unless the PFS priority string
284 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
285 <https://reproducible-builds.org/specs/source-date-epoch/>.
287 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
288 the check for any unsuccessful recipients did not notice the limit, and
289 erroneously found still-pending ones.
291 JH/08 Pipeline CHUNKING command and data together, on kernels that support
292 MSG_MORE. Only in-clear (not on TLS connections).
294 JH/09 Avoid using a temporary file during transport using dkim. Unless a
295 transport-filter is involved we can buffer the headers in memory for
296 creating the signature, and read the spool data file once for the
297 signature and again for transmission.
299 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
300 4.77 as the kernel support then wasn't solid, having issues in 64bit
301 mode. Now, it's been long enough. Add support for FreeBSD also.
303 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
304 case where the routing stage had gathered several addresses to send to
305 a host before calling the transport for the first, we previously failed
306 to close down TLS in the old transport process before passing the TCP
307 connection to the new process. The new one sent a STARTTLS command
308 which naturally failed, giving a failed delivery and bloating the retry
309 database. Investigation and fix prototype from Wolfgang Breyha.
311 JH/12 Fix check on SMTP command input synchronisation. Previously there were
312 false-negatives in the check that the sender had not preempted a response
313 or prompt from Exim (running as a server), due to that code's lack of
314 awareness of the SMTP input buffering.
316 PP/04 Add commandline_checks_require_admin option.
317 Exim drops privileges sanely, various checks such as -be aren't a
318 security problem, as long as you trust local users with access to their
319 own account. When invoked by services which pass untrusted data to
320 Exim, this might be an issue. Set this option in main configuration
321 AND make fixes to the calling application, such as using `--` to stop
324 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
325 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
326 responses to those, into a single TLS record each way (this usually means
327 a single packet). As a side issue, smtp_enforce_sync now works on TLS
330 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
331 affects you only if you're dancing at the edge of the param size limits.
332 If you are, and this message makes sense to you, then: raise the
333 configured limit or use OpenSSL 1.1. Nothing we can do for older
336 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
337 cmdline element to get the documented default one. Previously it was
340 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
341 are unsuable for non-message-specific reasons - eg. connection timeout,
342 banner-time rejection.
344 JH/16 Fix logging of delivery remote port, when specified by router, under
347 PP/06 Repair manualroute's ability to take options in any order, even if one
348 is the name of a transport.
351 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
353 JH/17 Change the list-building routines interface to use the expanding-string
354 triplet model, for better allocation and copying behaviour.
356 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
357 Previously it was constructed the first time a possibly-matching string
358 was met in the configuration file input during startup; now it is done
361 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
362 compatible one, to avoid the (poorly documented) possibility of a config
363 file in the working directory redirecting the DB files, possibly correpting
364 some existing file. CVE-2017-10140 assigned for BDB.
366 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
367 cache-hot. Previously, although the result was properly cached, the
368 initial verify call returned a defer.
370 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
371 the main verify for receipient in uncached-mode.
373 JH/22 Retire historical build files to an "unsupported" subdir. These are
374 defined as "ones for which we have no current evidence of testing".
376 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
377 if present. Previously it was ignored.
379 JH/24 Start using specified-initialisers in C structure init coding. This is
380 a C99 feature (it's 2017, so now considered safe).
382 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
383 if was a fixed-sized field and bitmask ops via macros; it is now more
386 PP/07 GitHub PR 56: Apply MariaDB build fix.
387 Patch provided by Jaroslav Škarvada.
389 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
390 during Coverity cleanups [4.87 JH/47]
391 Diagnosis and fix provided by Michael Fischer v. Mollard.
393 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
394 the right size to place the terminating semicolon on its own folded
395 line, the header hash was calculated to an incorrect value thanks to
396 the (relaxed) space the fold became.
398 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
399 and confused the parent.
401 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
402 which could crash as a result. This could lead to undeliverable messages.
404 JH/28 Logging: "next input sent too soon" now shows where input was truncated
407 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
408 matters on fast-turnover and PID-randomising systems, which were getting
409 out-of-order delivery.
411 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
412 a possibly-overlapping copy. The symptom was that "Remote host closed
413 connection in response to HELO" was logged instead of the actual 4xx
416 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
417 Previously only that bufferd was discarded, resulting in SYMTP command
420 JH/32 DKIM: when a message has multiple signatures matching an identity given
421 in dkim_verify_signers, run the dkim acl once for each. Previously only
422 one run was done. Bug 2189.
424 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
425 "panic the current process" to "deliberately defer". The panic log is
426 still written with the problem list name; the mail and reject logs now
427 get a temp-reject line for the message that was being handled, saying
428 something like "domains check lookup or other defer". The SMTP 451
429 message is still "Temporary local problem".
431 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
432 A crafted sequence of BDAT commands could result in in-use memory beeing
433 freed. CVE-2017-16943.
435 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
436 from SMTP input. Previously it was always done; now only done for DATA
437 and not BDAT commands. CVE-2017-16944.
439 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
440 to the message (such as an overlong header line). Previously this was
441 not done and we did not exit BDAT mode. Followon from the previous item
442 though a different problem.
448 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
449 than -2003 did; needs libidn2 in addition to libidn.
451 JH/02 The path option on a pipe transport is now expanded before use.
453 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
454 Patch provided by "Björn", documentation fix added too.
456 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
457 missing a wire-to-host endian conversion.
459 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
460 close after a BDAT command line could be taken as a following command,
461 giving a synch failure. Fix by only checking for synch immediately
462 before acknowledging the chunk.
464 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
465 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
466 macro. Patches provided by Josh Soref.
468 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
469 Previously we did not; the RFC seems ambiguous and VRFY is not listed
470 by IANA as a service extension. However, John Klensin suggests that we
473 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
474 the dkim code may be unix-mode line endings rather than smtp wire-format
475 CRLF, so prepend a CR to any bare LF.
477 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
478 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
480 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
481 the first were themselves being wrongly included in the feed into dkim
482 processing; with most chunk sizes in use this resulted in an incorrect
483 body hash calculated value.
485 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
486 DKIM signature block, for verification. Although advised against by
487 standards it is specifically not ruled illegal.
489 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
491 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
492 missing a body hash (the bh= tag).
494 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
495 It seems that HAProxy sends the Proxy Protocol information in clear and
496 only then does a TLS startup, so do the same.
498 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
499 TCP connections (such as for Spamd) unless the daemon successfully set
500 Fast Open mode on its listening sockets. This fixes breakage seen on
501 too-old kernels or those not configured for Fast Open, at the cost of
502 requiring both directions being enabled for TFO, and TFO never being used
503 by non-daemon-related Exim processes.
505 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
506 endings, at least on the first header line. Try to canonify any that get
507 past that check, despite the cost.
509 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
510 now limited to an arbitrary five deep, while parsing addresses with the
511 strip_excess_angle_brackets option enabled.
513 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
514 instead leave the unprompted TLS handshake in socket buffer for the
515 TLS library to consume.
517 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
519 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
521 JH/16 Drop variables when they go out of scope. Memory management drops a whole
522 region in one operation, for speed, and this leaves assigned pointers
523 dangling. Add checks run only under the testsuite which checks all
524 variables at a store-reset and panics on a dangling pointer; add code
525 explicitly nulling out all the variables discovered. Fixes one known
526 bug: a transport crash, where a dangling pointer for $sending_ip_address
527 originally assigned in a verify callout, is re-used.
529 PP/06 Drop '.' from @INC in various Perl scripts.
531 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
533 PP/08 Reduce a number of compilation warnings under clang; building with
534 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
535 should be warning-free.
537 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
539 HS/01 Fix portability problems introduced by PP/08 for platforms where
540 realloc(NULL) is not equivalent to malloc() [SunOS et al].
542 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
543 chunk. This allows us to accept broken chunked messages. We need a more
544 general solution here.
546 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
547 already-broken messages in the queue.
549 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
551 JH/19 Fix reference counting bug in routing-generated-address tracking.
557 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
558 supports it and a size is available (ie. the sending peer gave us one).
560 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
561 years of being deprecated). The replacements are the ACLs
562 acl_smtp_mime and acl_not_smtp_mime.
564 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
565 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
566 or even an in-clear connection were permitted. Now, if the host lookup
567 was dnssec and dane was requested then the host is only used if the
568 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
569 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
570 if one fails this test.
571 This means that a poorly-configured remote DNS will make it incommunicado;
572 but it protects against a DNS-interception attack on it.
574 JH/04 Bug 1810: make continued-use of an open smtp transport connection
575 non-noisy when a race steals the message being considered.
577 JH/05 If main configuration option tls_certificate is unset, generate a
578 self-signed certificate for inbound TLS connections.
580 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
581 in rewrites and routers.
583 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
584 and logged a warning sing 4.83; now they are a configuration file error.
586 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
587 (lacking @domain). Apply the same qualification processing as RCPT.
589 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
591 JH/10 Support ${sha256:} applied to a string (as well as the previous
594 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
595 a cutthrough deliver is pending, as we always want to make a connection.
596 This also avoids re-routing the message when later placing the cutthrough
597 connection after a verify cache hit.
598 Do not update it with the verify result either.
600 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
601 when routing results in more than one destination address.
603 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
604 signing (which inhibits the cutthrough capability). Previously only
605 the presence of an option was tested; now an expansion evaluating as
606 empty is permissible (obviously it should depend only on data available
607 when the cutthrough connection is made).
609 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
610 the relevant preceding SMTP command did not note the pipelining mode.
612 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
613 Previously they were not counted.
615 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
616 as one having no matching records. Previously we deferred the message
617 that needed the lookup.
619 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
620 distinguished as "(=".
622 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
623 for missing MX records. Previously it only worked for missing A records.
625 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
627 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
628 after the data-go-ahead and data-ack. Patch from Jason Betts.
630 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
631 even for a "none" policy. Patch from Tony Meyer.
633 JH/22 Fix continued use of a connection for further deliveries. If a port was
634 specified by a router, it must also match for the delivery to be
637 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
638 When one of the recipients of a message was unsuitable for the connection
639 (has no matching addresses), we lost track of needing to mark it
640 deferred. As a result mail would be lost.
642 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
644 JH/25 Decoding ACL controls is now done using a binary search; the source code
645 takes up less space and should be simpler to maintain. Merge the ACL
646 condition decode tables also, with similar effect.
648 JH/26 Fix problem with one_time used on a redirect router which returned the
649 parent address unchanged. A retry would see the parent address marked as
650 delivered, so not attempt the (identical) child. As a result mail would
653 JH/27 Fix a possible security hole, wherein a process operating with the Exim
654 UID can gain a root shell. Credit to http://www.halfdog.net/ for
655 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
658 JH/28 Enable {spool,log} filesystem space and inode checks as default.
659 Main config options check_{log,spool}_{inodes,space} are now
660 100 inodes, 10MB unless set otherwise in the configuration.
662 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
663 Previously it only applied to the main-section connection policy
666 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
668 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
669 by me. Added RFC7919 DH primes as an alternative.
671 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
672 are not in the system include path.
674 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
675 GnuTLS, when a session startup failed (eg because the client disconnected)
676 Exim did stdio operations after fclose. This was exposed by a recent
677 change which nulled out the file handle after the fclose.
679 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
680 signed directly by the cert-signing cert, rather than an intermediate
681 OCSP-signing cert. This is the model used by LetsEncrypt.
683 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
685 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
686 an incoming connection.
688 HS/02 Bug 1802: Do not half-close the connection after sending a request
691 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
692 fallback to "prime256v1".
694 JH/34 SECURITY: Use proper copy of DATA command in error message.
695 Could leak key material. Remotely exploitable. CVE-2016-9963.
701 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
702 and 3.4.4 - once the server is enabled to respond to an OCSP request
703 it does even when not requested, resulting in a stapling non-aware
704 client dropping the TLS connection.
706 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
707 support variable-length bit vectors. No functional change.
709 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
710 The I= interface field on outgoing lines is now after the H= remote
711 host field, same as incoming lines. There is a separate
712 outgoing_interface log selector which allows you to disable the
715 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
716 If not running log_selector +smtp_connection the mainlog would be held
717 open indefinitely after a "too many connections" event, including to a
718 deleted file after a log rotate. Leave the per net connection logging
719 leaving it open for efficiency as that will be quickly detected by the
720 check on the next write.
722 HS/01 Bug 1671: Fix post transport crash.
723 Processing the wait-<transport> messages could crash the delivery
724 process if the message IDs didn't exist for some reason. When
725 using 'split_spool_directory=yes' the construction of the spool
726 file name failed already, exposing the same netto behaviour.
728 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
729 mime_regex ACL conditions.
731 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
732 to DSN fail messages (bounces): remote IP, remote greeting, remote response
733 to HELO, local diagnostic string.
735 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
736 log line to debug. Even when configured with a tls authenticator many
737 client connections are expected to not authenticate in this way, so
738 an authenticate fail is not an error.
740 HS/02 Add the Exim version string to the process info. This way exiwhat
741 gives some more detail about the running daemon.
743 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
744 matter for fast-change records such as DNSBLs.
746 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
747 retry record, even if constant. There may be multiple transports with
748 different interface settings and the retry behaviour needs to be kept
751 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
753 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
755 JH/10 Bug 840: fix log_defer_output option of pipe transport
757 JH/11 Bug 830: use same host for all RCPTS of a message, even under
758 hosts_randomize. This matters a lot when combined with mua_wrapper.
760 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
761 ${quote_pgsql:<string>} operator.
763 JH/13 Bug 1708: avoid misaligned access in cached lookup.
765 JH/14 Change header file name for freeradius-client. Relevant if compiling
766 with Radius support; from the Gentoo tree and checked under Fedora.
768 JH/15 Bug 1712: Introduce $prdr_requested flag variable
770 JH/16 Bug 1714: Permit an empty string as expansion result for transport
771 option transport_filter, meaning no filtering.
773 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
775 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
776 defaults to "*" (all hosts). The variable is now available when not built
777 with TLS, default unset, mainly to enable keeping the testsuite sane.
778 If a server certificate is not supplied (via tls_certificate) an error is
779 logged, and clients will find TLS connections fail on startup. Presumably
780 they will retry in-clear.
781 Packagers of Exim are strongly encouraged to create a server certificate
782 at installation time.
784 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
785 with the $config_file variable.
787 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
788 in transport context, after the attempt, and per-recipient. The latter type
789 is per host attempted. The event data is the error message, and the errno
790 information encodes the lookup type (A vs. MX) used for the (first) host,
791 and the trailing two digits of the smtp 4xx response.
793 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
794 to write to mainlog (or rejectlog, paniclog) in the window between file
795 creation and permissions/ownership being changed. Particularly affects
796 installations where exicyclog is run as root, rather than exim user;
797 result is that the running daemon panics and dies.
799 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
801 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
802 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
803 "pri" and "weight". Note that the previous implicit priority given by the
804 list order is no longer honoured.
806 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
809 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
810 by defining SUPPORT_SOCKS.
812 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
813 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
814 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
815 variables are renamed to proxy_{local,external}_{address,port}.
817 JH/27 Move Internationalisation support from Experimental to mainline, enabled
818 for a build by defining SUPPORT_I18N
820 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
821 of the query string, and make ${quote_redis:} do that quoting.
823 JH/29 Move Events support from Experimental to mainline, enabled by default
824 and removable for a build by defining DISABLE_EVENT.
826 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
828 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
829 cached by the daemon.
831 JH/32 Move Redis support from Experimental to mainline, enabled for a build
832 by defining LOOKUP_REDIS. The libhiredis library is required.
834 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
835 keys are given for lookup.
837 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
838 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
839 only supported when built with TLS support. The PolarSSL SHA routines
840 are still used when the TLS library is too old for convenient support.
842 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
843 openssl_options), for security. OpenSSL forces this from version 1.1.0
844 server-side so match that on older versions.
846 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
847 allocation for $value could be released as the expansion processing
848 concluded, but leaving the global pointer active for it.
850 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
851 and to use the domains and local_parts ACL conditions.
853 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
854 incorrectly not doubled on cutthrough transmission, hence seen as a
855 body-termination at the receiving system - resulting in truncated mails.
856 Commonly the sender saw a TCP-level error, and retransmitted the message
857 via the normal store-and-forward channel. This could result in duplicates
858 received - but deduplicating mailstores were liable to retain only the
859 initial truncated version.
861 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
863 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
865 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
866 we're in there, support oversigning also; bug 1309.
868 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
870 HS/04 Add support for keep_environment and add_environment options.
872 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
873 either intentional arithmetic overflow during PRNG, or testing config-
876 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
877 delivery resulted in actual delivery. Cancel cutthrough before DATA
880 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
881 rejects a recipient: pass the reject to the originator.
883 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
884 Many were false-positives and ignorable, but it's worth fixing the
887 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
888 for the new environment-manipulation done at startup. Move the routines
889 from being local to tls.c to being global via the os.c file.
891 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
892 an extract embedded as result-arg for a map, the first arg for extract
893 is unavailable so we cannot tell if this is a numbered or keyed
894 extraction. Accept either.
900 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
903 JH/02 The smtp transport option "multi_domain" is now expanded.
905 JH/03 The smtp transport now requests PRDR by default, if the server offers
908 JH/04 Certificate name checking on server certificates, when exim is a client,
909 is now done by default. The transport option tls_verify_cert_hostnames
910 can be used to disable this per-host. The build option
911 EXPERIMENTAL_CERTNAMES is withdrawn.
913 JH/05 The value of the tls_verify_certificates smtp transport and main options
914 default to the word "system" to access the system default CA bundle.
915 For GnuTLS, only version 3.0.20 or later.
917 JH/06 Verification of the server certificate for a TLS connection is now tried
918 (but not required) by default. The verification status is now logged by
919 default, for both outbound TLS and client-certificate supplying inbound
922 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
925 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
926 Status Notification (bounce) messages are now MIME format per RFC 3464.
927 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
928 under the control of the dsn_advertise_hosts option, and routers may
929 have a dsn_lasthop option.
931 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
932 default, modifiable by a malware= option. The list separator for
933 the options can now be changed in the usual way. Bug 68.
935 JH/10 The smtp_receive_timeout main option is now expanded before use.
937 JH/11 The incoming_interface log option now also enables logging of the
938 local interface on delivery outgoing connections.
940 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
941 if the interface and destination host and port all match.
943 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
946 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
947 Patch from Andrew Lewis.
949 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
950 now supports optional time-restrictions, weighting, and priority
951 modifiers per server. Patch originally by <rommer@active.by>.
953 JH/16 The spamd_address main option now supports a mixed list of local
954 and remote servers. Remote servers can be IPv6 addresses, and
955 specify a port-range.
957 JH/17 Bug 68: The spamd_address main option now supports an optional
958 timeout value per server.
960 JH/18 Bug 1581: Router and transport options headers_add/remove can
961 now have the list separator specified.
963 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
966 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
969 JH/21 Support for the A6 type of dns record is withdrawn.
971 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
972 rather than the verbs used.
974 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
975 from 255 to 1024 chars.
977 JH/24 Verification callouts now attempt to use TLS by default.
979 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
980 are generic router options now. The defaults didn't change.
982 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
983 Original patch from Alexander Shikoff, worked over by JH.
985 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
986 files if ZCAT_COMMAND is not executable.
988 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
990 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
992 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
993 Normally benign, it bites when the pair was led to by a CNAME;
994 modern usage is to not canonicalize the domain to a CNAME target
995 (and we were inconsistent anyway for A-only vs AAAA+A).
997 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
999 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1000 when evaluating $sender_host_dnssec.
1002 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1003 $sender_helo_dnssec variable.
1005 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1007 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1009 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1011 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1012 documented as working, but never had. Support all but $spam_report.
1014 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1015 added for tls authenticator.
1017 HS/03 Add perl_taintmode main config option
1023 TL/01 When running the test suite, the README says that variables such as
1024 no_msglog_check are global and can be placed anywhere in a specific
1025 test's script, however it was observed that placement needed to be near
1026 the beginning for it to behave that way. Changed the runtest perl
1027 script to read through the entire script once to detect and set these
1028 variables, reset to the beginning of the script, and then run through
1029 the script parsing/test process like normal.
1031 TL/02 The BSD's have an arc4random API. One of the functions to induce
1032 adding randomness was arc4random_stir(), but it has been removed in
1033 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1034 function when detected.
1036 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1037 cause callback expansion.
1039 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1040 syntax errors in an expansion can be treated as a string instead of
1041 logging or causing an error, due to the internal use of bool_lax
1042 instead of bool when processing it.
1044 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1045 server certificates when making smtp deliveries.
1047 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1049 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1051 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1053 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1054 Merged patch from Sebastian Wiedenroth.
1056 JH/05 Fix results-pipe from transport process. Several recipients, combined
1057 with certificate use, exposed issues where response data items split
1058 over buffer boundaries were not parsed properly. This eventually
1059 resulted in duplicates being sent. This issue only became common enough
1060 to notice due to the introduction of connection certificate information,
1061 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1063 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1064 size buffer was used, resulting in syntax errors when an expansion
1067 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1068 version 3.3.6 or later.
1070 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1071 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1072 both become "event_action", the variables become $event_name, $event_data
1073 and $event_defer_errno. There is a new variable $verify_mode, usable in
1074 routers, transports and related events. The tls:cert event is now also
1075 raised for inbound connections, if the main configuration event_action
1078 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1079 early OCSP support, but no stapling (appears to be less than 1.0.0).
1081 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1082 server certificate names available under the smtp transport option
1083 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1086 JH/10 Time-related extraction expansions from certificates now use the main
1087 option "timezone" setting for output formatting, and are consistent
1088 between OpenSSL and GnuTLS compilations. Bug 1541.
1090 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1091 encoded parameter in the incoming message. Bug 1558.
1093 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1094 include certificate info, eximon was claiming there were spoolfile
1097 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1099 JH/14 Log delivery-related information more consistently, using the sequence
1100 "H=<name> [<ip>]" wherever possible.
1102 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1103 are problematic for Debian distribution, omit them from the release
1106 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1108 JH/16 Fix string representation of time values on 64bit time_t architectures.
1111 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1112 output list separator was used.
1117 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1118 checkers that were complaining about end of non-void function with no
1121 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1122 This was a regression introduced in 4.83 by another bugfix.
1124 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1126 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1127 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1133 TF/01 Correctly close the server side of TLS when forking for delivery.
1135 When a message was received over SMTP with TLS, Exim failed to clear up
1136 the incoming connection properly after forking off the child process to
1137 deliver the message. In some situations the subsequent outgoing
1138 delivery connection happened to have the same fd number as the incoming
1139 connection previously had. Exim would try to use TLS and fail, logging
1140 a "Bad file descriptor" error.
1142 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1143 utilities have not been installed.
1145 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1146 temporary space as the ACL may create new global variables.
1148 TL/01 LDAP support uses per connection or global context settings, depending
1149 upon the detected version of the libraries at build time.
1151 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1152 to extract and use the src ip:port in logging and expansions as if it
1153 were a direct connection from the outside internet. PPv2 support was
1154 updated based on HAProxy spec change in May 2014.
1156 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1158 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1159 Properly escape header and check for NULL return.
1161 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1164 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1166 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1167 characters in header names, implemented as a verify condition.
1168 Contributed by Michael Fischer v. Mollard.
1170 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1171 results permerror and temperror. Previous values are deprecated but
1172 still accepted. In a future release, err_perm and err_temp will be
1173 completely removed, which will be a backward incompatibility if the
1174 ACL tests for either of these two old results. Patch contributed by
1175 user bes-internal on the mailing list.
1177 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1179 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1180 selectors, in both main and reject logs.
1182 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1185 JH/07 Add malware type "sock" for talking to simple daemon.
1187 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1189 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1190 routers/transports under cutthrough routing.
1192 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1193 numbers. Touch up "bool" conditional to keep the same definition.
1195 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1197 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1199 JH/12 Expand items in router/transport headers_add or headers_remove lists
1200 individually rather than the list as a whole. Bug 1452.
1202 Required for reasonable handling of multiple headers_ options when
1203 they may be empty; requires that headers_remove items with embedded
1204 colons must have them doubled (or the list-separator changed).
1206 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1207 view the policy declared in the DMARC record. Currently, $dmarc_status
1208 is a combined value of both the record presence and the result of the
1211 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1213 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1214 dnslookup router and the smtp transport (applying to the forward
1217 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1218 of ldap servers used for a specific lookup. Patch provided by Heiko
1221 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1222 New variable $lookup_dnssec_authenticated for observability.
1224 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1225 Patch submitted by Lars Timman.
1227 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1229 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1230 Requires trusted mode and valid format message id, aborts otherwise.
1231 Patch contributed by Heiko Schlichting.
1233 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1234 certextract with support for various fields. Bug 1358.
1236 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1237 is requested by default, modifiable by smtp transport option
1240 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1241 operate on certificate variables to give certificate fingerprints
1242 Also new ${sha256:cert_variable}.
1244 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1246 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1247 Christian Aistleitner.
1249 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1251 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1252 file. Patch from Wolfgang Breyha.
1254 JH/25 Expand the coverage of the delivery $host and $host_address to
1255 client authenticators run in verify callout. Bug 1476.
1257 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1258 align with daemon_smtp_ports. Bug 72.
1260 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1261 support and error reporting did not work properly.
1263 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1264 and is readable. Patch from Andrew Colin Kissa.
1266 TL/14 Enhance documentation of ${run expansion and how it parses the
1267 commandline after expansion, particularly in the case when an
1268 unquoted variable expansion results in an empty value.
1270 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1272 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1273 by renaming away from it. Addresses GH issue 15, reported by
1276 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1278 TL/15 SECURITY: prevent double expansion in math comparison functions
1279 (can expand unsanitized data). Not remotely exploitable.
1286 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1288 PP/02 Make -n do something, by making it not do something.
1289 When combined with -bP, the name of an option is not output.
1291 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1294 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1295 $sender_host_name and config options to manage this, and basic check
1298 PP/05 DSCP support for outbound connections and control modifier for inbound.
1300 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1301 (Only plugin which currently uses this is kerberos4, which nobody should
1302 be using, but we should make it available and other future plugins might
1303 conceivably use it, even though it would break NAT; stuff *should* be
1304 using channel bindings instead).
1306 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1307 name; added for Sendmail compatibility; requires admin caller.
1308 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1309 just ignore it); requires trusted caller.
1310 Also parse but ignore: -Ac -Am -X<logfile>
1313 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1315 TL/02 Add +smtp_confirmation as a default logging option.
1317 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1318 Patch by Magnus Holmgren from 2007-02-20.
1320 TL/04 Bugzilla 1281 - Spec typo.
1321 Bugzilla 1283 - Spec typo.
1322 Bugzilla 1290 - Spec grammar fixes.
1324 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1326 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1328 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1329 mailing list by by Dmitry Isaikin.
1331 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1333 JH/02 Support "G" suffix to numbers in ${if comparisons.
1335 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1337 NM/01 Bugzilla 1197 - Spec typo
1338 Bugzilla 1196 - Spec examples corrections
1340 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1342 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1343 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1346 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1347 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1349 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1350 "acl {{name}{arg}...}", and optional args on acl condition
1353 JH/05 Permit multiple router/transport headers_add/remove lines.
1355 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1357 JH/07 Avoid using a waiting database for a single-message-only transport.
1358 Performance patch from Paul Fisher. Bugzilla 1262.
1360 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1363 JH/09 Add $headers_added variable, with content from use of ACL modifier
1364 add_header (but not yet added to the message). Bugzilla 199.
1366 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1367 Pulled from Bugzilla 817 by Wolfgang Breyha.
1369 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1371 (nb: this is the same fix as in Exim 4.80.1)
1373 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1376 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1379 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1381 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1382 advertise SMTP AUTH mechanism to us, instead of a generic
1383 protocol violation error. Also, make Exim more robust to bad
1384 data from the Dovecot auth socket.
1386 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1388 When a queue runner is handling a message, Exim first routes the
1389 recipient addresses, during which it prunes them based on the retry
1390 hints database. After that it attempts to deliver the message to
1391 any remaining recipients. It then updates the hints database using
1394 So if a recipient address works intermittently, it can get repeatedly
1395 deferred at routing time. The retry hints record remains fresh so the
1396 address never reaches the final cutoff time.
1398 This is a fairly common occurrence when a user is bumping up against
1399 their storage quota. Exim had some logic in its local delivery code
1400 to deal with this. However it did not apply to per-recipient defers
1401 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1403 This change adds a proper retry rule check during routing so that the
1404 final cutoff time is checked against the message's age. We only do
1405 this check if there is an address retry record and there is not a
1406 domain retry record; this implies that previous attempts to handle
1407 the address had the retry_use_local_parts option turned on. We use
1408 this as an approximation for the destination being like a local
1409 delivery, as in LMTP.
1411 I suspect this new check makes the old local delivery cutoff check
1412 redundant, but I have not verified this so I left the code in place.
1414 TF/02 Correct gecos expansion when From: is a prefix of the username.
1416 Test 0254 submits a message to Exim with the header
1420 When I ran the test suite under the user fanf2, Exim expanded
1421 the header to contain my full name, whereas it should have added
1422 a Resent-Sender: header. It erroneously treats any prefix of the
1423 username as equal to the username.
1425 This change corrects that bug.
1427 GF/01 DCC debug and logging tidyup
1428 Error conditions log to paniclog rather than rejectlog.
1429 Debug lines prefixed by "DCC: " to remove any ambiguity.
1431 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1433 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1434 Bug spotted by Jeremy Harris; was flawed since initial commit.
1435 Would have resulted in OCSP responses post-SNI triggering an Exim
1436 NULL dereference and crash.
1438 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1440 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1441 Bug detection, analysis and fix by Samuel Thibault.
1442 Bugzilla 1331, Debian bug #698092.
1444 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1446 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1447 Server implementation by Todd Lyons, client by JH.
1448 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1449 config variable "prdr_enable" controls whether the server
1450 advertises the facility. If the client requests PRDR a new
1451 acl_data_smtp_prdr ACL is called once for each recipient, after
1452 the body content is received and before the acl_smtp_data ACL.
1453 The client is controlled by both of: a hosts_try_prdr option
1454 on the smtp transport, and the server advertisement.
1455 Default client logging of deliveries and rejections involving
1456 PRDR are flagged with the string "PRDR".
1458 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1459 fclose(). Diagnosis by Todd Lyons.
1461 PP/17 Update configure.default to handle IPv6 localhost better.
1462 Patch by Alain Williams (plus minor tweaks).
1465 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1466 This is now consistent with GnuTLS, and is now documented: the
1467 previous undocumented portable approach to treating the option as
1468 unset was to force an expansion failure. That still works, and
1469 an empty string is now equivalent.
1471 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1472 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1473 not performing validation itself.
1475 PP/20 Added force_command boolean option to pipe transport.
1476 Patch from Nick Koston, of cPanel Inc.
1478 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1481 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1483 PP/21 Fix eximon continuous updating with timestamped log-files.
1484 Broken in a format-string cleanup in 4.80, missed when I repaired the
1485 other false fix of the same issue.
1486 Report and fix from Heiko Schlichting.
1489 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1490 Report from Prashanth Katuri.
1492 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1493 It's SecureTransport, so affects any MacOS clients which use the
1494 system-integrated TLS libraries, including email clients.
1496 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1497 using a MIME ACL for non-SMTP local injection.
1498 Report and assistance in diagnosis by Warren Baker.
1500 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1502 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1504 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1505 last id that failed so it may be referenced in subsequent ACL's.
1507 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1510 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1511 ldap library initialization, allowing self-signed CA's to be
1512 used. Also properly sets require_cert option later in code by
1513 using NULL (global ldap config) instead of ldap handle (per
1514 session). Bug diagnosis and testing by alxgomz.
1516 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1517 the src/util/ subdirectory.
1519 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1520 renamed to Transport Post Delivery Action by Jeremy Harris, as
1523 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1524 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1525 redis_servers = needs to be configured which will be used by the redis
1526 lookup. Patch from Warren Baker, of The Packet Hub.
1528 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1530 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1531 hostname or reverse DNS when processing a host list. Used suggestions
1532 from multiple comments on this bug.
1534 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1536 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1537 Missed a few lines, added it to make the runtest require no keyboard
1540 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1541 contains upper case chars. Make router use caseful_local_part.
1543 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1544 support when GnuTLS has been built with p11-kit.
1550 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1552 This, or similar/improved, will also be change PP/11 of 4.82.
1558 PP/01 Handle short writes when writing local log-files.
1559 In practice, only affects FreeBSD (8 onwards).
1560 Bugzilla 1053, with thanks to Dmitry Isaikin.
1562 NM/01 Bugzilla 949 - Documentation tweak
1564 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1567 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1569 PP/02 Implemented gsasl authenticator.
1571 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1573 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1574 `pkg-config foo` for cflags/libs.
1576 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1577 with rest of GSASL and with heimdal_gssapi.
1579 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1580 `pkg-config foo` for cflags/libs for the TLS implementation.
1582 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1583 properties get this fed in as external SSF. A number of robustness
1584 and debugging improvements to the cyrus_sasl authenticator.
1586 PP/08 cyrus_sasl server now expands the server_realm option.
1588 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1589 Patch by Jeremy Harris.
1591 PP/10 Added dbmjz lookup type.
1593 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1595 PP/12 MAIL args handles TAB as well as SP, for better interop with
1596 non-compliant senders.
1597 Analysis and variant patch by Todd Lyons.
1599 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1600 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1601 Patch from Dirk Mueller <dmueller@suse.com>
1603 PP/13 tls_peerdn now print-escaped for spool files.
1604 Observed some $tls_peerdn in wild which contained \n, which resulted
1605 in spool file corruption.
1607 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1608 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1609 or write after TLS renegotiation, which otherwise led to messages
1612 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1613 as a tracking header (ie: a signed header comes before the signature).
1614 Patch from Wolfgang Breyha.
1616 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1617 comma-sep list; embedded commas doubled.
1619 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1621 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1623 Report and patch from Dmitry Banschikov.
1625 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1626 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1627 failures. We appear to now support TLS1.1+ with Exim.
1629 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1630 lets Exim select keys and certificates based upon TLS SNI from client.
1631 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1632 before an outbound SMTP session. New log_selector, +tls_sni.
1634 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1635 NULL dereference. Report and patch from Alun Jones.
1637 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1638 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1639 Not seeing resolver debug output on NetBSD, but suspect this is a
1640 resolver implementation change.
1642 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1643 Left warnings. Added "eximon gdb" invocation mode.
1645 PP/21 Defaulting "accept_8bitmime" to true, not false.
1647 PP/22 Added -bw for inetd wait mode support.
1649 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1650 locate the relevant includes and libraries. Made this the default.
1652 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1653 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1655 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1656 This may cause build issues on older platforms.
1658 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1659 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1660 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1661 Added SNI support via GnuTLS too.
1662 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1664 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1666 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1667 Applied second patch from Janne, implementing suggestion to default
1668 multiple-strings-in-record handling to match SPF spec.
1670 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1672 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1673 read-only, out of scope).
1674 Patch from Wolfgang Breyha, report from Stuart Northfield.
1676 PP/29 Fix three issues highlighted by clang analyser static analysis.
1677 Only crash-plausible issue would require the Cambridge-specific
1678 iplookup router and a misconfiguration.
1679 Report from Marcin Mirosław.
1681 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1683 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1684 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1685 As part of this, removing so much warning spew let me fix some minor
1686 real issues in debug logging.
1688 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1689 assignment on my part. Fixed.
1691 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1692 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1693 Janne Snabb (who went above and beyond: thank you).
1695 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1696 string otherwise requires a connection and a bunch more work and it's
1697 relatively easy to get wrong. Should also expose TLS library linkage
1700 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1701 64-bit ${eval} (JH/03).
1703 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1704 GNU libc to support some of the 64-bit stuff, should not lead to
1705 conflicts. Defined before os.h is pulled in, so if a given platform
1706 needs to override this, it can.
1708 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1709 protection layer was required, which is not implemented.
1710 Bugzilla 1254, patch from Wolfgang Breyha.
1712 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1713 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1714 tls_dhparam take prime identifiers. Also unbreak combination of
1715 OpenSSL+DH_params+TLSSNI.
1717 PP/39 Disable SSLv2 by default in OpenSSL support.
1723 PP/01 Solaris build fix for Oracle's LDAP libraries.
1724 Bugzilla 1109, patch from Stephen Usher.
1726 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1728 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1731 TF/02 Fix a couple more cases where we did not log the error message
1732 when unlink() failed. See also change 4.74-TF/03.
1734 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1735 lock up or crash if it happened to be inside a call to libc when it
1736 got a SIGUSR1 from exiwhat.
1738 The SIGUSR1 handler appends the current process status to the process
1739 log which is later printed by exiwhat. It used to use the general
1740 purpose logging code to do this, but several functions it calls are
1741 not safe for signals.
1743 The new output code in the SIGUSR1 handler is specific to the process
1744 log, and simple enough that it's easy to inspect for signal safety.
1745 Removing some special cases also simplifies the general logging code.
1746 Removing the spurious timestamps from the process log simplifies
1749 TF/04 Improved ratelimit ACL condition.
1751 The /noupdate option has been deprecated in favour of /readonly which
1752 has clearer semantics. The /leaky, /strict, and /readonly update modes
1753 are mutually exclusive. The update mode is no longer included in the
1754 database key; it just determines when the database is updated. (This
1755 means that when you upgrade Exim will forget old rate measurements.)
1757 Exim now checks that the per_* options are used with an update mode that
1758 makes sense for the current ACL. For example, when Exim is processing a
1759 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1760 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1761 must specify per_mail/readonly. If you omit the update mode it defaults to
1762 /leaky where that makes sense (as before) or /readonly where required.
1764 The /noupdate option is now undocumented but still supported for
1765 backwards compatibility. It is equivalent to /readonly except that in
1766 ACLs where /readonly is required you may specify /leaky/noupdate or
1767 /strict/noupdate which are treated the same as /readonly.
1769 A useful new feature is the /count= option. This is a generalization
1770 of the per_byte option, so that you can measure the throughput of other
1771 aggregate values. For example, the per_byte option is now equivalent
1772 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1774 The per_rcpt option has been generalized using the /count= mechanism
1775 (though it's more complicated than the per_byte equivalence). When it is
1776 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1777 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1778 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1779 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1780 Note that using per_rcpt with a non-readonly update mode in more than
1781 one ACL will cause the recipients to be double-counted. (The per_mail
1782 and per_byte options don't have this problem.)
1784 The handling of very low rates has changed slightly. If the computed rate
1785 is less than the event's count (usually one) then this event is the first
1786 after a long gap. In this case the rate is set to the same as this event's
1787 count, so that the first message of a spam run is counted properly.
1789 The major new feature is a mechanism for counting the rate of unique
1790 events. The new per_addr option counts the number of different
1791 recipients that someone has sent messages to in the last time period. It
1792 behaves like per_rcpt if all the recipient addresses are different, but
1793 duplicate recipient addresses do not increase the measured rate. Like
1794 the /count= option this is a general mechanism, so the per_addr option
1795 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1796 example, measure the rate that a client uses different sender addresses
1797 with the options per_mail/unique=$sender_address. There are further
1798 details in the main documentation.
1800 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1802 TF/06 Removed a few PCRE remnants.
1804 TF/07 Automatically extract Exim's version number from tags in the git
1805 repository when doing development or release builds.
1807 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1808 Bugzilla 879. Patch from Paul Fisher.
1810 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1811 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1814 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1816 PP/05 Variable $av_failed, true if the AV scanner deferred.
1817 Bugzilla 1078. Patch from John Horne.
1819 PP/06 Stop make process more reliably on build failure.
1820 Bugzilla 1087. Patch from Heiko Schlittermann.
1822 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1823 Bugzilla 1089. Patch from Heiko Schlittermann.
1825 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1826 Bugzilla 1131. Patch from Holger Weiß.
1828 PP/09 Handle IPv6 addresses with SPF.
1829 Bugzilla 860. Patch from Wolfgang Breyha.
1831 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1833 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1836 PP/11 match_* no longer expand right-hand-side by default.
1837 New compile-time build option, EXPAND_LISTMATCH_RHS.
1838 New expansion conditions, "inlist", "inlisti".
1840 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1842 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1844 PP/14 fix log_write() format string regression from TF/03.
1845 Bugzilla 1152. Patch from Dmitry Isaikin.
1851 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1853 PP/02 Harmonised TLS library version reporting; only show if debugging.
1854 Layout now matches that introduced for other libraries in 4.74 PP/03.
1856 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1858 PP/04 New "dns_use_edns0" global option.
1860 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1863 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1864 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1866 TK/01 Updated PolarSSL code to 0.14.2.
1867 Bugzilla 1097. Patch from Andreas Metzler.
1869 PP/07 Catch divide-by-zero in ${eval:...}.
1870 Fixes bugzilla 1102.
1872 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1875 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1876 format-string attack -- SECURITY: remote arbitrary code execution.
1878 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1879 time unintentionally subject to list matching rules, letting the header
1880 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1881 arbitrary string expansion). This allowed for information disclosure.
1883 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1884 INT_MIN/-1 -- value coerced to INT_MAX.
1890 NM/01 Workaround for PCRE version dependency in version reporting
1893 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1894 This fixes portability to compilers other than gcc, notably
1895 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1897 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1898 makefiles for portability to HP-UX and POSIX correctness.
1900 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1901 Also via indented variable definition in the Makefile.
1902 (Debugging by Oliver Heesakkers).
1904 PP/02 Restore caching of spamd results with expanded spamd_address.
1905 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1907 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1908 Improves build reliability. Fix from: Frank Elsner
1910 NM/02 Fix wide character breakage in the rfc2047 coding
1911 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1913 NM/03 Allow underscore in dnslist lookups
1914 Fixes bug 1026. Patch from Graeme Fowler
1916 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1917 Code patches from Adam Ciarcinski of NetBSD.
1919 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1922 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1923 is logged, to avoid truncation. Patch from John Horne.
1925 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1926 Patch from Jakob Hirsch.
1928 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1929 SQL string expansion failure details.
1930 Patch from Andrey Oktyabrski.
1932 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1933 Patch from Simon Arlott.
1935 PP/09 New lookups functionality failed to compile on old gcc which rejects
1936 extern declarations in function scope.
1937 Patch from Oliver Fleischmann
1939 PP/10 Use sig_atomic_t for flags set from signal handlers.
1940 Check getgroups() return and improve debugging.
1941 Fixed developed for diagnosis in bug 927 (which turned out to be
1944 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1945 Patch from Mark Zealey.
1947 PP/12 Bugzilla 1056: Improved spamd server selection.
1948 Patch from Mark Zealey.
1950 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1951 Based on patch from Heiko Schlittermann.
1953 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1954 Patch from Uwe Doering, sign-off by Michael Haardt.
1956 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1957 variable declaration deep within a block. Bug and patch from
1960 PP/15 lookups-Makefile IRIX compatibility coercion.
1962 PP/16 Make DISABLE_DKIM build knob functional.
1964 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1965 Patch by Simon Arlott
1967 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1968 variable argument macros. Our copy now differs from upstream.
1974 TF/01 Failure to get a lock on a hints database can have serious
1975 consequences so log it to the panic log.
1977 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1978 controlled using the smtp_confirmation log selector.
1980 TF/03 Include the error message when we fail to unlink a spool file.
1982 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1983 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1984 for maintaining out-of-tree patches for some time.
1986 PP/01 Bugzilla 139: Documentation and portability issues.
1987 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1988 Handle per-OS dynamic-module compilation flags.
1990 PP/02 Let /dev/null have normal permissions.
1991 The 4.73 fixes were a little too stringent and complained about the
1992 permissions on /dev/null. Exempt it from some checks.
1993 Reported by Andreas M. Kirchwitz.
1995 PP/03 Report version information for many libraries, including
1996 Exim version information for dynamically loaded libraries. Created
1997 version.h, now support a version extension string for distributors
1998 who patch heavily. Dynamic module ABI change.
2000 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2001 privilege escalation vulnerability whereby the Exim run-time user
2002 can cause root to append content of the attacker's choosing to
2005 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2008 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2009 If dropping privileges for untrusted macros, we disabled normal logging
2010 on the basis that it would fail; for the Exim run-time user, this is not
2011 the case, and it resulted in successful deliveries going unlogged.
2012 Fixed. Reported by Andreas Metzler.
2018 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2019 only prepend for the Resent-* case. Fixes regression introduced in
2020 Exim 4.70 by NM/22 for Bugzilla 607.
2022 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2023 increasing numbers of administrators be bitten by this.
2025 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2027 PP/03 Bugzilla 994: added openssl_options main configuration option.
2029 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2031 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2033 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2035 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2036 without a peer certificate, leading to a segfault because of an
2037 assumption that peers always have certificates. Be a little more
2038 paranoid. Problem reported by Martin Tscholak.
2040 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2041 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2042 NB: ClamAV planning to remove STREAM in "middle of 2010".
2043 CL also introduces -bmalware, various -d+acl logging additions and
2044 more caution in buffer sizes.
2046 PP/09 Implemented reverse_ip expansion operator.
2048 PP/10 Bugzilla 937: provide a "debug" ACL control.
2050 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2052 PP/12 Bugzilla 973: Implement --version.
2054 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2056 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2058 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2060 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2061 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2062 ignore trailing whitespace.
2064 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2066 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2069 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2070 Notification from Dr Andrew Aitchison.
2072 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2073 ExtendedDetectionInfo response format.
2074 Notification from John Horne.
2076 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2079 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2080 XSL and documented dependency on system catalogs, with examples of how
2083 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2086 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2087 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2088 configuration file which is writeable by the Exim user or group.
2090 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2091 of configuration files to cover files specified with the -C option if
2092 they are going to be used with root privileges, not just the default
2095 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2096 option (effectively making it always true).
2098 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2099 files to be used while preserving root privileges.
2101 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2102 that rogue child processes cannot use them.
2104 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2105 run-time user, instead of root.
2107 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2108 Exim run-time user without dropping privileges.
2110 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2111 result string, instead of calling string_vformat() twice with the same
2114 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2115 for other users. Others should always drop root privileges if they use
2116 -C on the command line, even for a whitelisted configure file.
2118 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2120 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2126 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2127 $data_path, and $header_path variables; fixed documentation bugs and
2130 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2131 exipick to access non-standard spools, including the "frozen" queue
2134 NM/01 Bugzilla 965: Support mysql stored procedures.
2135 Patch from Alain Williams
2137 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2139 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2140 Patch from Andreas Metzler
2142 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2143 Patch from Kirill Miazine
2145 NM/05 Bugzilla 671: Added umask to procmail example.
2147 JJ/03 installed exipick 20100323.0, fixing doc bug
2149 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2150 directory. Notification and patch from Dan Rosenberg.
2152 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2154 TK/02 Improve log output when DKIM signing operation fails.
2156 MH/01 Treat the transport option dkim_domain as a colon separated
2157 list, not as a single string, and sign the message with each element,
2158 omitting multiple occurences of the same signer.
2160 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2161 Bugzilla 985, 986. Patch by Simon Arlott
2163 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2164 Patch by Simon Arlott
2166 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2167 MBX locking. Notification from Dan Rosenberg.
2173 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2175 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2177 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2179 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2181 NM/04 Bugzilla 915: Flag broken perl installation during build.
2187 TK/01 Added patch by Johannes Berg that expands the main option
2188 "spamd_address" if it starts with a dollar sign.
2190 TK/02 Write list of recipients to X-Envelope-Sender header when building
2191 the mbox-format spool file for content scanning (suggested by Jakob
2194 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2195 (http://www.dcc-servers.net/) support via dccifd. Activated by
2196 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2198 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2199 by Mark Daniel Reidel <mr@df.eu>.
2201 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2202 When building exim an external PCRE library is now needed -
2203 PCRE is a system library on the majority of modern systems.
2204 See entry on PCRE_LIBS in EDITME file.
2206 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2207 conversation. Added nologin parameter to request.
2208 Patch contributed by Kirill Miazine.
2210 TF/01 Do not log submission mode rewrites if they do not change the address.
2212 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2214 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2215 log files in place. Contributed by Roberto Lima.
2217 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2219 TF/03 Bugzilla 615: When checking the local_parts router precondition
2220 after a local_part_suffix or local_part_prefix option, Exim now
2221 does not use the address's named list lookup cache, since this
2222 contains cached lookups for the whole local part.
2224 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2225 Robert Millan. Documentation is in experimental-spec.txt.
2227 TF/04 Bugzilla 668: Fix parallel build (make -j).
2229 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2231 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2232 Patch provided by Jan Srzednicki.
2234 TF/05 Leading white space used to be stripped from $spam_report which
2235 wrecked the formatting. Now it is preserved.
2237 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2238 that they are available at delivery time.
2240 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2242 TF/08 TLS error reporting now respects the incoming_interface and
2243 incoming_port log selectors.
2245 TF/09 Produce a more useful error message if an SMTP transport's hosts
2246 setting expands to an empty string.
2248 NM/06 Bugzilla 744: EXPN did not work under TLS.
2249 Patch provided by Phil Pennock.
2251 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2252 Patch provided by Richard Godbee.
2254 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2255 acl_smtp_notquit, added index entry.
2257 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2258 Patch provided by Eugene Bujak.
2260 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2261 accept(). Patch provided by Maxim Dounin.
2263 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2264 Patch provided by Phil Pennock.
2266 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2268 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2269 Patch provided by Brad "anomie" Jorsch.
2271 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2272 Patch provided by Dean Brooks.
2274 TK/05 Add native DKIM support (does not depend on external libraries).
2276 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2277 Patch provided by Graeme Fowler.
2279 NM/16 Bugzilla 851: Documentation example syntax fix.
2281 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2283 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2286 NM/19 Bugzilla 745: TLS version reporting.
2287 Patch provided by Phil Pennock.
2289 NM/20 Bugzilla 167: bool: condition support.
2290 Patch provided by Phil Pennock.
2292 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2293 clients. Patch provided by Phil Pennock.
2295 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2296 Patch provided by Brad "anomie" Jorsch.
2298 NM/23 Bugzilla 687: Fix misparses in eximstats.
2299 Patch provided by Heiko Schlittermann.
2301 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2302 Patch provided by Heiko Schlittermann.
2304 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2305 plus update to original patch.
2307 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2309 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2310 Patch provided by David Brownlee.
2312 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2314 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2316 NM/30 Bugzilla 888: TLS documentation bugfixes.
2318 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2320 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2321 Unlike the original bugzilla I have changed all shell scripts in src tree.
2323 NM/33 Bugzilla 898: Transport filter timeout fix.
2324 Patch by Todd Rinaldo.
2326 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2327 Patch by Serge Demonchaux.
2329 NM/35 Bugzilla 39: Base64 decode bug fixes.
2330 Patch by Jakob Hirsch.
2332 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2334 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2336 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2338 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2344 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2345 ALT-N's libdkim that I have put here:
2346 http://duncanthrax.net/exim-experimental/
2348 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2349 were called 'true' and it seems that C99 defines that as a reserved
2350 keyword to be used with 'bool' variable types. That means you could
2351 not include C99-style headers which use bools without triggering
2352 build errors in sieve.c.
2354 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2355 as mailq or other aliases. Changed the --help handling significantly
2356 to do whats expected. exim_usage() emits usage/help information.
2358 SC/01 Added the -bylocaldomain option to eximstats.
2360 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2362 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2364 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2370 PH/01 Another patch from the Sieve maintainer.
2372 PH/02 When an IPv6 address is converted to a string for single-key lookup
2373 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2374 dots are used instead of colons so that keys in lsearch files need not
2375 contain colons. This was done some time before quoting was made available
2376 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2377 (notated using the quote facility) so as to distinguish them from IPv4
2378 keys. This meant that lookups for IP addresses in host lists did not work
2379 for iplsearch lookups.
2381 This has been fixed by arranging for IPv6 addresses to be expressed with
2382 colons if the lookup type is iplsearch. This is not incompatible, because
2383 previously such lookups could never work.
2385 The situation is now rather anomalous, since one *can* have colons in
2386 ordinary lsearch keys. However, making the change in all cases is
2387 incompatible and would probably break a number of configurations.
2389 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2392 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2393 conversion specification without a maximum field width, thereby enabling
2394 a rogue spamd server to cause a buffer overflow. While nobody in their
2395 right mind would setup Exim to query an untrusted spamd server, an
2396 attacker that gains access to a server running spamd could potentially
2397 exploit this vulnerability to run arbitrary code as the Exim user.
2399 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2400 $primary_hostname instead of what libspf2 thinks the hosts name is.
2402 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2403 a directory entry by the name of the lookup key. Previously, if a
2404 symlink pointed to a non-existing file or a file in a directory that
2405 Exim lacked permissions to read, a lookup for a key matching that
2406 symlink would fail. Now it is enough that a matching directory entry
2407 exists, symlink or not. (Bugzilla 503.)
2409 PH/03 The body_linecount and body_zerocount variables are now exported in the
2412 PH/04 Added the $dnslist_matched variable.
2414 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2415 This means they are set thereafter only if the connection becomes
2418 PH/06 Added the client_condition to authenticators so that some can be skipped
2419 by clients under certain conditions.
2421 PH/07 The error message for a badly-placed control=no_multiline_responses left
2422 "_responses" off the end of the name.
2424 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2426 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2427 (without spaces) instead of just copying the configuration text.
2429 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2431 PH/11 Added $max_received_linelength.
2433 PH/12 Added +ignore_defer and +include_defer to host lists.
2435 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2436 way in which PCRE > 7.0 is built.
2438 PH/14 Implemented queue_only_load_latch.
2440 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2441 MAIL command. The effect was to mangle the value on 64-bit systems.
2443 PH/16 Another patch from the Sieve maintainer.
2445 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2447 PH/18 If a system quota error occurred while trying to create the file for
2448 a maildir delivery, the message "Mailbox is full" was not appended to the
2449 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2450 only to a quota excession during the actual writing of the file.
2452 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2453 characters?) which causes problems in log lines. The DN values are now
2454 passed through string_printing() before being added to log lines.
2456 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2457 and InterBase are left for another time.)
2459 PH/21 Added message_body_newlines option.
2461 PH/22 Guard against possible overflow in moan_check_errorcopy().
2463 PH/23 POSIX allows open() to be a macro; guard against that.
2465 PH/24 If the recipient of an error message contained an @ in the local part
2466 (suitably quoted, of course), incorrect values were put in $domain and
2467 $local_part during the evaluation of errors_copy.
2473 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2474 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2477 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2478 issue a MAIL command.
2480 PH/02 In an ACL statement such as
2482 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2484 if a client was not listed at all, or was listed with a value other than
2485 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2486 the condition was not true (as it should be), so access was not denied.
2487 The bug was that the ! inversion was incorrectly passed on to the second
2488 item. This has been fixed.
2490 PH/03 Added additional dnslists conditions == and =& which are different from
2491 = and & when the dns lookup returns more than one IP address.
2493 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2494 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2496 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2497 FSYNC, which compiles an option called disable_fsync that allows for
2498 bypassing fsync(). The documentation is heavily laced with warnings.
2500 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2502 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2503 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2504 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2505 including adding "make clean"; (3) Added -fPIC when compiling the test
2506 dynamically loaded module, to get rid of a warning.
2508 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2509 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2510 = 0s. The bug is otherwise harmless.
2512 PH/07 There was a bug in the dovecot authenticator such that the value of
2513 $auth1 could be overwritten, and so not correctly preserved, after a
2514 successful authentication. This usually meant that the value preserved by
2515 the server_setid option was incorrect.
2517 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2519 PH/09 Installed PCRE release 7.0.
2521 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2522 run for batched SMTP input. It is now run at the start of every message
2523 in the batch. While fixing this I discovered that the process information
2524 (output by running exiwhat) was not always getting set for -bs and -bS
2525 input. This is fixed, and it now also says "batched" for BSMTP.
2527 PH/11 Added control=no_pipelining.
2529 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2530 patch, slightly modified), and move the expansion of helo_data till after
2531 the connection is made in the smtp transport (so it can use these
2534 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2536 PH/14 Added log_selector = +pid.
2538 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2540 PH/16 Add ${if forany and ${if forall.
2542 PH/17 Added dsn_from option to vary the From: line in DSNs.
2544 PH/18 Flush SMTP output before performing a callout, unless control =
2545 no_callout_flush is set.
2547 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2548 was true (the default) a successful delivery failed to delete the retry
2549 item, thus causing premature timeout of the address. The bug is now
2552 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2554 PH/21 Long custom messages for fakedefer and fakereject are now split up
2555 into multiline responses in the same way that messages for "deny" and
2556 other ACL rejections are.
2558 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2559 with slight modification.
2561 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2562 draft, changing the syntax and factoring some duplicate code".
2564 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2565 for deliveries of the second and subsequent messages over the same SMTP
2568 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2569 ${reduce, with only minor "tidies".
2571 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2573 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2574 expansion side effects.
2576 PH/27 When a message times out after an over-quota error from an Exim-imposed
2577 quota, the bounce message says "mailbox is full". This message was not
2578 being given when it was a system quota that was exceeded. It now should
2581 MH/03 Made $recipients available in local_scan(). local_scan() already has
2582 better access to the recipient list through recipients_list[], but
2583 $recipients can be useful in postmaster-provided expansion strings.
2585 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2586 in the case of a MAIL command with additional options following the
2587 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2588 were accidentally chopped off.
2590 PH/29 SMTP synchronization checks are implemented when a command is read -
2591 there is a check that no more input is waiting when there shouldn't be
2592 any. However, for some commands, a delay in an ACL can mean that it is
2593 some time before the response is written. In this time, more input might
2594 arrive, invalidly. So now there are extra checks after an ACL has run for
2595 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2596 pipelining has not been advertised.
2598 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2600 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2601 returned in the SMTP response when smtp_return_error_details was false.
2602 This has been fixed.
2604 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2605 instead of the C I/O that was originally supplied, because problems were
2606 reported on Solaris.
2608 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2609 Exim which did not show up earlier: it was assuming that a call to
2610 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2611 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2612 was a macro that became an assignment, so it seemed to work. This has
2613 changed to a proper function call with a void return, hence the compile
2614 error. Exim's code has been fixed.
2616 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2619 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2621 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2623 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2624 because it turns out to be too big or there is a timeout) there is no
2625 "Completed" line in the log. When some messages of this type were
2626 selected by exigrep, they were listed as "not completed". Others were
2627 picked up by some special patterns. I have improved the selection
2628 criteria to be more general.
2630 PH/38 The host_find_failed option in the manualroute router can now be set
2631 to "ignore", to completely ignore a host whose IP address cannot be
2632 found. If all hosts are ignored, the behaviour is controlled by the new
2633 host_all_ignored option.
2635 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2636 homing or because of multiple MX records with /mx) generated more than
2637 one IP address, and the following item turned out to be the local host,
2638 all the secondary addresses of the first item were incorrectly removed
2639 from the list, along with the local host and any following hosts (which
2640 is what is supposed to happen).
2642 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2643 whoever called Exim into the -H file. In the case of the daemon it was
2644 behaving confusingly. When first started, it used values for whoever
2645 started the daemon, but after a SIGHUP it used the Exim user (because it
2646 calls itself on a restart). I have changed the code so that it now always
2649 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2650 message are rejected with the same error (e.g. no authentication or bad
2651 sender address), and a DATA command is nevertheless sent (as can happen
2652 with PIPELINING or a stupid MUA), the error message that was given to the
2653 RCPT commands is included in the rejection of the DATA command. This is
2654 intended to be helpful for MUAs that show only the final error to their
2657 PH/42 Another patch from the Sieve maintainer.
2659 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2660 Eximstats - Fixed some broken HTML links and added missing column headers
2662 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2663 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2665 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2667 PH/43 Yet another patch from the Sieve maintainer.
2669 PH/44 I found a way to check for a TCP/IP connection going away before sending
2670 the response to the final '.' that terminates a message, but only in the
2671 case where the client has not sent further data following the '.'
2672 (unfortunately, this is allowed). However, in many cases there won't be
2673 any further data because there won't be any more messages to send. A call
2674 to select() can be used: if it shows that the input is "ready", there is
2675 either input waiting, or the socket has been closed. An attempt to read
2676 the next input character can distinguish the two cases. Previously, Exim
2677 would have sent an OK response which the client would never have see.
2678 This could lead to message repetition. This fix should cure that, at
2679 least in a lot of common cases.
2681 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2682 advertised in response to EHLO.
2688 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2689 fixed by 4.65/MH/01 (is this a record?) are fixed:
2691 (i) An empty string was always treated as zero by the numeric comparison
2692 operators. This behaviour has been restored.
2694 (ii) It is documented that the numeric comparison operators always treat
2695 their arguments as decimal numbers. This was broken in that numbers
2696 starting with 0 were being interpreted as octal.
2698 While fixing these problems I realized that there was another issue that
2699 hadn't been noticed. Values of message_size_limit (both the global option
2700 and the transport option) were treated as octal if they started with 0.
2701 The documentation was vague. These values are now always treated as
2702 decimal, and I will make that clear in the documentation.
2708 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2709 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2712 MH/01 Don't check that the operands of numeric comparison operators are
2713 integers when their expansion is in "skipping" mode (fixes bug
2714 introduced by 4.64-PH/07).
2716 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2717 child addresses, Exim now panics and dies. Previously, because the count
2718 is held in a short int, deliveries were likely to be lost. As such a
2719 large number of recipients for a single message is ridiculous
2720 (performance will be very, very poor), I have chosen to impose a limit
2721 rather than extend the field.
2727 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2728 leftover -K file (the existence of which was triggered by #402).
2729 While we were at it, introduced process PID as part of the -K
2730 filename. This should rule out race conditions when creating
2733 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2734 processing considerably. Previous code took too long for large mails,
2735 triggering a timeout which in turn triggers #401.
2737 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2738 in the DK code in transports.c. sendfile() is not really portable,
2739 hence the _LINUX specificness.
2741 TF/01 In the add_headers option to the mail command in an Exim filter,
2742 there was a bug that Exim would claim a syntax error in any
2743 header after the first one which had an odd number of characters
2746 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2747 callout verification, Exim cached a "reject" for the entire domain. This
2748 is correct for most verifications, but it is not correct for a recipient
2749 verification with use_sender or use_postmaster set, because in that case
2750 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2751 case of MAIL FROM:<> rejection from other early rejections (e.g.
2752 rejection of HELO). When verifying a recipient using a non-null MAIL
2753 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2754 Whatever the result of the callout, the value of the domain cache is
2755 left unchanged (for any other kind of callout, getting as far as trying
2756 RCPT means that the domain itself is ok).
2758 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2761 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2762 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2763 session, the daemon ignores ECONNECT errors and logs others; it now
2764 ignores EPIPE as well.
2766 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2767 (quoted-printable decoding).
2769 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2770 later the small subsequent patch to fix an introduced bug.
2772 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2774 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2776 PH/08 An error is now given if message_size_limit is specified negative.
2778 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2779 to be given (somewhat) arbitrary names.
2781 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2784 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2785 miscellaneous code fixes
2787 PH/10 Added the log_reject_target ACL modifier to specify where to log
2790 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2791 hostname. This is wrong, because it relates to the incoming message (and
2792 probably the interface on which it is arriving) and not to the outgoing
2793 callout (which could be using a different interface). This has been
2794 changed to use the value of the helo_data option from the smtp transport
2795 instead - this is what is used when a message is actually being sent. If
2796 there is no remote transport (possible with a router that sets up host
2797 addresses), $smtp_active_hostname is used.
2799 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2800 tweaks were necessary in order to get it to work (see also 21 below):
2801 (a) The code assumed that strncpy() returns a negative number on buffer
2802 overflow, which isn't the case. Replaced with Exim's string_format()
2804 (b) There were several signed/unsigned issues. I just did the minimum
2805 hacking in of casts. There is scope for a larger refactoring.
2806 (c) The code used strcasecmp() which is not a standard C function.
2807 Replaced with Exim's strcmpic() function.
2808 (d) The code set only $1; it now sets $auth1 as well.
2809 (e) A simple test gave the error "authentication client didn't specify
2810 service in request". It would seem that Dovecot has changed its
2811 interface. Fortunately there's a specification; I followed it and
2812 changed what the client sends and it appears to be working now.
2814 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2817 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2818 address A is aliased to B and C, where B exists and C does not. Without
2819 -v the output is "A verified" because verification stops after a
2820 successful redirection if more than one address is generated. However,
2821 with -v the child addresses are also verified. Exim was outputting "A
2822 failed to verify" and then showing the successful verification for C,
2823 with its parentage. It now outputs "B failed to verify", showing B's
2824 parentage before showing the successful verification of C.
2826 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2827 look up a TXT record in a specific list after matching in a combined
2830 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2831 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2832 they consult the DNS. I had assumed they would set it the way they
2833 wanted; and indeed my experiments on Linux seem to show that in some
2834 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2835 To be on the safe side, however, I have now made the interface to
2836 host_find_byname() similar to host_find_bydns(), with an argument
2837 containing the DNS resolver options. The host_find_byname() function now
2838 sets these options at its start, just as host_find_bydns() does. The smtp
2839 transport options dns_qualify_single and dns_search_parents are passed to
2840 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2841 of host_find_byname() use the default settings of RES_DEFNAMES
2842 (qualify_single) but not RES_DNSRCH (search_parents).
2844 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2845 spool_read_header() do less string testing, by means of a preliminary
2846 switch on the second character of optional "-foo" lines. (This is
2847 overdue, caused by the large number of possibilities that now exist.
2848 Originally there were few.) While I was there, I also converted the
2849 str(n)cmp tests so they don't re-test the leading "-" and the first
2850 character, in the hope this might squeeze out yet more improvement.
2852 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2853 flag allowing group syntax was set by the header_syntax check but not
2854 turned off, possible causing trouble later; (2) The flag was not being
2855 set at all for the header_verify test, causing "group"-style headers to
2856 be rejected. I have now set it in this case, and also caused header_
2857 verify to ignore an empty address taken from a group. While doing this, I
2858 came across some other cases where the code for allowing group syntax
2859 while scanning a header line wasn't quite right (mostly, not resetting
2860 the flag correctly in the right place). These bugs could have caused
2861 trouble for malformed header lines. I hope it is now all correct.
2863 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2864 with the "reply" argument non-NULL. The code, however (which originally
2865 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2866 but it didn't always do it. This confused somebody who was copying the
2867 code for some other use. I have removed all the tests.
2869 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2870 feature that was used to support insecure browsers during the U.S. crypto
2871 embargo. It requires special client support, and Exim is probably the
2872 only MTA that supported it -- and would never use it because real RSA is
2873 always available. This code has been removed, because it had the bad
2874 effect of slowing Exim down by computing (never used) parameters for the
2875 RSA_EXPORT functionality.
2877 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2878 authenticator to fail if there's a tab character in the incoming data
2879 (there should never be unless someone is messing about, as it's supposed
2880 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2881 if the connection is using TLS or if the remote IP is the same as the
2882 local IP, and the "valid-client-cert option" if a client certificate has
2885 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2886 authenticators. This can be used for authorization after authentication
2887 succeeds. (In the case of plaintext, it servers for both authentication
2890 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2891 if any retry times were supplied.
2893 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2894 connection, where there is no client IP address to check. In this
2895 situation, the verify now always succeeds.
2897 PH/25 Applied John Jetmore's -Mset patch.
2899 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2901 PH/27 In a string expansion for a processed (not raw) header when multiple
2902 headers of the same name were present, leading whitespace was being
2903 removed from all of them, but trailing whitespace was being removed only
2904 from the last one. Now trailing whitespace is removed from each header
2905 before concatenation. Completely empty headers in a concatenation (as
2906 before) are ignored.
2908 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2909 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2911 PH/29 [Removed. This was a change that I later backed out, and forgot to
2912 correct the ChangeLog entry (that I had efficiently created) before
2913 committing the later change.]
2915 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2916 address errors (4xx response to RCPT) over the same connection as other
2917 messages routed to the same hosts. Such deliveries are always "forced",
2918 so retry times are not inspected. This resulted in far too many retries
2919 for the affected addresses. The effect occurred only when there were more
2920 hosts than the hosts_max_try setting in the smtp transport when it had
2921 the 4xx errors. Those hosts that it had tried were not added to the list
2922 of hosts for which the message was waiting, so if all were tried, there
2923 was no problem. Two fixes have been applied:
2925 (i) If there are any address or message errors in an SMTP delivery, none
2926 of the hosts (tried or untried) are now added to the list of hosts
2927 for which the message is waiting, so the message should not be a
2928 candidate for sending over the same connection that was used for a
2929 successful delivery of some other message. This seems entirely
2930 reasonable: after all the message is NOT "waiting for some host".
2931 This is so "obvious" that I'm not sure why it wasn't done
2932 previously. Hope I haven't missed anything, but it can't do any
2933 harm, as the worst effect is to miss an optimization.
2935 (ii) If, despite (i), such a delivery is accidentally attempted, the
2936 routing retry time is respected, so at least it doesn't keep
2937 hammering the server.
2939 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2940 in ${readsocket because some servers need this prod.
2942 PH/32 Added some extra debug output when updating a wait-xxx database.
2944 PH/33 The hint "could be header name not terminated by colon", which has been
2945 given for certain expansion errors for a long time, was not being given
2946 for the ${if def:h_colon_omitted{... case.
2948 PH/34 The spec says: "With one important exception, whenever a domain list is
2949 being scanned, $domain contains the subject domain." There was at least
2950 one case where this was not true.
2952 PH/35 The error "getsockname() failed: connection reset by peer" was being
2953 written to the panic log as well as the main log, but it isn't really
2954 panic-worthy as it just means the connection died rather early on. I have
2955 removed the panic log writing for the ECONNRESET error when getsockname()
2958 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2959 runs only) independently of the message's sender address. This meant
2960 that, if the 4xx error was in fact related to the sender, a different
2961 message to the same recipient with a different sender could confuse
2962 things. In particular, this can happen when sending to a greylisting
2963 server, but other circumstances could also provoke similar problems.
2964 I have changed the default so that the retry time for these errors is now
2965 based a combination of the sender and recipient addresses. This change
2966 can be overridden by setting address_retry_include_sender=false in the
2969 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2970 remote server are returned as part of bounce messages. This was not
2971 happening for LMTP over a pipe (the lmtp transport), but now it is the
2972 same for both kinds of LMTP.
2974 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2975 in header lines that were in fact CNAMEs. This is no longer the case.
2977 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2978 and queue runs started by the daemon processed all messages. This has
2979 been fixed so that -R and -S can now usefully be given with -q<time>.
2981 PH/40 Import PCRE release 6.7 (fixes some bugs).
2983 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2985 PH/42 Give an error if -q is specified more than once.
2987 PH/43 Renamed the variables $interface_address and $interface_port as
2988 $received_ip_address and $received_port, to make it clear that these
2989 values apply to message reception, and not to the outgoing interface when
2990 a message is delivered. (The old names remain recognized, of course.)
2992 PH/44 There was no timeout on the connect() call when using a Unix domain
2993 socket in the ${readsocket expansion. There now is.
2995 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2996 be meaningful with "accept".
2998 SC/01 Eximstats V1.43
2999 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3001 SC/02 Eximstats V1.44
3002 Use a glob alias rather than an array ref in the generated
3003 parser. This improves both readability and performance.
3005 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3006 Collect SpamAssassin and rejection statistics.
3007 Don't display local sender or destination tables unless
3008 there is data to show.
3009 Added average volumes into the top table text output.
3011 SC/04 Eximstats V1.46
3012 Collect data on the number of addresses (recipients)
3013 as well as the number of messages.
3015 SC/05 Eximstats V1.47
3016 Added 'Message too big' to the list of mail rejection
3017 reasons (thanks to Marco Gaiarin).
3019 SC/06 Eximstats V1.48
3020 Mainlog lines which have GMT offsets and are too short to
3021 have a flag are now skipped.
3023 SC/07 Eximstats V1.49 (Alain Williams)
3024 Added the -emptyok flag.
3026 SC/08 Eximstats V1.50
3027 Fixes for obtaining the IP address from reject messages.
3029 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3030 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3031 whitespace changes from 4.64-PH/27
3033 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3036 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3037 are found, allow negative numbers in numeric criteria)
3039 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3041 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3042 to match changes made in 4.64-PH/43
3044 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3046 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3047 "domain" is actually an IP address, and force a failure. This locks out
3048 those revolvers/nameservers that support "A-for-A" lookups, in
3049 contravention of the specifications.
3051 PH/48 When a host name was looked up from an IP address, and the subsequent
3052 forward lookup of the name timed out, the host name was left in
3053 $sender_host_name, contrary to the specification.
3055 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3056 restricted to single-key lookups, Exim was not diagnosing an error if
3057 * or *@ was used with a query-style lookup.
3059 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3061 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3062 long ago, but noone interested enough thought of it. Let's just say that
3063 the "1.1" means that there are some new functions that weren't there at
3064 some point in the past.
3066 PH/51 Error processing for expansion failure of helo_data from an smtp
3067 transport during callout processing was broken.
3069 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3070 tested/used via the -bh/-bhc/-bs options.
3072 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3073 bug, fixed in subsequent PCRE releases).
3075 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3076 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3078 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3084 SC/01 Use a glob alias rather than an array ref in eximstats generated
3085 parser. This improves both readability and performance.
3087 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3088 Don't display local sender or destination tables in eximstats unless
3089 there is data to show.
3090 Added average volumes into the eximstats top table text output.
3092 SC/03 Collect data on the number of addresses (recipients) as well
3093 as the number of messages in eximstats.
3095 TF/01 Correct an error in the documentation for the redirect router. Exim
3096 does (usually) call initgroups() when daemonizing.
3098 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3099 with consistent privilege compared to when running as a daemon.
3101 TF/03 Note in the spec that $authenticated_id is not set for local
3102 submissions from trusted users.
3104 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3105 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3107 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3108 by adding some example configuration directives to the default
3109 configuration file. A little bit of work is required to uncomment the
3110 directives and define how usernames and passwords are checked, but
3111 there is now a framework to start from.
3113 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3114 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3115 without this. I don't know how relevant this is to other LDAP libraries.
3117 PH/02 Add the verb name to the "unknown ACL verb" error.
3119 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3121 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3123 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3124 directory not expanded when it should be if an expanded home directory
3125 was set for the address (which is overridden by the transport).
3127 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3130 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3131 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3132 because it is too late at that time, and has no effect.
3134 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3135 security issue with \' (bugzilla #107). I could not use the
3136 PQescapeStringConn() function, because it needs a PGconn value as one of
3139 PH/08 When testing addresses using -bt, indicate those final addresses that
3140 are duplicates that would not cause an additional delivery. At least one
3141 person was confused, thinking that -bt output corresponded to deliveries.
3142 (Suppressing duplicates isn't a good idea as you lose the information
3143 about possibly different redirections that led to the duplicates.)
3145 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3146 systems where poll() doesn't work, in particular OS X.
3148 PH/10 Added more information to debugging output for retry time not reached.
3150 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3151 operations in malware.c.
3153 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3156 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3157 syslog_duplication set false, log lines that would normally be written
3158 both the the main log and to the reject log were not written to syslog at
3161 PH/14 In the default configuration, change the use of "message" in ACL warn
3162 statements to "add_header".
3164 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3165 not followed by a command (e.g. "seen endif").
3167 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3168 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3171 PH/17 Added extra conditions to the default value of delay_warning_condition
3175 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3176 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3177 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3180 The Auto-Submitted: and various List- headers are standardised, whereas I
3181 don't think Precedence: ever was.
3183 PH/18 Refactored debugging code in route_finduser() to show more information,
3184 in particular, the error code if getpwnam() issues one.
3186 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3187 This is apparently needed in addition to the PH/07 change above to avoid
3188 any possible encoding problems.
3190 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3191 but not after initializing Perl.
3193 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3194 output them only if debugging. By default they are written stderr,
3195 apparently, which is not desirable.
3197 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3200 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3203 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3205 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3206 authenticated or an ident call has been made. Suppress the default
3207 values for $authenticated_id and $authenticated_sender (but permit -oMai
3208 and -oMas) when testing with -bh.
3210 PH/24 Re-jigged the order of the tests in the default configuration so that the
3211 tests for valid domains and recipients precede the DNS black list and CSA
3212 tests, on the grounds that those ones are more expensive.
3214 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3215 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3216 This bug exists in every version of Exim that I still have, right back to
3219 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3220 However, an attempt to turn on case-sensitivity in a regex key by
3221 including (?-i) didn't work because the subject string was already
3222 lowercased, and the effects were non-intuitive. It turns out that a
3223 one-line patch can be used to allow (?-i) to work as expected.
3229 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3230 other effects) broke the use of negated acl sub-conditions.
3232 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3235 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3236 "Deny" causes Exim to reject the incoming connection with a 554 error.
3237 Unfortunately, if there is a major crisis, such as a disk failure,
3238 tcp-wrappers gives "deny", whereas what one would like would be some
3239 kind of temporary error. A kludge has been added to help with this.
3240 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3241 554 error is used if errno is still zero or contains ENOENT (which occurs
3242 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3245 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3247 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3248 errors. Otherwise a message that provokes a temporary error (when other
3249 messages do not) can cause a whole host to time out.
3251 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3252 addresses were routed directly to files or pipes from a redirect router.
3253 File deliveries just didn't batch; pipe deliveries might have suffered
3256 PH/06 A failure to get a lock for a hints database would erroneously always say
3257 "Failed to get write lock", even when it was really a read lock.
3259 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3260 of 0600. This has been changed to use the value of the lockfile_mode
3261 option (which defaults to 0600).
3263 PH/08 Applied small patch from the Sieve maintainer.
3265 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3266 folder from quota calculations, a direct delivery into this folder messed
3267 up the contents of the maildirsize file. This was because the regex was
3268 used only to exclude .Trash (or whatever) when the size of the mailbox
3269 was calculated. There was no check that a delivery was happening into an
3270 excluded directory. This bug has been fixed by ignoring all quota
3271 processing for deliveries into excluded directories.
3273 PH/10 Added the maildirfolder_create_regex option to appendfile.
3279 PH/01 The code for finding all the local interface addresses on a FreeBSD
3280 system running IPv6 was broken. This may well have applied to all BSD
3281 systems, as well as to others that have similar system calls. The broken
3282 code found IPv4 interfaces correctly, but gave incorrect values for the
3283 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3284 that it would not match correctly against @[] and not recognize the IPv6
3287 PH/02 The ipliteral router was not recognizing addresses of the form user@
3288 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3290 PH/03 Added disable_ipv6.
3292 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3293 decoded form, because it is most often used to construct To: headers
3294 lines in autoreplies, and the decoded form may well be syntactically
3295 invalid. However, $reply_address has leading white space removed, and all
3296 newlines turned into spaces so that the autoreply transport does not
3299 PH/05 If group was specified without a user on a router, and no group or user
3300 was specified on a transport, the group from the router was ignored.
3302 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3303 for visible compile-time settings that can be used to change these
3304 numbers, for those that want even more. Backwards compatibility with old
3305 spool files has been maintained. However, going back to a previous Exim
3306 release will lost any variables that are in spool files.
3308 PH/07 Two small changes when running in the test harness: increase delay when
3309 passing a TCP/IP connection to a new process, in case the original
3310 process has to generate a bounce, and remove special handling of
3311 127.0.0.2 (sic), which is no longer necessary.
3313 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3314 be the same on different OS.
3316 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3319 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3320 whether --show-vars was specified or not
3322 JJ/02 exipick: Added support for new ACL variable spool format introduced
3325 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3326 syntactically invalid From: or Reply-to: line, and a filter used this to
3327 generate an autoreply, and therefore failed to obtain an address for the
3328 autoreply, Exim could try to deliver to a non-existent relative file
3329 name, causing unrelated and misleading errors. What now happens is that
3330 it logs this as a hard delivery error, but does not attempt to create a
3333 PH/11 The exinext utility has a -C option for testing purposes, but although
3334 the given file was scanned by exinext itself; it wasn't being passed on
3335 when Exim was called.
3337 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3338 an end-of-file indication when reading a command response.
3340 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3341 compiled. In many other places in Exim, IPv6 addresses are always
3342 recognized, so I have changed this. It also means that IPv4 domain
3343 literals of the form [IPV4:n.n.n.n] are now always recognized.
3345 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3346 used if the router is not running as root, for example, when verifying at
3347 ACL time, or when using -bh. The debugging output from this situation was
3348 non-existent - all you got was a failure to exec. I have made two
3351 (a) Failures to set uid/gid, the current directory, or a process leader
3352 in a subprocess such as that created by queryprogram now generate
3353 suitable debugging output when -d is set.
3355 (b) The queryprogram router detects when it is not running as root,
3356 outputs suitable debugging information if -d is set, and then runs
3357 the subprocess without attempting to change uid/gid.
3359 PH/15 Minor change to Makefile for building test_host (undocumented testing
3362 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3363 additional section of a DNS packet that returns MX or SRV records.
3364 Instead, it always explicitly searches for A/AAAA records. This avoids
3365 major problems that occur when a DNS server includes only records of one
3366 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3367 fixed another bug: if SRV records were looked up and the corresponding
3368 address records were *not* found in the additional section, the port
3369 values from the SRV records were lost.
3371 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3372 using the correct key (the original address) when searching the retry
3373 rules in order to find which one to use for generating the retry hint.
3375 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3376 adding the default one. Similarly, if it contains a Reply-To: header, the
3377 errors_reply_to option, if set, is not used.
3379 PH/19 When calculating a retry time, Exim used to measure the "time since
3380 failure" by looking at the "first failed" field in the retry record. Now
3381 it does not use this if it is later than than the arrival time of the
3382 message. Instead it uses the arrival time. This makes for better
3383 behaviour in cases where some deliveries succeed, thus re-setting the
3384 "first failed" field. An example is a quota failure for a huge message
3385 when small messages continue to be delivered. Without this change, the
3386 "time since failure" will always be short, possible causing more frequent
3387 delivery attempts for the huge message than are intended.
3388 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3390 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3391 $1, $2, $3) because the numerical variables can be reset during some
3392 expansion items (e.g. "match"), thereby losing the authentication data.
3394 PH/21 Make -bV show the size of off_t variables so that the test suite can
3395 decide whether to run tests for quotas > 2G.
3397 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3398 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3399 filecount value is greater than 2G or if a quota value is greater than 2G
3400 on a system where the size of off_t is not greater than 4, a panic error
3403 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3404 never match. The debug and -bh output now contains an explicit error
3405 message indicating a malformed IPv4 address or mask.
3407 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3408 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3409 PH/23 above applies.
3411 PH/25 Do not write to syslog when running in the test harness. The only
3412 occasion when this arises is a failure to open the main or panic logs
3413 (for which there is an explicit test).
3415 PH/26 Added the /no_tell option to "control=freeze".
3417 PH/27 If a host name lookup failed very early in a connection, for example, if
3418 the IP address matched host_lookup and the reverse lookup yielded a name
3419 that did not have a forward lookup, an error message of the form "no IP
3420 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3421 could be logged. Now it outputs the IP address instead of "NULL".
3423 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3424 allows the sender and the authenticated sender to be set when
3425 submitting a message from within Exim. Since child_open_exim() is
3426 documented for local_scan(), the new function should be too.
3428 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3429 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3430 results in an empty string is now treated as unset.
3432 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3434 PH/31 Added sender_verify_fail logging option.
3436 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3437 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3438 tidied the source and removed it altogether.
3440 PH/33 When a queue run was abandoned because the load average was too high, a
3441 log line was always written; now it is written only if the queue_run log
3442 selector is set. In addition, the log line for abandonment now contains
3443 information about the queue run such as the pid. This is always present
3444 in "start" and "stop" lines but was omitted from the "abandon" line.
3446 PH/34 Omit spaces between a header name and the colon in the error message that
3447 is given when verify = headers_syntax fails (if there are lots of them,
3448 the message gets confusing).
3450 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3451 names, as there are now some PTR records that contain slashes. This check
3452 is only to protect against broken name servers that fall over on strange
3453 characters, so the fact that it applies to all lookups doesn't matter.
3455 PH/36 Now that the new test suite is complete, we can remove some of the
3456 special code in Exim that was needed for the old test suite. For example,
3457 sorting DNS records because real resolvers return them in an arbitrary
3458 order. The new test suite's fake resolver always returns records in the
3461 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3462 bounces) except when queue_only is set, to avoid logging races between
3463 the different processes.
3465 PH/38 Panic-die if .include specifies a non-absolute path.
3467 PH/39 A tweak to the "H" retry rule from its user.
3469 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3470 a label. They prevented compilation on older perls.
3472 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3473 a warning to be raised on newish perls.
3475 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3476 on queue. Changes to match documented behaviour of showing count of
3477 messages matching specified criteria.
3479 PH/40 Changed the default ident timeout from 30s to 5s.
3481 PH/41 Added support for the use of login_cap features, on those BSD systems
3482 that have them, for controlling the resources used by pipe deliveries.
3484 PH/42 The content-scanning code uses fopen() to create files in which to put
3485 message data. Previously it was not paying any attention to the mode of
3486 the files. Exim runs with umask(0) because the rest of the code creates
3487 files with open(), and sets the required mode explicitly. Thus, these
3488 files were ending up world-writeable. This was not a big issue, because,
3489 being within the spool directory, they were not world-accessible. I have
3490 created a function called modefopen, which takes an additional mode
3491 argument. It sets umask(777), creates the file, chmods it to the required
3492 mode, then resets the umask. All the relevant calls to fopen() in the
3493 content scanning code have been changed to use this function.
3495 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3496 to 24 hours. This avoids potential overflow problems when processing G
3497 and H retry rules. I suspect nobody ever tinkers with this value.
3499 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3501 PH/45 When the plaintext authenticator is running as a client, the server's
3502 challenges are checked to ensure they are valid base64 strings. By
3503 default, the authentication attempt is cancelled if an invalid string is
3504 received. Setting client_ignore_invalid_base64 true ignores these errors.
3505 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3506 they are received. Thus, the responses can be made to depend on the
3507 challenges. If an invalid string is ignored, an empty string is placed in
3510 PH/46 Messages that are created by the autoreply transport now contains a
3511 References: header, in accordance with RFCs 2822 and 3834.
3513 PH/47 Added authenticated_sender_force to the smtp transport.
3515 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3517 PH/49 Installed latest patch from the Sieve maintainer.
3519 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3520 was also set, the appendfile transport was unnecessarily scanning a
3521 directory of message files (e.g. for maildir delivery) to find the count
3522 of files (along with the size), even though it did not need this
3523 information. It now does the scan only if it needs to find either the
3524 size of the count of files.
3526 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3528 PH/52 Two bugs concerned with error handling when the smtp transport is
3531 (i) Exim was not creating retry information for temporary errors given
3532 for individual recipients after the DATA command when the smtp transport
3533 was used in LMTP mode. This meant that they could be retried too
3534 frequently, and not timed out correctly.
3536 (ii) Exim was setting the flag that allows error details to be returned
3537 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3538 recipients that were returned after the DATA command.
3540 PH/53 This is related to PH/52, but is more general: for any failing address,
3541 when detailed error information was permitted to be returned to the
3542 sender, but the error was temporary, then after the final timeout, only
3543 "retry timeout exceeded" was returned. Now it returns the full error as
3544 well as "retry timeout exceeded".
3546 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3547 do this, and (what is worse) MTAs that accept it.
3549 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3550 will now be deprecated.
3552 PH/56 New os.c-cygwin from the Cygwin maintainer.
3554 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3555 formats (previously only available in exim formats via -bpr, -bpru,
3556 and -bpra. Now also available in native and exiqgrep formats)
3558 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3559 with very large, slow to parse queues
3561 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3563 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3565 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3566 responses to authentication challenges, though it was showing the
3567 challenges; (ii) I've removed the CR characters from the debug output for
3570 PH/58 Allow for the insertion of a newline as well as a space when a string
3571 is turned into more than one encoded-word during RFC 2047 encoding. The
3572 Sieve code now uses this.
3574 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3575 data_4xx, lost_connection, tls_required.
3577 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3578 message was being sent as an SMTP response.
3580 PH/61 Add -l and -k options to exicyclog.
3582 PH/62 When verifying, if an address was redirected to one new address, so that
3583 verification continued, and the new address failed or deferred after
3584 having set something in $address_data, the value of $address_data was not
3585 passed back to the ACL. This was different to the case when no
3586 redirection occurred. The value is now passed back in both cases.
3588 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3589 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3590 use login_cap.h, so on its own it isn't the distinguishing feature. The
3591 new name refers directly to the setclassresources() function.
3593 PH/65 Added configuration files for NetBSD3.
3595 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3597 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3598 is preferred over IPv4.
3600 PH/68 The bounce_return_message and bounce_return_body options were not being
3601 honoured for bounces generated during the reception of non-SMTP messages.
3602 In particular, this applied to messages rejected by the ACL. This bug has
3603 been fixed. However, if bounce_return_message is true and bounce_return_
3604 body is false, the headers that are returned for a non-SMTP message
3605 include only those that have been read before the error was detected.
3606 (In the case of an ACL rejection, they have all been read.)
3608 PH/69 The HTML version of the specification is now built in a directory called
3609 spec_html instead of spec.html, because the latter looks like a path with
3610 a MIME-type, and this confuses some software.
3612 PH/70 Catch two compiler warnings in sieve.c.
3614 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3615 function verify_get_ident() calls ip_connect() to connect a socket, but
3616 if the "connect()" function timed out, ip_connect() used to close the
3617 socket. However, verify_get_ident() also closes the socket later, and in
3618 between Exim writes to the log, which may get opened at this point. When
3619 the socket was closed in ip_connect(), the log could get the same file
3620 descriptor number as the socket. This naturally causes chaos. The fix is
3621 not to close the socket in ip_connect(); the socket should be closed by
3622 the function that creates it. There was only one place in the code where
3623 this was missing, in the iplookup router, which I don't think anybody now
3624 uses, but I've fixed it anyway.
3626 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3627 well as to direct DNS lookups. Otherwise the handling of names in host
3628 lists is inconsistent and therefore confusing.
3634 PH/01 Two changes to the default runtime configuration:
3636 (1) Move the checks for relay_from_hosts and authenticated clients from
3637 after to before the (commented out) DNS black list checks.
3639 (2) Add control=submission to the relay_from_hosts and authenticated
3640 clients checks, on the grounds that messages accepted by these
3641 statements are most likely to be submissions.
3643 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3645 (1) Generate an error if the third argument for the ${prvs expansion is
3648 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3651 (3) Reset the variables that are obtained from the first argument of
3652 ${prvscheck and used in the second argument before leaving the code,
3653 because their memory is reclaimed, so using them afterwards may do
3656 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3657 one (it's much easier than Tom thought :-).
3659 (5) Because of (4), we can now allow for the use of $prvscheck_result
3660 inside the third argument.
3662 PH/03 For some reason, the default setting of PATH when running a command from
3663 a pipe transport was just "/usr/bin". I have changed it to
3666 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3667 anything to be listed in the output from -bV.
3669 PH/05 When a filter generated an autoreply, the entire To: header line was
3670 quoted in the delivery log line, like this:
3672 => >A.N.Other <ano@some.domain> <original@ddress> ...
3674 This has been changed so that it extracts the operative address. There
3675 may be more than one such address. If so, they are comma-separated, like
3678 => >ano@some.domain,ona@other.domain <original@ddress> ...
3680 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3681 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3682 not being looked up in the rDNS to get a host name, Exim was showing the
3683 IP address twice in Received: lines, even though the IP addresses were
3684 identical. For example:
3686 Received: from [1.2.3.4] (helo=[1.2.3.4])
3688 However, if the real host name was known, it was omitting the HELO data
3689 if it matched the actual IP address. This has been tidied up so that it
3690 doesn't show the same IP address twice.
3692 PH/07 When both +timestamp and +memory debugging was on, the value given by
3693 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3694 called by the debug printing, thereby overwriting the timestamp buffer.
3695 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3697 PH/08 When the original message was included in an autoreply transport, it
3698 always said "this is a copy of the message, including all the headers",
3699 even if body_only or headers_only was set. It now gives an appropriate
3702 PH/09 Applied a patch from the Sieve maintainer which:
3704 o fixes some comments
3705 o adds the (disabled) notify extension core
3706 o adds some debug output for the result of if/elsif tests
3707 o points to the current vacation draft in the documentation
3708 and documents the missing references header update
3712 o fixes a bug in processing the envelope test (when testing
3713 multiple envelope elements, the last element determined the
3716 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3717 Electronic Mail") by including:
3719 Auto-submitted: auto-generated
3721 in the messages that it generates (bounce messages and others, such as
3722 warnings). In the case of bounce messages for non-SMTP messages, there was
3723 also a typo: it was using "Auto_submitted" (underscore instead of
3724 hyphen). Since every message generated by Exim is necessarily in response
3725 to another message, thes have all been changed to:
3727 Auto-Submitted: auto-replied
3729 in accordance with these statements in the RFC:
3731 The auto-replied keyword:
3733 - SHOULD be used on messages sent in direct response to another
3734 message by an automatic process,
3736 - MUST NOT be used on manually-generated messages,
3738 - MAY be used on Delivery Status Notifications (DSNs) and Message
3739 Disposition Notifications (MDNs),
3741 - MUST NOT be used on messages generated by automatic or periodic
3742 processes, except for messages which are automatic responses to
3745 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3746 to the default Received: header definition.
3748 PH/12 Added log selector acl_warn_skipped (default on).
3750 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3751 variables because (a) they are in the wrong storage pool and (b) even if
3752 they were copied, it wouldn't work properly because of the caching.
3754 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3755 checking when decoding. Apparently there are clients that generate
3756 overlong encoded strings. Why am I not surprised?
3758 PH/15 If the first argument of "${if match_address" was not empty, but did not
3759 contain an "@" character, Exim crashed. Now it writes a panic log message
3760 and treats the condition as false.
3762 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3764 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3765 extension "envelope-auth". The code is finished and in agreement with
3766 other implementations, but there is no documentation so far and in fact,
3767 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3768 not changing the active code.
3770 Print executed "if" and "elsif" statements when debugging is used. This
3771 helps a great deal to understand what a filter does.
3773 Document more things not specified clearly in RFC3028. I had all this
3774 sorted out, when out of a sudden new issues came to my mind. Oops."
3776 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3779 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3780 canonical form (as documented). However, after a host name lookup from
3781 the IP address, check_host() was doing a simple string comparison with
3782 addresses acquired from the DNS when checking that the found name did
3783 have the original IP as one of its addresses. Since any found IPv6
3784 addresses are likely to be in abbreviated form, the comparison could
3785 fail. Luckily, there already exists a function for doing the comparison
3786 by converting both addresses to binary, so now that is used instead of
3787 the text comparison.
3789 PH/20 There was another similar case to PH/19, when a complete host name was
3790 given in a host list; looking up its IP address could give an abbreviated
3791 form, whereas the current host's name might or might not be abbreviated.
3792 The same fix has been applied.
3798 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3799 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3802 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3803 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3805 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3807 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3808 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3809 Also, for "Auto-Submitted", treat anything other than "no" as
3810 non-personal, in accordance with RFC 3834. (Previously it treated
3811 anything starting "auto-" as non-personal.)
3813 TF/01 The control=submission/name=... option had a problem with syntax
3814 errors if the name included a slash character. The /name= option
3815 now slurps the rest of the string, so it can include any characters
3816 but it must come last in the list of options (after /sender_retain
3819 PH/05 Some modifications to the interface to the fake nameserver for the new
3827 TK/01 Added the "success_on_redirect" address verification option. See
3828 NewStuff for rationale and an example.
3830 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3832 PH/02 Patch to exigrep to allow it to work on syslog lines.
3834 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3835 fread() to skip over the body file's header line, because in Cygwin the
3836 header line is locked and is inaccessible.
3838 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3839 co-exist for some time) to make it clear that it is the Exim ID that is
3840 referenced, not the Message-ID: header line.
3842 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3843 string_format() function, because snprintf() does not exist on all
3846 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3847 ${stat: expansion item.
3849 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3850 protocol synchronization error", to keep the pedants happy.
3852 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3853 well as for IRIX systems, when gcc is being used. See the host.c source
3856 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3858 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3861 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3862 transport and to the smtp transport in LMTP mode.
3864 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3866 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3867 run a filter in a subprocess. This could lead to confusion in subsequent
3868 lookups in the parent process. There should also be a search_tidyup() at
3869 the end of the subprocess.
3871 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3872 only if the host matched helo_try_verify_hosts, which caused the
3873 verification to occur when the EHLO/HELO command was issued. The ACL just
3874 tested the remembered result. Now, if a previous verification attempt has
3875 not happened, "verify = helo" does it there and then.
3877 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3879 TK/03 Fix log output including CR from clamd.
3881 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3882 exist provoked a memory error which could cause a segfault.
3884 PH/15 Installed PCRE 6.2
3886 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3888 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3889 of the problem. Specifically, suggested +O2 rather than +O1 for the
3892 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3894 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3895 record, the port was not correctly logged when the outgoing_port log
3896 selector was set (it logged the transort's default port).
3898 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3899 fallback_hosts, and "hosts" in the smtp transport.
3901 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3902 host errors such as "Connection refused".
3904 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3905 authentication with radiusclient 0.4.9:
3907 - Error returned from rc_read_config was caught wrongly
3908 - Username/password not passed on to radius server due to wrong length.
3910 The presumption is that some radiusclient API changes for 4.51/PH/17
3911 were not taken care of correctly. The code is still untested by me (my
3912 Linux distribution still has 0.3.2 of radiusclient), but it was
3913 contributed by a Radius user.
3915 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3916 expanding the "port" option of the smtp transport.
3918 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3919 while reading a MIME header. Thanks to Tom Hughes for a patch.
3921 PH/24 Include config.h inside local_scan.h so that configuration settings are
3924 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3925 that in an ACL for RCPT (for example), you can examine exactly what was
3928 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3929 commands, but it was not correctly comparing the address with the actual
3930 client host address. Thus, it would show the EHLO address in Received:
3931 header lines when this was not necessary.
3933 PH/27 Added the % operator to ${eval:}.
3935 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3936 it should be ignoring failures (because with -C, for example, it has lost
3937 privilege). It wasn't ignoring creation failures other than "already
3940 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3943 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3944 because an input line was too long, either on its own, or by virtue of
3945 too many continuations, the temporary file was not being removed, and the
3946 return code was incorrect.
3948 PH/31 Missing "BOOL" in function definition in filtertest.c.
3950 PH/32 Applied Sieve patches from the maintainer.
3952 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3954 PH/33 Added "verify = not_blind".
3956 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3957 Local/Makefile (with some defaults set). These are used in built scripts
3958 such as exicyclog, but they have never been used in the exim_install
3959 script (though there are many overriding facilities there). I have
3960 arranged that the exim_install script now takes note of these two
3963 PH/35 Installed configuration files for Dragonfly.
3965 PH/36 When a locally submitted message by a trusted user did not contain a
3966 From: header, and the sender address was obtained from -f or from an SMTP
3967 MAIL command, and the trusted user did not use -F to supply a sender
3968 name, $originator_name was incorrectly used when constructing a From:
3969 header. Furthermore, $originator_name was used for submission mode
3970 messages from external hosts without From: headers in a similar way,
3971 which is clearly wrong.
3973 PH/37 Added control=suppress_local_fixups.
3975 PH/38 When log_selector = +received_sender was set, and the addition of the
3976 sender made the log line's construction buffer exactly full, or one byte
3977 less than full, an overflow happened when the terminating "\n" was
3980 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3981 when the result of a list match is failure because a DNS lookup failed.
3983 PH/40 RM_COMMAND is now used in the building process.
3985 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3986 the "build-* directories that it finds.
3988 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3989 address in a domain literal was a prefix of an interface address.
3991 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3992 when verifying a sender address, unless rewrite_headers is false.
3994 PH/44 Wrote a long comment about why errors_to addresses are verified as
3995 recipients, not senders.
3997 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3998 the ratelimit ACL was added.
4000 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4002 PH/46 Added extra information about PostgreSQL errors to the error string.
4004 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4005 avoiding the need to install special zones in a real server. This is
4006 backwards compatible; if it can't find the fake resolver, it drops back.
4007 Thus, both old and new test suites can be run.
4009 TF/02 Added util/ratelimit.pl
4011 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4012 clock is set back in time.
4014 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4015 Candler <B.Candler@pobox.com>.
4017 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4018 OK for addresses that are the result of redirections.
4020 PH/48 A number of further additions for the benefit of the new test suite,
4021 including a fake gethostbyname() that interfaces to the fake DNS resolver
4024 TF/06 The fix for widen_domains has also been applied to qualify_single and
4025 search_parents which are the other dnslookup options that can cause
4028 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4031 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4033 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4034 service" if it encountered an MX record with an empty target hostname.
4035 The message is now "an MX or SRV record indicated no SMTP service".
4037 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4038 if the DNS of the sending site is misconfigured. This is quite a
4039 common situation. This change restores the behaviour of treating a
4040 helo verification defer as a failure.
4042 PH/51 If self=fail was set on a router, the bounce message did not include the
4043 actual error message.
4049 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4051 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4052 command itself ended in error, the underlying message about the transport
4053 filter timeout was being overwritten with the pipe command error. Now the
4054 underlying error message should be appended to the second error message.
4056 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4058 PH/02 Reduce the amount of output that "make" produces by default. Full output
4059 can still be requested.
4061 PH/03 The warning log line about a condition test deferring for a "warn" verb
4062 was being output only once per connection, rather than after each
4063 occurrence (because it was using the same function as for successful
4064 "warn" verbs). This seems wrong, so I have changed it.
4066 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4067 it should not have, which might have caused a crash in the right
4068 circumstances, but probably never did.
4070 PH/04 Installed a modified version of Tony Finch's patch to make submission
4071 mode fix the return path as well as the Sender: header line, and to
4072 add a /name= option so that you can make the user's friendly name appear
4075 TF/03 Added the control = fakedefer ACL modifier.
4077 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4078 Mark Lowes for thorough testing.
4080 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4082 TK/03 Merged latest SRS patch from Miles Wilton.
4084 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4085 with the definition in sysexits.h (which is #included earlier).
4086 Fortunately, Exim does not actually use EX_OK. The code used to try to
4087 preserve the sysexits.h value, by assuming that macro definitions were
4088 scanned for macro replacements. I have been disabused of this notion,
4089 so now the code just undefines EX_OK before #including unistd.h.
4091 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4092 in the smtp transport. When a block could not be written in a single
4093 write() function, the timeout was being re-applied to each part-write.
4094 This seems wrong - if the receiver was accepting one byte at a time it
4095 would take for ever. The timeout is now adjusted when this happens. It
4096 doesn't have to be particularly precise.
4098 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4099 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4101 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4102 without a domain if the check to <postmaster@domain> fails.
4104 SC/01 Eximstats: added -xls and the ability to specify output files
4105 (patch written by Frank Heydlauf).
4107 SC/02 Eximstats: use FileHandles for outputting results.
4109 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4111 SC/04 Eximstats: fixed display of large numbers with -nvr option
4113 SC/05 Eximstats: fixed merging of reports with empty tables.
4115 SC/06 Eximstats: added the -include_original_destination flag
4117 SC/07 Eximstats: removed tabs and trailing whitespace.
4119 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4121 TK/06 MBOX spool code: Add real "From " MBOX separator line
4122 so the .eml file is really in mbox format (even though
4123 most programs do not really care). Patch from Alex Miller.
4125 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4126 The latter is generated from $received_to and is only set if the
4127 message has one envelope recipient. SA can use these headers,
4128 obviously out-of-the-box. Patch from Alex Miller.
4130 PH/08 The ${def test on a variable was returning false if the variable's
4131 value was "0", contrary to what the specification has always said!
4132 The result should be true unless the variable is empty.
4134 PH/09 The syntax error of a character other than { following "${if
4135 def:variable_name" (after optional whitespace) was not being diagnosed.
4136 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4137 accidental colon was present, for example, could give incorrect results.
4139 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4140 result is used (not including appendfile, where other changes are about
4143 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4144 This involved changing a lot of size variables from int to off_t. It
4145 should work with maildirs and everything.
4147 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4148 spamd dying while we are connected to it.
4150 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4153 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4154 function for BDB 4.3.
4156 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4158 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4159 the sources for PCRE have been re-arranged and more files are now
4162 PH/15 The code I had for printing potentially long long variables in PH/11
4163 above was not the best (it lost precision). The length of off_t variables
4164 is now inspected at build time, and an appropriate printing format (%ld
4165 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4166 to be "long long int" or "long int". This is needed for the internal
4167 formatting function string_vformat().
4169 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4170 the configuration file to be ":syslog", then the script "guesses" where
4171 the logs files are, rather than using the compiled in default. In our
4172 case the guess is not the same as the compiled default, so the script
4173 suddenly stopped working when I started to use syslog. The patch checks
4174 to see if log_file_path is "". If so, it attempts to read it from exim
4175 with no configuration file to get the compiled in version, before it
4176 falls back to the previous guessing code."
4178 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4179 implementing BATV in an Exim configuration. See NewStuff for the gory
4182 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4183 Makefile that are specific to HP-UX.
4185 PH/18 If the "use_postmaster" option was set for a recipient callout together
4186 with the "random" option, the postmaster address was used as the MAIL
4187 FROM address for the random test, but not for the subsequent recipient
4188 test. It is now used for both.
4190 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4191 patch removes a few documentation additions to RFC 3028, because the
4192 latest draft now contains them. It adds the new en;ascii-case comparator
4193 and a new error check for 8bit text in MIME parts. Comparator and
4194 require names are now matched exactly. I enabled the subaddress
4195 extension, but it is not well tested yet (read: it works for me)."
4197 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4198 rework some of the code of TK/09 above to avoid the hardwired use of
4199 "%lld" and "long long". Replaced the call to snprintf() with a call to
4202 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4203 records point to non-existent hosts", "retry timeout exceeded", and
4204 "retry time not reached for any host after a long failure period".
4206 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4207 experimental DomainKeys support:
4209 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4210 (2) On an error such as an illegally used "control", the wrong name for
4211 the control was given.
4213 These problems did NOT occur unless DomainKeys support was compiled.
4215 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4217 PH/24 Added ${if match_ip condition.
4219 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4220 databases so that it will be absolutely obvious if a crash occurs in the
4221 DB library. This is a regular occurrence (often caused by mis-matched
4224 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4225 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4226 picked up on a user's system that detects such things. There doesn't seem
4227 to be a gcc warning option for this - only an attribute that has to be
4228 put on the function's prototype. It seems that in Fedora Core 4 they have
4229 set this on a number of new functions. No doubt there will be more in due
4232 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4233 specify a transport. However, if an address that was verified by such a
4234 router was the subject of a callout, Exim crashed because it tried to
4235 read the rcpt_include_affixes from the non-existent transport. Now it
4236 just assumes that the setting of that option is false. This bug was
4237 introduced by 4.51/PH/31.
4239 PH/28 Changed -d+all to exclude +memory, because that information is very
4240 rarely of interest, but it makes the output a lot bigger. People tend to
4241 do -d+all out of habit.
4243 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4244 code in os-type was giving problems when libc.so lives in lib64, like on
4247 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4248 aren't the modern standard, and it seems that some systems' include files
4249 don't always have them. Exim was already checking for some of the newer
4250 ones like T_AAAA, and defining it itself. I've added checks for all the
4251 record types that Exim uses.
4253 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4254 not automatically generating a new one, as it is supposed to. This
4255 prevented TLS from working. If the file did exist, but contained invalid
4256 data, a new version was generated, as expected. It was only the case of a
4257 non-existent file that was broken.
4259 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4260 with a change in libdomainkeys > 0.64.
4262 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4263 from DNS. If the selector record carries the flag, it now has
4264 precedence over the domain-wide flag.
4266 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4268 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4269 the use of a port name that isn't defined in /etc/services) occurred, the
4270 message was deferred as in a normal delivery, and thus remained on the
4271 spool, instead of being failed because of the mua_wrapper setting. This
4272 is now fixed, and I tidied up some of the mua_wrapper messages at the
4275 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4276 the messages in a hash of arrays rather than using individual hashes.
4277 This is a bit cleaner and results in dramatic memory savings, albeit
4278 at a slight CPU cost.
4280 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4281 as requested by Marc Sherman.
4283 SC/10 Eximstats: added histograms for user specified patterns as requested
4286 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4288 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4289 fopen() in the content-scanning modules that did not already have it.
4295 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4296 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4298 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4300 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4302 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4303 to test Sieve filters that use "vacation".
4305 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4306 that changes the way the GnuTLS parameters are stored in the cache file.
4307 The new format can be generated externally. For backward compatibility,
4308 if the data in the cache doesn't make sense, Exim assumes it has read an
4309 old-format file, and it generates new data and writes a new file. This
4310 means that you can't go back to an older release without removing the
4313 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4314 work if there are any delivery delays because "one_time" forces the
4315 parent to be marked "delivered", so its unseen clone is never tried
4316 again. For this reason, Exim now forbids the simultaneous setting of
4319 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4320 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4321 entry if you want to know the details. The fix, however, neglected to
4322 consider the case when local delivery batching is involved. The test for
4323 "previously delivered" was not happening when checking to see if an
4324 address could be batched with a previous (undelivered) one; under
4325 certain circumstances this could lead to multiple deliveries to the same
4328 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4329 in its include files, and this causes problems building Exim.
4331 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4332 header_syntax) but Exim was just ignoring anything given after a slash.
4333 In particular, this caused confusion with an attempt to use "verify =
4334 reverse_host_lookup/defer_ok". An error is now given when options are
4335 supplied for verify items that do not have them. (Maybe reverse_host_
4336 lookup should have a defer_ok option, but that's a different point.)
4338 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4339 defined by RFC 821) to 2048, because there were problems with some AUTH
4340 commands, and RFC 1869 says the size should be increased for extended
4341 SMTP commands that take arguments.
4343 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4346 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4347 "unknown" error; now it says that the functionality isn't in the binary.
4349 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4350 an address' error message when a string expansion fails (syntax or
4351 whatever). Otherwise the password may appear in the log. Following change
4352 PH/42 below, there is no longer a chance of it appearing in a bounce
4355 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4357 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4358 of its addresses was ever tried. (Bugzilla bug #2.)
4360 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4361 the result incorrectly in the debug output. (It correctly added a newline
4362 to what was transported.)
4364 TF/01 Added $received_time.
4366 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4367 commented out examples of how to interface to a virus scanner and to
4368 SpamAssassin. Also added commented examples of av_scanner and
4369 spamd_address settings.
4371 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4372 and controls are allowed in which ACLs. There were a couple of minor
4373 errors. Some of the entries in the conditions table (which is a table of
4374 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4375 negation of where the condition IS allowed.
4377 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4379 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4380 header file does not have a version number, so I've had to invent a new
4381 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4382 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4383 radiusclient), but it was contributed by a Radius user.
4385 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4386 files or directories, for OpenSSL.
4388 PH/19 When an Exim process that is running as root has to create an Exim log
4389 file, it does so in a subprocess that runs as exim:exim so as to get the
4390 ownership right at creation (otherwise, other Exim processes might see
4391 the file with the wrong ownership). There was no test for failure of this
4392 fork() call, which would lead to the process getting stuck as it waited
4393 for a non-existent subprocess. Forks do occasionally fail when resources
4394 run out. I reviewed all the other calls to fork(); they all seem to check
4397 PH/20 When checking for unexpected SMTP input at connect time (before writing
4398 the banner), Exim was not dealing correctly with a non-positive return
4399 from the read() function. If the client had disconnected by this time,
4400 the result was a log entry for a synchronization error with an empty
4401 string after "input=" when read() returned zero. If read() returned -1
4402 (an event I could not check), uninitialized data bytes were printed.
4403 There were reports of junk text (parts of files, etc) appearing after
4406 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4408 PH/22 Added support for macro redefinition, and (re)definition in between
4409 driver and ACL definitions.
4411 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4412 forgetting to use the resulting value; it was using the unexpanded value.
4414 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4415 hadn't been configured. The fix is from Juergen Kreileder, who
4416 understands it better than I do:
4418 "Here's what I see happening with three configured cyrus_sasl
4419 authenticators configured (plain, login, cram-md5):
4421 On startup auth_cyrus_sasl_init() gets called for each of these.
4422 This means three calls to sasl_listmech() without a specified mech_list.
4423 => SASL tests which mechs of all available mechs actually work
4424 => three warnings about OTP not working
4425 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4427 With the patch, sasl_listmech() also gets called three times. But now
4428 SASL's mech_list option is set to the server_mech specified in the the
4429 authenticator. Or in other words, the answer from sasl_listmech()
4430 gets limited to just the mech you're testing for (which is different
4432 => the return list contains just 'plain' or 'login', 'cram-md5' or
4433 nothing depending on the value of ob->server_mech.
4435 I've just tested the patch: Authentication still works fine,
4436 unavailable mechs specified in the exim configuration are still
4437 caught, and the auth.log warnings about OTP are gone."
4439 PH/25 When debugging is enabled, the contents of the command line are added
4440 to the debugging output, even when log_selector=+arguments is not
4443 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4444 answer is "GNU", and only if the return is "GNU/something" is the answer
4447 PH/27 $acl_verify_message is now set immediately after the failure of a
4448 verification in an ACL, and so is available in subsequent modifiers. In
4449 particular, the message can be preserved by coding like this:
4451 warn !verify = sender
4452 set acl_m0 = $acl_verify_message
4454 Previously, $acl_verify_message was set only while expanding "message"
4455 and "log_message" when a very denied access.
4457 PH/28 Modified OS/os.c-Linux with
4459 -#ifndef OS_LOAD_AVERAGE
4460 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4462 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4463 nomenclature these days.)
4465 PH/29 Installed patch from the Sieve maintainer that adds the options
4466 sieve_useraddress and sieve_subaddress to the redirect router.
4468 PH/30 In these circumstances:
4469 . Two addresses routed to the same list of hosts;
4470 . First host does not offer TLS;
4471 . First host accepts first address;
4472 . First host gives temporary error to second address;
4473 . Second host offers TLS and a TLS session is established;
4474 . Second host accepts second address.
4475 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4476 and peerdn, if requested) that were in fact used only for the second
4479 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4480 attention to any local part prefix or suffix that was matched by the
4481 router that accepted the address. It now behaves in the same way as it
4482 does for delivery: the affixes are removed from the local part unless
4483 rcpt_include_affixes is set on the transport.
4485 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4486 timeout during the DATA phase of an incoming message.
4488 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4489 applied a patch sanctioned by the Sieve maintainer.
4491 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4492 the uid or gid is negative. A case of a negative gid caused this to be
4493 noticed. The fix allows for either to be negative.
4495 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4496 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4499 PH/36 The change PH/12 above was broken. Fixed it.
4501 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4502 the grounds that routing the same address twice would always produce the
4503 same answer. This might have been true once, but it is certainly no
4504 longer true now. Routing a child address may depend on the previous
4505 routing that produced that child. Some complicated redirection strategies
4506 went wrong when messages had multiple recipients, and made Exim's
4507 behaviour dependent on the order in which the addresses were given.
4509 I have moved the duplicate checking until after the routing is complete.
4510 Exim scans the addresses that are assigned to local and remote
4511 transports, and removes any duplicates. This means that more work will be
4512 done, as duplicates will always all be routed, but duplicates are
4513 presumably rare, so I don't expect this is of any significance.
4515 For deliveries to pipes, files, and autoreplies, the duplicate checking
4516 still happens during the routing process, since they are not going to be
4519 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4520 It corrects a timeout issue with spamd. This is Ian's comment: "The
4521 background is that sometimes spamd either never reads data from a
4522 connection it has accepted, or it never writes response data. The exiscan
4523 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4524 blindly assumes that writes won't block so it may never time out."
4526 PH/39 Allow G after quota size as well as K and M.
4528 PH/40 The value set for $authenticated_id in an authenticator may not contain
4529 binary zeroes or newlines because the value is written to log lines and
4530 to spool files. There was no check on this. Now the value is run through
4531 the string_printing() function so that such characters are converted to
4532 printable escape sequences.
4534 PH/41 $message_linecount is a new variable that contains the total number of
4535 lines in the message. Compare $body_linecount, which is the count for the
4538 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4539 bounce and delay warning messages, except in certain special cases, which
4542 (a) An SMTP error message from a remote host;
4543 (b) A message specified in a :fail: redirection;
4544 (c) A message specified in a "fail" command in a system filter;
4545 (d) A message specified in a FAIL return from the queryprogram router;
4546 (e) A message specified by the cannot_route_message router option.
4548 In these cases only, Exim does include the error details in bounce and
4549 warning messages. There are also a few cases where bland messages such
4550 as "unrouteable address" or "local delivery error" are given.
4552 PH/43 $value is now also set for the "else" part of a ${run expansion.
4554 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4555 being worked on, but at least Exim now implements the latest version to
4558 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4559 process to complete was treated as a delivery failure, a timeout while
4560 writing the message to the pipe was logged, but erroneously treated as a
4561 successful delivery. Such timeouts include transport filter timeouts. For
4562 consistency with the overall process timeout, these timeouts are now
4563 treated as errors, giving rise to delivery failures by default. However,
4564 there is now a new Boolean option for the pipe transport called
4565 timeout_defer, which, if set TRUE, converts the failures into defers for
4566 both kinds of timeout. A transport filter timeout is now identified in
4569 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4570 systems where "make" and "gmake" are different, calling "gmake" at top
4571 level broke things. I've arranged for the value of $(MAKE) to be passed
4572 from the Makefile to this script so that it can call the same version of
4576 A note about Exim versions 4.44 and 4.50
4577 ----------------------------------------
4579 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4580 changes of various kinds. As a consequence, a big documentation update was
4581 needed. This delayed the release for rather longer than seemed good, especially
4582 in the light of a couple of (minor) security issues. Therefore, the changes
4583 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4584 release. So 4.44 and 4.50 are in effect two different branches that both start
4587 I have left the 4.50 change log unchanged; it contains all the changes since
4588 4.43. The change log for 4.44 is below; many of its items are identical to
4589 those for 4.50. This seems to be the most sensible way to preserve the
4590 historical information.
4596 1. Minor wording change to the doc/README.SIEVE file.
4598 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4599 computation of the current number of files was incorrect.
4601 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4602 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4605 4. Give more explanation in the error message when the command for a transport
4606 filter fails to execute.
4608 5. There are several places where Exim runs a non-Exim command in a
4609 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4610 was being done only for the command run by the queryprogram router. It is
4611 now done for all such subprocesses. The other cases are: ${run, transport
4612 filters, and the commands run by the lmtp and pipe transports.
4614 6. Added CONFIGURE_GROUP build-time option.
4616 7. Some older OS have a limit of 256 on the maximum number of file
4617 descriptors. Exim was using setrlimit() to set 1000 as a large value
4618 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4619 systems. I've change it so that if it can't get 1000, it tries for 256.
4621 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4622 was an oversight, and furthermore, ever since the addition of extra
4623 controls (e.g. 4.43/32), the checks on when to allow different forms of
4624 "control" were broken. There should now be diagnostics for all cases when a
4625 control that does not make sense is encountered.
4627 9. Added the /retain_sender option to "control=submission".
4629 10. $recipients is now available in the predata ACL (oversight).
4631 11. Tidy the search cache before the fork to do a delivery from a message
4632 received from the command line. Otherwise the child will trigger a lookup
4633 failure and thereby defer the delivery if it tries to use (for example) a
4634 cached ldap connection that the parent has called unbind on.
4636 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4637 of $address_data from the recipient verification was clobbered by the
4638 sender verification.
4640 13. The value of address_data from a sender verification is now available in
4641 $sender_address_data in subsequent conditions in the ACL statement.
4643 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4645 15. Added a new option "connect=<time>" to callout options, to set a different
4648 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4649 was its contents. (It was OK if the option was not defined at all.)
4651 17. A "Completed" log line is now written for messages that are removed from
4652 the spool by the -Mrm option.
4654 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4655 information about exactly what failed.
4657 19. Added -dd to debug only the daemon process.
4659 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4660 handles timeouts, both on the server side and network timeouts. Renamed the
4661 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4663 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4664 It is now set to "smtps".
4666 22. $host_address is now set to the target address during the checking of
4667 ignore_target_hosts.
4669 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4670 being passed; this would have caused $sender_host_name to have been used if
4671 matching the list had actually called for a host name (not very likely,
4672 since this list is usually IP addresses). A host name is now passed as
4675 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4676 code that specifies a non-restarting handler (typically sigaction() in
4677 modern systems) in an attempt to fix a rare and obscure crash bug.
4679 25. Narrowed the window for a race in the daemon that could cause it to ignore
4680 SIGCHLD signals. This is not a major problem, because they are used only to
4681 wake it up if nothing else does.
4683 26. A malformed maildirsize file could cause Exim to calculate negative values
4684 for the mailbox size or file count. Odd effects could occur as a result.
4685 The maildirsize information is now recalculated if the size or filecount
4688 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4689 support for a long time. Removed HAVE_SYS_VFS_H.
4691 28. Installed the latest version of exipick from John Jetmore.
4693 29. In an address list, if the pattern was not a regular expression, an empty
4694 subject address (from a bounce message) matched only if the pattern was an
4695 empty string. Non-empty patterns were not even tested. This was the wrong
4696 because it is perfectly reasonable to use an empty address as part of a
4697 database query. An empty address is now tested by patterns that are
4698 lookups. However, all the other forms of pattern expect the subject to
4699 contain a local part and a domain, and therefore, for them, an empty
4700 address still always fails if the pattern is not itself empty.
4702 30. Exim went into a mad DNS loop when attempting to do a callout where the
4703 host was specified on an smtp transport, and looking it up yielded more
4704 than one IP address.
4706 31. Re-factored the code for checking spool and log partition space into a
4707 function that finds that data and another that does the check. The former
4708 is then used to implement four new variables: $spool_space, $log_space,
4709 $spool_inodes, and $log_inodes.
4711 32. The RFC2047 encoding function was originally intended for short strings
4712 such as real names; it was not keeping to the 75-character limit for
4713 encoded words that the RFC imposes. It now respects the limit, and
4714 generates multiple encoded words if necessary. To be on the safe side, I
4715 have increased the buffer size for the ${rfc2047: expansion operator from
4718 33. It is now permitted to omit both strings after an "if" condition; if the
4719 condition is true, the result is "true". As before, when the second string
4720 is omitted, a false condition yields an empty string. This makes it less
4721 cumbersome to write custom ACL and router conditions.
4723 34. Failure to deliver a bounce message always caused it to be frozen, even if
4724 there was an errors_to setting on the router. The errors_to setting is now
4727 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4728 canonical form (fully expanded) before being placed in
4729 $sender_host_address.
4731 36. The table in the code that translates DNS record types into text (T_A to
4732 "A" for instance) was missing entries for NS and CNAME. It is just possible
4733 that this could have caused confusion if both these types were looked up
4734 for the same domain, because the text type is used as part of Exim's
4735 per-process caching. But the chance of anyone hitting this buglet seems
4738 37. The dnsdb lookup has been extended in a number of ways.
4740 (1) There is a new type, "zns", which walks up the domain tree until it
4741 finds some nameserver records. It should be used with care.
4743 (2) There is a new type, "mxh", which is like "mx" except that it returns
4744 just the host names, not the priorities.
4746 (3) It is now possible to give a list of domains (or IP addresses) to be
4747 looked up. The behaviour when one of the lookups defers can be
4748 controlled by a keyword.
4750 (4) It is now possible to specify the separator character for use when
4751 multiple records are returned.
4753 38. The dnslists ACL condition has been extended: it is now possible to supply
4754 a list of IP addresses and/or domains to be looked up in a particular DNS
4757 39. Added log_selector=+queue_time_overall.
4759 40. When running the queue in the test harness, wait just a tad after forking a
4760 delivery process, to get repeatability of debugging output.
4762 41. Include certificate and key file names in error message when GnuTLS fails
4763 to set them up, because the GnuTLS error message doesn't include the name
4764 of the failing file when there is a problem reading it.
4766 42. Allow both -bf and -bF in the same test run.
4768 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4770 44. The "Exiscan patch" is now merged into the mainline Exim source.
4772 45. Sometimes the final signoff response after QUIT could fail to get
4773 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4774 < 0 before doing a fflush(). This bug looks as though it goes back to the
4775 introduction of TLS in release 3.20, but "sometimes" must have been rare
4776 because the tests only now provoked it.
4778 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4779 (this can affect the format of dates).
4781 47. exim_tidydb, when checking for the continued existence of a message for
4782 which it has found a message-specific retry record, was not finding
4783 messages that were in split spool directories. Consequently, it was
4784 deleting retry records that should have stayed in existence.
4786 48. Steve fixed some bugs in eximstats.
4788 49. The SPA authentication driver was not abandoning authentication and moving
4789 on to the next authenticator when an expansion was forced to fail,
4790 contradicting the general specification for all authenticators. Instead it
4791 was generating a temporary error. It now behaves as specified.
4793 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4794 (the order specifies the preference for clients). The order is now AES256,
4795 AES128, 3DES, ARCFOUR128.
4797 51. Small patch to Sieve code - explicitly set From: when generating an
4800 52. Exim crashed if a remote delivery caused a very long error message to be
4801 recorded - for instance if somebody sent an entire SpamAssassin report back
4802 as a large number of 550 error lines. This bug was coincidentally fixed by
4803 increasing the size of one of Exim's internal buffers (big_buffer) that
4804 happened as part of the Exiscan merge. However, to be on the safe side, I
4805 have made the code more robust (and fixed the comments that describe what
4808 53. Now that there can be additional text after "Completed" in log lines (if
4809 the queue_time_overall log selector is set), a one-byte patch to exigrep
4810 was needed to allow it to recognize "Completed" as not the last thing in
4813 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4814 patch that reportedly fixes this has been added. I am not expert enough to
4815 create a test for it. This is what the patch creator wrote:
4817 "I found a little strange behaviour of ldap code when working with
4818 Windows 2003 AD Domain, where users was placed in more than one
4819 Organization Units. When I tried to give exim partial DN, the exit code
4820 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4821 But simultaneously result of request was absolutely normal ldap result,
4822 so I produce this patch..."
4824 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4825 REFERENCE, so I have modified the code to exclude the patch when that macro
4828 55. Some experimental protocols are using DNS PTR records for new purposes. The
4829 keys for these records are domain names, not reversed IP addresses. The
4830 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4831 leaves it alone. Component reversal etc. now happens only for IP addresses.
4834 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4836 57. Double the size of the debug message buffer (to 2048) so that more of very
4837 long debug lines gets shown.
4839 58. The exicyclog utility now does better if the number of log files to keep
4840 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4842 59. Two changes related to the smtp_active_hostname option:
4844 (1) $smtp_active_hostname is now available as a variable.
4845 (2) The default for smtp_banner uses $smtp_active_hostname instead
4846 of $primary_hostname.
4848 60. The host_aton() function is supposed to be passed a string that is known
4849 to be a valid IP address. However, in the case of IPv6 addresses, it was
4850 not checking this. This is a hostage to fortune. Exim now panics and dies
4851 if the condition is not met. A case was found where this could be provoked
4852 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4853 components; fortuitously, this particular loophole had already been fixed
4854 by change 4.50/55 above.
4856 If there are any other similar loopholes, the new check in host_aton()
4857 itself should stop them being exploited. The report I received stated that
4858 data on the command line could provoke the exploit when Exim was running as
4859 exim, but did not say which command line option was involved. All I could
4860 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4861 running as the user.
4864 61. There was a buffer overflow vulnerability in the SPA authentication code
4865 (which came originally from the Samba project). I have added a test to the
4866 spa_base64_to_bits() function which I hope fixes it.
4869 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4870 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4872 63. The daemon start-up calls getloadavg() while still root for those OS that
4873 need the first call to be done as root, but it missed one case: when
4874 deliver_queue_load_max is set with deliver_drop_privilege. This is
4875 necessary for the benefit of the queue runner, because there is no re-exec
4876 when deliver_drop_privilege is set.
4878 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4879 This has been fixed.
4881 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4882 was in use, was not putting the data itself into the right store pool;
4883 consequently, it could be overwritten for a subsequent message in the same
4884 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4887 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4889 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4890 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4891 the function were treating the return as a boolean value, which happened to
4892 work because 0=false and not-0=true, but is not correct code.
4894 68. The host_aton() function was not handling scoped IPv6 addresses (those
4895 with, for example, "%eth0" on the end) correctly.
4897 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4898 negated items (that is, ~something) in unsigned ints. Some compilers
4899 apparently mutter when there is no cast.
4901 70. If an address verification called from an ACL failed, and did not produce a
4902 user-specific message (i.e. there was only a "system" message), nothing was
4903 put in $acl_verify_message. In this situation, it now puts the system
4906 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4907 session; change 4.31/43 added the unwanted input to the log line - except
4908 that it did not do this in the start of session case. It now does.
4910 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4911 This is wrong and can cause the other end to generate a synchronization
4912 error if it is another Exim or anything else that does the synchronization
4913 check. A QUIT command is no longer sent after a timeout.
4915 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4916 during host lookups.
4918 74. The defer_ok option of callout verification was not working if it was used
4919 when verifying addresses in header lines, that is, for this case:
4921 verify = header_sender/callout=defer_ok
4923 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4924 those file descriptors could be used for SMTP connections. If anything
4925 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4926 could be sent to the SMTP client, causing chaos. The daemon now opens
4927 stdin, stdout, and stderr to /dev/null when it puts itself into the
4930 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4931 log by default. The user can override this with suitable Perl magic.
4933 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4934 the log message when discard triggers, was not working for the DATA ACL or
4935 for the non-SMTP ACL.
4937 78. Error message wording change in sieve.c.
4939 79. If smtp_accept_max_per_host was set, the number of connections could be
4940 restricted to fewer than expected, because the daemon was trying to set up
4941 a new connection before checking whether the processes handling previous
4942 connections had finished. The check for completed processes is now done
4943 earlier. On busy systems, this bug wouldn't be noticed because something
4944 else would have woken the daemon, and it would have reaped the completed
4947 80. If a message was submitted locally by a user whose login name contained one
4948 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4949 It caused a spool format error. I have fixed the spool reading code. A
4950 related problem was that the "from" clause in the Received: line became
4951 illegal because of the space(s). It is now covered by ${quote_local_part.
4953 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4956 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4958 83. Patch from Sieve maintainer to fix unterminated string problem in
4959 "vacation" handling.
4961 84. Some minor changes to the Linux configuration files to help with other
4962 OS variants using glibc.
4964 85. One more patch for Sieve to update vacation handling to latest spec.
4967 ----------------------------------------------------
4968 See the note above about the 4.44 and 4.50 releases.
4969 ----------------------------------------------------
4975 1. Change 4.43/35 introduced a bug that caused file counts to be
4976 incorrectly computed when quota_filecount was set in an appendfile
4979 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4980 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4983 3. Give more explanation in the error message when the command for a transport
4984 filter fails to execute.
4986 4. There are several places where Exim runs a non-Exim command in a
4987 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4988 was being done only for the command run by the queryprogram router. It is
4989 now done for all such subprocesses. The other cases are: ${run, transport
4990 filters, and the commands run by the lmtp and pipe transports.
4992 5. Some older OS have a limit of 256 on the maximum number of file
4993 descriptors. Exim was using setrlimit() to set 1000 as a large value
4994 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4995 systems. I've change it so that if it can't get 1000, it tries for 256.
4997 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4998 was an oversight, and furthermore, ever since the addition of extra
4999 controls (e.g. 4.43/32), the checks on when to allow different forms of
5000 "control" were broken. There should now be diagnostics for all cases when a
5001 control that does not make sense is encountered.
5003 7. $recipients is now available in the predata ACL (oversight).
5005 8. Tidy the search cache before the fork to do a delivery from a message
5006 received from the command line. Otherwise the child will trigger a lookup
5007 failure and thereby defer the delivery if it tries to use (for example) a
5008 cached ldap connection that the parent has called unbind on.
5010 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5011 of $address_data from the recipient verification was clobbered by the
5012 sender verification.
5014 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5015 was its contents. (It was OK if the option was not defined at all.)
5017 11. A "Completed" log line is now written for messages that are removed from
5018 the spool by the -Mrm option.
5020 12. $host_address is now set to the target address during the checking of
5021 ignore_target_hosts.
5023 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5024 being passed; this would have caused $sender_host_name to have been used if
5025 matching the list had actually called for a host name (not very likely,
5026 since this list is usually IP addresses). A host name is now passed as
5029 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5030 code that specifies a non-restarting handler (typically sigaction() in
5031 modern systems) in an attempt to fix a rare and obscure crash bug.
5033 15. Narrowed the window for a race in the daemon that could cause it to ignore
5034 SIGCHLD signals. This is not a major problem, because they are used only to
5035 wake it up if nothing else does.
5037 16. A malformed maildirsize file could cause Exim to calculate negative values
5038 for the mailbox size or file count. Odd effects could occur as a result.
5039 The maildirsize information is now recalculated if the size or filecount
5042 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5043 support for a long time. Removed HAVE_SYS_VFS_H.
5045 18. Updated exipick to current release from John Jetmore.
5047 19. Allow an empty sender to be matched against a lookup in an address list.
5048 Previously the only cases considered were a regular expression, or an
5051 20. Exim went into a mad DNS lookup loop when doing a callout where the
5052 host was specified on the transport, if the DNS lookup yielded more than
5055 21. The RFC2047 encoding function was originally intended for short strings
5056 such as real names; it was not keeping to the 75-character limit for
5057 encoded words that the RFC imposes. It now respects the limit, and
5058 generates multiple encoded words if necessary. To be on the safe side, I
5059 have increased the buffer size for the ${rfc2047: expansion operator from
5062 22. Failure to deliver a bounce message always caused it to be frozen, even if
5063 there was an errors_to setting on the router. The errors_to setting is now
5066 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5067 canonical form (fully expanded) before being placed in
5068 $sender_host_address.
5070 24. Updated eximstats to version 1.33
5072 25. Include certificate and key file names in error message when GnuTLS fails
5073 to set them up, because the GnuTLS error message doesn't include the name
5074 of the failing file when there is a problem reading it.
5076 26. Expand error message when OpenSSL has problems setting up cert/key files.
5079 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5080 (this can affect the format of dates).
5082 28. exim_tidydb, when checking for the continued existence of a message for
5083 which it has found a message-specific retry record, was not finding
5084 messages that were in split spool directories. Consequently, it was
5085 deleting retry records that should have stayed in existence.
5087 29. eximstats updated to version 1.35
5088 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5089 1.35 - bugfix such that pie charts by volume are generated correctly
5091 30. The SPA authentication driver was not abandoning authentication and moving
5092 on to the next authenticator when an expansion was forced to fail,
5093 contradicting the general specification for all authenticators. Instead it
5094 was generating a temporary error. It now behaves as specified.
5096 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5097 (the order specifies the preference for clients). The order is now AES256,
5098 AES128, 3DES, ARCFOUR128.
5100 31. Small patch to Sieve code - explicitly set From: when generating an
5103 32. Exim crashed if a remote delivery caused a very long error message to be
5104 recorded - for instance if somebody sent an entire SpamAssassin report back
5105 as a large number of 550 error lines. This bug was coincidentally fixed by
5106 increasing the size of one of Exim's internal buffers (big_buffer) that
5107 happened as part of the Exiscan merge. However, to be on the safe side, I
5108 have made the code more robust (and fixed the comments that describe what
5111 33. Some experimental protocols are using DNS PTR records for new purposes. The
5112 keys for these records are domain names, not reversed IP addresses. The
5113 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5114 leaves it alone. Component reversal etc. now happens only for IP addresses.
5117 34. The host_aton() function is supposed to be passed a string that is known
5118 to be a valid IP address. However, in the case of IPv6 addresses, it was
5119 not checking this. This is a hostage to fortune. Exim now panics and dies
5120 if the condition is not met. A case was found where this could be provoked
5121 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5122 components; fortuitously, this particular loophole had already been fixed
5123 by change 4.50/55 or 4.44/33 above.
5125 If there are any other similar loopholes, the new check in host_aton()
5126 itself should stop them being exploited. The report I received stated that
5127 data on the command line could provoke the exploit when Exim was running as
5128 exim, but did not say which command line option was involved. All I could
5129 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5130 running as the user.
5133 35. There was a buffer overflow vulnerability in the SPA authentication code
5134 (which came originally from the Samba project). I have added a test to the
5135 spa_base64_to_bits() function which I hope fixes it.
5138 36. The daemon start-up calls getloadavg() while still root for those OS that
5139 need the first call to be done as root, but it missed one case: when
5140 deliver_queue_load_max is set with deliver_drop_privilege. This is
5141 necessary for the benefit of the queue runner, because there is no re-exec
5142 when deliver_drop_privilege is set.
5144 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5145 was in use, was not putting the data itself into the right store pool;
5146 consequently, it could be overwritten for a subsequent message in the same
5147 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5150 38. Sometimes the final signoff response after QUIT could fail to get
5151 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5152 < 0 before doing a fflush(). This bug looks as though it goes back to the
5153 introduction of TLS in release 3.20, but "sometimes" must have been rare
5154 because the tests only now provoked it.
5160 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5161 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5162 Now it returns an integer. A number of calls were still expecting a T/F
5163 return. Fortuitously, in all cases, the tests worked in OK situations,
5164 which is the norm. However, things would have gone wrong on any write
5165 failures on the smtp file descriptor. This function is used when sending
5166 messages over SMTP and also when doing verify callouts.
5168 2. When Exim is called to do synchronous delivery of a locally submitted
5169 message (the -odf or -odi options), it no longer closes stderr before doing
5172 3. Implemented the mua_wrapper option.
5174 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5176 5. Implemented the functions header_remove(), header_testname(),
5177 header_add_at_position(), and receive_remove_recipient(), and exported them
5180 6. If an ACL "warn" statement specified the addition of headers, Exim already
5181 inserted X-ACL-Warn: at the start if there was no header name. However, it
5182 was not making this test for the second and subsequent header lines if
5183 there were newlines in the string. This meant that an invalid header could
5184 be inserted if Exim was badly configured.
5186 7. Allow an ACL "warn" statement to add header lines at the start or after all
5187 the Received: headers, as well as at the end.
5189 8. Added the rcpt_4xx retry error code.
5191 9. Added postmaster_mailfrom=xxx to callout verification option.
5193 10. Added mailfrom=xxxx to the callout verification option, for verify=
5196 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5197 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5199 12. Inserted some casts to stop certain compilers warning when using pointer
5200 differences as field lengths or precisions in printf-type calls (mostly
5201 affecting debugging statements).
5203 13. Added optional readline() support for -be (dynamically loaded).
5205 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5206 same clock tick as a message's arrival, so that its received time was the
5207 same as the "first fail" time on the retry record, and that message
5208 remained on the queue past the ultimate address timeout, every queue runner
5209 would try a delivery (because it was past the ultimate address timeout) but
5210 after another failure, the ultimate address timeout, which should have then
5211 bounced the address, did not kick in. This was a "< instead of <=" error;
5212 in most cases the first failure would have been in the next clock tick
5213 after the received time, and all would be well.
5215 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5216 being recognized when the domain list was tested by the match_domain
5217 condition in an expansion string.
5219 16. Added the ${str2b64: operator.
5221 17. Exim was always calling setrlimit() to set a large limit for the number of
5222 processes, without checking whether the existing limit was already
5223 adequate. (It did check for the limit on file descriptors.) Furthermore,
5224 errors from getrlimit() and setrlimit() were being ignored. Now they are
5225 logged to the main and panic logs, but Exim does carry on, to try to do its
5226 job under whatever limits there are.
5228 18. Imported PCRE 5.0.
5230 19. Trivial typo in log message " temporarily refused connection" (the leading
5233 20. If the log selector return_path_on_delivery was set and an address was
5234 redirected to /dev/null, the delivery process crashed because it assumed
5235 that a return path would always be set for a "successful" delivery. In this
5236 case, the whole delivery is bypassed as an optimization, and therefore no
5239 21. Internal re-arrangement: the function for sending a challenge and reading
5240 a response while authentication was assuming a zero-terminated challenge
5241 string. It's now changed to take a pointer and a length, to allow for
5242 binary data in such strings.
5244 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5246 23. Exim was not respecting finduser_retries when seeking the login of the
5247 uid under which it was called; it was always trying 10 times. (The default
5248 setting of finduser_retries is zero.) Also, it was sleeping after the final
5249 failure, which is pointless.
5251 24. Implemented tls_on_connect_ports.
5253 25. Implemented acl_smtp_predata.
5255 26. If the domain in control=submission is set empty, Exim assumes that the
5256 authenticated id is a complete email address when it generates From: or
5257 Sender: header lines.
5259 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5260 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5261 chown and chgrp in /bin and hostname in /usr/bin.
5263 28. Exim was keeping the "process log" file open after each use, just as it
5264 does for the main log. This opens the possibility of it remaining open for
5265 long periods when the USR1 signal hits a daemon. Occasional processlog
5266 errors were reported, that could have been caused by this. Anyway, it seems
5267 much more sensible not to leave this file open at all, so that is what now
5270 29. The long-running daemon process does not normally write to the log once it
5271 has entered its main loop, and it closes the log before doing so. This is
5272 so that log files can straightforwardly be renamed and moved. However,
5273 there are a couple of unusual error situations where the daemon does write
5274 log entries, and I had neglected to close the log afterwards.
5276 30. The text of an SMTP error response that was received during a remote
5277 delivery was being truncated at 512 bytes. This is too short for some of
5278 the long messages that one sometimes sees. I've increased the limit to
5281 31. It is now possible to make retry rules that apply only when a message has a
5282 specific sender, in particular, an empty sender.
5284 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5285 it possible to be selective about when SMTP synchronization is enforced.
5287 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5289 32. Implemented hosts_connection_nolog.
5291 33. Added an ACL for QUIT.
5293 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5296 35. Added mailbox_size and mailbox_filecount to appendfile.
5298 36. Added control = no_multiline_responses to ACLs.
5300 37. There was a bug in the logic of the code that waits for the clock to tick
5301 in the case where the clock went backwards by a substantial amount such
5302 that the microsecond fraction of "now" was more than the microsecond
5303 fraction of "then" (but the whole seconds number was less).
5305 38. Added support for the libradius Radius client library this is found on
5306 FreeBSD (previously only the radiusclient library was supported).
5312 1. When certain lookups returned multiple values in the form name=value, the
5313 quoting of the values was not always being done properly. Specifically:
5314 (a) If the value started with a double quote, but contained no whitespace,
5316 (b) If the value contained whitespace other than a space character (i.e.
5317 tabs or newlines or carriage returns) it was not quoted.
5318 This fix has been applied to the mysql and pgsql lookups by writing a
5319 separate quoting function and calling it from the lookup code. The fix
5320 should probably also be applied to nisplus, ibase and oracle lookups, but
5321 since I cannot test any of those, I have not disturbed their existing code.
5323 2. A hit in the callout cache for a specific address caused a log line with no
5324 reason for rejecting RCPT. Now it says "Previous (cached) callout
5325 verification failure".
5327 3. There was an off-by-one bug in the queryprogram router. An over-long
5328 return line was truncated at 256 instead of 255 characters, thereby
5329 overflowing its buffer with the terminating zero. As well as fixing this, I
5330 have increased the buffer size to 1024 (and made a note to document this).
5332 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5333 when Exim is waiting for an SMTP response from a remote server, Exim
5334 restarts its select() call on the socket, thereby resetting its timeout.
5335 This is not a problem when such interrupts are rare. Somebody set up a cron
5336 job to run exiwhat every 2 minutes, which is less than the normal select()
5337 timeout (5 or 10 minutes). This meant that the select() timeout never
5338 kicked in because it was always reset. I have fixed this by comparing the
5339 time when an interrupt arrives with the time at the start of the first call
5340 to select(). If more time than the timeout has elapsed, the interrupt is
5341 treated as a timeout.
5343 5. Some internal re-factoring in preparation for the addition of Sieve
5344 extensions (by MH). In particular, the "personal" test is moved to a
5345 separate function, and given an option for scanning Cc: and Bcc: (which is
5346 not set for Exim filters).
5348 6. When Exim created an email address using the login of the caller as the
5349 local part (e.g. when creating a From: or Sender: header line), it was not
5350 quoting the local part when it contained special characters such as @.
5352 7. Installed new OpenBSD configuration files.
5354 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5355 try to make them clearer.
5357 9. Callout options, other than the timeout value, were being ignored when
5358 verifying sender addresses in header lines. For example, when using
5360 verify = header_sender/callout=no_cache
5362 the cache was (incorrectly) being used.
5364 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5365 only the Cygwin environment.
5367 11. When return_path_on_delivery was set as a log selector, if different remote
5368 addresses in the same message used different return paths and parallel
5369 remote delivery occurred, the wrong values would sometimes be logged.
5370 (Whenever a remote delivery process finished, the return path value from
5371 the most recently started remote delivery process was logged.)
5373 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5374 lines when AUTH and/or TLS are in use. This is the "received protocol"
5375 field. Exim used to use "asmtp" for authenticated SMTP, without any
5376 indication (in the protocol name) for TLS use. Now it follows the RFC and
5377 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5378 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5379 names appear in log lines as well as in Received: header lines.
5381 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5382 extensions, and comparison tests, and to fix some bugs.
5384 14. Changes to the "personal" filter test:
5386 (1) The test was buggy in that it was just doing the equivalent of
5387 "contains" tests on header lines. For example, if a user's address was
5388 anne@some.where, the "personal" test would incorrectly be true for
5390 To: susanne@some.where
5392 This test is now done by extracting each address from the header in turn,
5393 and checking the entire address. Other tests that are part of "personal"
5394 are now done using regular expressions (for example, to check local parts
5395 of addresses in From: header lines).
5397 (2) The list of non-personal local parts in From: addresses has been
5398 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5399 taken from the Sieve specification recommendations.
5401 (3) If the message contains any header line starting with "List-" it is
5402 treated as non-personal.
5404 (4) The test for "circular" in the Subject: header line has been removed
5405 because it now seems ill-conceived.
5407 15. Minor typos in src/EDITME comments corrected.
5409 16. Installed latest exipick from John Jetmore.
5411 17. If headers_add on a router specified a text string that was too long for
5412 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5413 of string_sprintf() is now avoided.
5415 18. $message_body_size was not set (it was always zero) when running the DATA
5416 ACL and the local_scan() function.
5418 19. For the "mail" command in an Exim filter, no default was being set for
5419 the once_repeat time, causing a random time value to be used if "once" was
5420 specified. (If the value happened to be <= 0, no repeat happened.) The
5421 default is now 0s, meaning "never repeat". The "vacation" command was OK
5422 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5423 (I found it when inspecting the code).
5425 20. There is now an overall timeout for performing a callout verification. It
5426 defaults to 4 times the callout timeout, which applies to individual SMTP
5427 commands during the callout. The overall timeout applies when there is more
5428 than one host that can be tried. The timeout is checked before trying the
5429 next host. This prevents very long delays if there are a large number of
5430 hosts and all are timing out (e.g. when the network connections are timing
5431 out). The value of the overall timeout can be changed by specifying an
5432 additional sub-option for "callout", called "maxwait". For example:
5434 verify = sender/callout=5s,maxwait=20s
5436 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5437 to the end before writing, but this should make it even safer).
5439 22. Exim was forgetting that it had advertised PIPELINING for the second and
5440 subsequent messages on an SMTP connection. It was also not resetting its
5441 memory on STARTTLS and an internal HELO.
5443 23. When Exim logs an SMTP synchronization error within a session, it now
5444 records whether PIPELINING has been advertised or not.
5446 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5447 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5448 rather than long int.
5450 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5452 26. Added the never_mail option to autoreply.
5458 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5459 crash if the getsockname() call failed; this can happen if a connection is
5460 closed very soon after it is established. The problem was simply in the
5461 order in which certain operations were done, causing Exim to try to write
5462 to the SMTP stream before it had set up the file descriptor. The bug has
5463 been fixed by making things happen in the correct order.
5469 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5470 before the connection was closed, thus losing the rejection response.
5472 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5473 some early Solaris releases, but causes trouble in current releases where
5474 socklen_t is defined.
5476 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5479 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5482 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5483 that releasing the top of it at the end releases what was used for sub-
5484 expansions (unless the block got too big). However, discard this block if
5485 the first thing is a variable or header, so that we can use its block when
5486 it is dynamic (useful for very large $message_headers, for example).
5488 6. Lookups now cache *every* query, not just the most recent. A new, separate
5489 store pool is used for this. It can be recovered when all lookup caches are
5490 flushed. Lookups now release memory at the end of their result strings.
5491 This has involved some general refactoring of the lookup sources.
5493 7. Some code has been added to the store_xxx() functions to reduce the amount
5494 of flapping under certain conditions.
5496 8. log_incoming_interface used to affect only the <= reception log lines. Now
5497 it causes the local interface and port to be added to several more SMTP log
5498 lines, for example "SMTP connection from", and rejection lines.
5500 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5502 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5504 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5505 could be overwritten at the end of the current message (or the start of a
5506 new message if it was set in a HELO ACL). The value is now preserved for
5507 the duration of the SMTP connection.
5509 12. If a transport had a headers_rewrite setting, and a matching header line
5510 contained an unqualified address, that address was qualified, even if it
5511 did not match any rewriting rules. The underlying bug was that the values
5512 of the flags that permit the existence of unqualified sender and recipient
5513 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5514 non-local messages, and by -bnq for local messages) were not being
5515 preserved with the message after it was received.
5517 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5518 "next input=" as part of the text comprising the host identity instead of
5519 the correct text. The code was using the same buffer for two different
5520 strings. However, depending on which order the printing function evaluated
5521 its arguments, the bug did not always show up. Under Linux, for example, my
5522 test suite worked just fine.
5524 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5525 doesn't work with some older versions of Perl. It has been changed to "my",
5526 which in any case is probably the better facility to use.
5528 15. A really picky compiler found some instances of statements for creating
5529 error messages that either had too many or two few arguments for the format
5532 16. The size of the buffer for calls to the DNS resolver has been increased
5533 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5534 for addresses that have a lot of PTR records. This alleviates a problem; it
5535 does not fully solve it.
5537 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5538 buffer now truncates the list and logs the incident, which is the same
5539 action as happens when Exim is looking up a host name and its aliases.
5540 Previously in this situation something unpredictable would happen;
5541 sometimes it was "internal error: store_reset failed".
5543 18. If a server dropped the connection unexpectedly when an Exim client was
5544 using GnuTLS and trying to read a response, the client delivery process
5545 crashed while trying to generate an error log message.
5547 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5548 string, for example:
5550 warn message = H1: something\nH2: something
5552 the text was added as a single header line from Exim's point of view
5553 though it ended up OK in the delivered message. However, searching for the
5554 second and subsequent header lines using $h_h2: did not work. This has been
5555 fixed. Similarly, if a system filter added multiple headers in this way,
5556 the routers could not see them.
5558 20. Expanded the error message when iplsearch is called with an invalid key to
5559 suggest using net-iplsearch in a host list.
5561 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5562 ACLs are no longer actually imposed (and a message to that effect is
5565 22. If a "gecos" field in a passwd entry contained escaped characters, in
5566 particular, if it contained a \" sequence, Exim got it wrong when building
5567 a From: or a Sender: header from that name. A second bug also caused
5568 incorrect handling when an unquoted " was present following a character
5569 that needed quoting.
5571 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5572 was not being matched caselessly.
5574 24. Arranged for all hyphens in the exim.8 source to be escaped with
5577 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5578 properly. Recipient callout cache records were still being keyed to include
5579 the sender, even when use_sender was set false. This led to far more
5580 callouts that were necessary. The sender is no longer included in the key
5581 when use_sender is false.
5583 26. Added "control = submission" modifier to ACLs.
5585 27. Added the ${base62d: operator to decode base 62 numbers.
5587 28. dnsdb lookups can now access SRV records.
5589 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5590 the configuration file.
5592 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5593 (-v) mode. This makes the output for a verbose queue run more intelligible.
5595 31. Added a use_postmaster feature to recipient callouts.
5597 32. Added the $body_zerocount variable, containing the number of binary zero
5598 bytes in the message body.
5600 33. The time of last modification of the "new" subdirectory is now used as the
5601 "mailbox time last read" when there is a quota error for a maildir
5604 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5606 35. Added +ignore_unknown as a special item in host lists.
5608 36. Code for decoding IPv6 addresses in host lists is now included, even if
5609 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5610 address was recognized as an IP address, but was then not correctly decoded
5611 into binary, causing unexpected and incorrect effects when compared with
5618 1. Very minor rewording of debugging text in manualroute to say "list of
5619 hosts" instead of "hostlist".
5621 2. If verify=header_syntax was set, and a header line with an unqualified
5622 address (no domain) and a large number of spaces between the end of the
5623 name and the colon was received, the reception process suffered a buffer
5624 overflow, and (when I tested it) crashed. This was caused by some obsolete
5625 code that should have been removed. The fix is to remove it!
5627 3. When running in the test harness, delay a bit after writing a bounce
5628 message to get a bit more predictability in the log output.
5630 4. Added a call to search_tidyup() just before forking a reception process. In
5631 theory, someone could use a lookup in the expansion of smtp_accept_max_
5632 per_host which, without the tidyup, could leave open a database connection.
5634 5. Added the variables $recipient_data and $sender_data which get set from a
5635 lookup success in an ACL "recipients" or "senders" condition, or a router
5636 "senders" option, similar to $domain_data and $local_part_data.
5638 6. Moved the writing of debug_print from before to after the "senders" test
5641 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5642 problems for message scanning, either using a data ACL, or using
5643 local_scan() because the Received: header was not generated till after they
5644 were called (in order to set the time as the time of reception completion).
5645 I have revised the way this works. The header is now generated after the
5646 body is received, but before the ACL or local_scan() are called. After they
5647 are run, the timestamp in the header is updated.
5653 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5654 before starting a queue runner without re-exec. This happened only when
5655 deliver_drop_privilege was set or when the Exim user was set to root. The
5656 effect of the bug was that timeouts during subsequent deliveries caused
5657 crashes instead of being properly handled. The handler is now left at its
5658 default (and expected) setting.
5660 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5661 message, again when deliver_drop_privilege is set or Exim is run as root.
5662 The bug described in (1) was not present in this case, but the tidying up
5663 of the other signals was missing. I have made the two cases consistent.
5665 3. The ignore_target_hosts setting on a manualroute router was being ignored
5666 for hosts that were looked up using the /MX notation.
5668 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5671 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5672 operated on the sender address. After changing the $sender_address to <>
5673 for the sender address verify, Exim was re-instated it as the original
5674 (before rewriting) address, but remembering that it had rewritten it, so it
5675 wasn't rewriting it again. This bug also had the effect of breaking the
5676 sender address verification caching when the sender address was rewritten.
5678 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5679 This has been changed so that if the ip literal address matches
5680 ignore_target_hosts, the router declines.
5682 7. Added expansion conditions match_domain, match_address, and match_local_
5683 part (NOT match_host).
5685 8. The placeholder for the Received: header didn't have a length field set.
5687 9. Added code to Exim itself and to exim_lock to test for a specific race
5688 condition that could lead to file corruption when using MBX delivery. The
5689 issue is with the lockfile that is created in /tmp. If this file is removed
5690 after a process has opened it but before that process has acquired a lock,
5691 there is the potential for a second process to recreate the file and also
5692 acquire a lock. This could lead to two Exim processes writing to the file
5693 at the same time. The added code performs the same test as UW imapd; it
5694 checks after acquiring the lock that its file descriptor still refers to
5695 the same named file.
5697 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5698 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5701 11. Added the smtp_active_hostname option. If used, this will typically be made
5702 to depend on the incoming interface address. Because $interface_address is
5703 not set up until the daemon has forked a reception process, error responses
5704 that can happen earlier (such as "too many connections") no longer contain
5707 12. If an expansion in a condition on a "warn" statement fails because a lookup
5708 defers, the "warn" statement is abandoned, and the next ACL statement is
5709 processed. Previously this caused the whole ACL to be aborted.
5711 13. Added the iplsearch lookup type.
5713 14. Added ident_timeout as a log selector.
5715 15. Added tls_certificate_verified as a log selector.
5717 16. Added a global option tls_require_ciphers (compare the smtp transport
5718 option of the same name). This controls incoming TLS connections.
5720 17. I finally figured out how to make tls_require_ciphers do a similar thing
5721 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5722 before starting the TLS session.
5724 18. Tabs are now shown as \t in -bP output.
5726 19. If the log selector return_path_on_delivery was set, Exim crashed when
5727 bouncing a message because it had too many Received: header lines.
5729 20. If two routers both had headers_remove settings, and the first one included
5730 a superfluous trailing colon, the final name in the first list and the
5731 first name in the second list were incorrectly joined into one item (with a
5732 colon in the middle).
5738 1. Added -C and -D options to the exinext utility, mainly to make it easier
5739 to include in the automated testing, but these could be helpful when
5740 multiple configurations are in use.
5742 2. The exinext utility was not formatting the output nicely when there was
5743 an alternate port involved in the retry record key, nor when there was a
5744 message id as well (for retries that were specific to a specific message
5745 and a specific host). It was also confused by IPv6 addresses, because of
5746 the additional colons they contain. I have fixed the IPv4 problem, and
5747 patched it up to do a reasonable job for IPv6.
5749 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5750 delivery, the log line now contains "pipelined" if PIPELINING was used.
5752 4. An SMTP transport process used to panic and die if the bind() call to set
5753 an explicit outgoing interface failed. This has been changed; it is now
5754 treated in the same way as a connect() failure.
5756 5. A reference to $sender_host_name in the part of a conditional expansion
5757 that was being skipped was still causing a DNS lookup. This no longer
5760 6. The def: expansion condition was not recognizing references to header lines
5761 that used bh_ and bheader_.
5763 7. Added the _cache feature to named lists.
5765 8. The code for checking quota_filecount in the appendfile transport was
5766 allowing one more file than it should have been.
5768 9. For compatibility with Sendmail, the command line option
5776 and sets the incoming protocol and host name (for trusted callers). The
5777 host name and its colon can be omitted when only the protocol is to be set.
5778 Note the Exim already has two private options, -pd and -ps, that refer to
5779 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5780 "s", but I don't think that's a major issue.
5782 10. A number of refactoring changes to the code, none of which should affect
5785 (a) The number of logging options was getting close to filling up the
5786 32-bit word that was used as a bit map. I have split them into two classes:
5787 those that are passed in the argument to log_write(), and those that are
5788 only ever tested independently outside of that function. These are now in
5789 separate 32-bit words, so there is plenty of room for expansion again.
5790 There is no change in the user interface or the logging behaviour.
5792 (b) When building, for example, log lines, the code previously used a
5793 macro that called string_cat() twice, in order to add two strings. This is
5794 not really sufficiently general. Furthermore, there was one instance where
5795 it was actually wrong because one of the argument was used twice, and in
5796 one call a function was used. (As it happened, calling the function twice
5797 did not affect the overall behaviour.) The macro has been replaced by a
5798 function that can join an arbitrary number of extra strings onto a growing
5801 (c) The code for expansion conditions now uses a table and a binary chop
5802 instead of a serial search (which was left over from when there were very
5803 few conditions). Also, it now recognizes conditions like "pam" even when
5804 the relevant support is not compiled in: a suitably worded error message is
5805 given if an attempt is made to use such a condition.
5807 11. Added ${time_interval:xxxxx}.
5809 12. A bug was causing one of the ddress fields not to be passed back correctly
5810 from remote delivery subprocesses. The field in question was not being
5811 subsequently used, so this caused to problems in practice.
5813 13. Added new log selectors queue_time and deliver_time.
5815 14. Might have fixed a bug in maildirsizefile handling that threw up
5816 "unexpected character" debug warnings, and recalculated the data
5817 unnecessarily. In any case, I expanded the warning message to give more
5820 15. Added the message "Restricted characters in address" to the statements in
5821 the default ACL that block characters like @ and % in local parts.
5823 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5824 Three changes have been made:
5826 (a) There was a serious bug; a negative response to MAIL caused the whole
5827 recipient domain to be cached as invalid, thereby blocking all messages
5828 to all local parts at the same domain, from all senders. This bug has
5829 been fixed. The domain is no longer cached after a negative response to
5830 MAIL if the sender used is not empty.
5832 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5835 (c) A new callout option, "use_sender" has been added for people who want
5836 the modified behaviour.
5842 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5845 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5846 indeed breaks things for older releases.
5848 3. Added additional logging to the case where there is a problem reading data
5849 from a filter that is running in a subprocess using a pipe, in order to
5850 try to track down a specific problem.
5852 4. Testing facility fudge: when running in the test harness and attempting
5853 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5854 getting "No route to host". Convert this to a timeout.
5856 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5859 6. Some OS don't have socklen_t but use size_t instead. This affects the
5860 fifth argument of getsockopt() amongst other things. This is now
5861 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5862 can be set for individual OS. I have set it for SunOS5, OSF1, and
5863 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5864 some earlier ones do not.
5866 7. Change 4.30/15 was not doing the test caselessly.
5868 8. The standard form for an IPv6 address literal was being rejected by address
5869 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5870 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5871 this, as well as the form without the "IPv6" on the front (but only when
5872 address literals are enabled, of course).
5874 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5876 10. Exim crashed if a message with an empty sender address specified by -f
5877 encountered a router with an errors_to setting. This could be provoked only
5878 by a command such as
5882 where an empty string was supplied; "<>" did not hit this bug.
5884 11. Installed PCRE release 4.5.
5886 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5887 remained set. It is now erased.
5889 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5890 times from message ids (which are base 36 rather than the normal 62).
5892 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5893 were being counted as actual protocol errors, and logged if the log
5894 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5895 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5896 and DATA following a set of rejected RCPTs do not count as protocol errors.
5897 In other words, Exim assumes they were pipelined, though this may not
5898 actually be the case. Of course, in all cases the client gets an
5899 appropriate error code.
5901 15. If a lookup fails in an ACL condition, a message about the failure may
5902 be available; it is used if testing the ACL cannot continue, because most
5903 such messages specify what the cause of the deferral is. However, some
5904 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5905 that caused an old message to be retained and used if a later statement
5906 caused a defer, replacing the real cause of the deferral.
5908 16. If an IP address had so many PTR records that the DNS lookup buffer
5909 was not large enough to hold them, Exim could crash while trying to process
5910 the truncated data. It now detects and logs this case.
5912 17. Further to 4.21/58, another change has been made: if (and only if) the
5913 first line of a message (the first header line) ends with CRLF, a bare LF
5914 in a subsequent header line has a space inserted after it, so as not to
5915 terminate the header.
5917 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5918 unnecessarily, used atoi() instead of strtol(), and didn't check the
5919 termination when getting file sizes from file names by regex.
5921 19. Completely re-implemented the support for maildirsize files, in the light
5922 of a number of problems with the previous contributed implementation
5923 (4.30/29). In particular:
5925 . If the quota is zero, the maildirsize file is maintained, but no quota is
5928 . If the maildir directory does not exist, it is created before any attempt
5929 to write a maildirsize file.
5931 . The quota value in the file is just a cache; if the quota is changed in
5932 the transport, the new value overrides.
5934 . A regular expression is available for excluding directories from the
5937 20. The autoreply transport checks the characters in options that define the
5938 message's headers; it allows continued headers, but it was checking with
5939 isspace() after an embedded newline instead of explicitly looking for a
5942 21. If all the "regular" hosts to which an address was routed had passed their
5943 expiry times, and had not reached their retry times, the address was
5944 bounced, even if fallback hosts were defined. Now Exim should go on to try
5947 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5948 equivalent code in the SMTP transport. Some hosts send humungous responses
5949 to HELO/EHLO, more than 1024 it seems.
5951 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5952 gives compiler warnings in some environments. I've now done it "properly",
5955 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5956 (because of problems with the built-in one) was declared to return uschar *
5957 instead of char *, causing compiler failure.
5959 25. Fixed a file descriptor leak when processing alias/forward files.
5961 26. Fixed a minor format string issue in dbfn.c.
5963 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5965 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5966 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5969 29. When a sender address is verified, it is cached, to save repeating the test
5970 when there is more than one recipient in a message. However, when the
5971 verification involves a callout, it is possible for different callout
5972 options to be set for different recipients. It is too complicated to keep
5973 track of this in the cache, so now Exim always runs a verification when a
5974 callout is required, relying on the callout cache for the optimization.
5975 The overhead is duplication of the address routing, but this should not be
5978 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5979 to be verified with callout=postmaster, and the main callout worked but the
5980 postmaster check failed, the verification correctly failed. However, if a
5981 subsequent RCPT command asked for sender verification *without* the
5982 postmaster check, incorrect caching caused this verification also to fail,
5985 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5986 it was not caching the DNS options (qualify_single, search_parents) that
5987 were used when the lookup failed. A subsequent lookup with different
5988 options therefore always gave the same answer, though there were cases
5989 where it should not have. (Example: a "domains = !$mx_any" option on a
5990 dnslookup router: the "domains" option is always processed without any
5991 widening, but the router might have qualify_single set.) Now Exim uses the
5992 cached value only when the same options are set.
5994 32. Added John Jetmore's "exipick" utility to the distribution.
5996 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5997 than a timeout (e.g. a certificate is required, and is not provided), an
5998 Exim server now closes the connection immediately. Previously it waited for
5999 the client to close - but if the client is SSL, it seems that they each
6000 wait for each other, leading to a delay before one of them times out.
6002 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6003 maintained 0.8.x compatibility because I don't think many are using it, and
6004 it is clearly obsolete.
6006 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6009 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6010 client certificate was expired. A simple patch fixes this, though I don't
6011 understand the full logic of why the verify callback is called multiple
6014 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6015 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6016 which causes problems with some clients (such as the Certicom SSL Plus
6017 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6018 disables the coutermeasure allowing Eudora to connect."
6020 38. Exim was not checking that a write() to a log file succeeded. This could
6021 lead to Bad Things if a log got too big, in particular if it hit a file
6022 size limit. Exim now panics and dies if it cannot write to a log file, just
6023 as it does if it cannot open a log file.
6025 39. Modified OS/Makefile-Linux so that it now contains
6027 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6029 The two -D definitions ensure that Exim is compiled with large file
6030 support, which makes it possible to handle log files that are bigger than
6033 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6034 instance) a domain was checked against a named list that involved a lookup,
6035 causing $domain_data to be set, then another domain was checked against the
6036 same list, then the first domain was re-checked, the value of $domain_data
6037 after the final check could be wrong. In particular, if the second check
6038 failed, it could be set empty. This bug probably also applied to
6041 41. The strip_trailing_dot option was not being applied to the address given
6042 with the -f command-line option.
6044 42. The code for reading a message's header from the spool was incrementing
6045 $received_count, but never initializing it. This meant that the value was
6046 incorrect (doubled) while delivering a message in the same process in which
6047 it was received. In the most common configuration of Exim, this never
6048 happens - a fresh exec is done - but it can happen when
6049 deliver_drop_privilege is set.
6051 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6052 it now includes up to 150 characters of the unexpected data in the log
6055 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6056 and building data strings. The size of both of these buffers was 10 000
6057 bytes - far larger than anybody would *ever* want, thought I. Needless to
6058 say, somebody hit the limit. I have increased the maximum line length to
6059 20 000 and the maximum data length of concatenated lines to 100 000. I have
6060 also fixed two bugs, because there was no checking on these buffers. Tsk,
6061 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6062 buffer is too small.
6064 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6065 lsearch lookups. Now it does.
6067 46. When parsing a route_list item in a manualroute router, a fixed-length
6068 buffer was used for the list of hosts. I made this 1024 bytes long,
6069 thinking that nobody would ever have a list of hosts that long. Wrong.
6070 Somebody had a whole pile of complicated expansion conditions, and the
6071 string was silently truncated, leading to an expansion error. It turns out
6072 that it is easier to change to an unlimited length (owing to other changes
6073 that have happened since this code was originally written) than to build
6074 structure for giving a limitation error. The length of the item that
6075 expands into the list of hosts is now unlimited.
6077 47. The lsearch lookup could not handle data where the length of text line was
6078 more than 4095 characters. Such lines were truncated, leading to shortened
6079 data being returned. It should now handle lines of any length.
6081 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6082 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6085 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6086 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6087 no longer happens in comment lines. A list of replacements is now placed
6088 at the head of all of the source files, except those whose only change is
6089 to replace PERL_COMMAND in the very first #! line.
6091 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6092 messages on the queue, with a bottom-up merge sort, using code contributed
6093 by Michael Haardt. This should make operations like -bp somewhat faster on
6094 large queues. It won't affect queue runners, except when queue_run_in_order
6097 51. Installed eximstats 1.31 in the distribution.
6099 52. Added support for SRV lookups to the dnslookup router.
6101 53. If an ACL referred to $message_body or $message_body_end, the value was not
6102 reset for any messages that followed in the same SMTP session.
6104 54. The store-handling optimization for building very long strings was not
6105 differentiating between the different store pools. I don't think this
6106 actually made any difference in practice, but I've tidied it.
6108 55. While running the routers to verify a sender address, $sender_address
6109 was still set to the sender address. This is wrong, because when routing to
6110 send a bounce to the sender, it would be empty. Therefore, I have changed
6111 it so that, while verifying a sender address, $sender_address is set to <>.
6112 (There is no change to what happens when verifying a recipient address.)
6114 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6115 target A or AAAA records (if not already returned) without resetting the
6116 qualify_single or search_parents options of the DNS resolver. These are
6117 inappropriate in this case because the targets of MX and SRV records must
6118 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6119 target that, when qualified, matched something in the local domain. These
6120 two options are now turned off when doing these lookups.
6122 57. It seems that at least some releases of Reiserfs (which does not have the
6123 concept of a fixed number of inodes) returns zero and not -1 for the
6124 number of available inodes. This interacted badly with check_spool_inodes,
6125 which assumed that -1 was the "no such thing" setting. What I have done is
6126 to check that the total number of inodes is greater than zero before doing
6127 the test of how many are available.
6129 58. When a "warn" ACL statement has a log_message modifier, the message is
6130 remembered, and not repeated. This is to avoid a lot of repetition when a
6131 message has many recipients that cause the same warning to be written.
6132 However, Exim was preserving the list of already written lines for an
6133 entire SMTP session, which doesn't seem right. The memory is now reset if a
6134 new message is started.
6136 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6137 debugging output unless log_rewrite was also set.
6139 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6140 of (int)(handle) when we know that handle contains (void *)(-1).
6142 61. The Exim daemon panic-logs an error return when it closes the incoming
6143 connection. However "connection reset by peer" seems to be common, and
6144 isn't really an error worthy of noting specially, so that particular error
6147 62. When Exim is trying to find all the local interfaces, it used to panic and
6148 die if the ioctl to get the interface flags failed. However, it seems that
6149 on at least one OS (Solaris 9) it is possible to have an interface that is
6150 included in the list of interfaces, but for which you get a failure error
6151 for this call. This happens when the interface is not "plumbed" into a
6152 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6153 failure of the "get flags" call assumes that the interface is down.
6155 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6156 makes life easier for people who are doing arithmetic on fields extracted
6157 from dates, where you often get leading zeros that should not be
6158 interpreted as octal.
6160 64. Added qualify_domain to the redirect router, to override the global
6163 65. If a pathologically long header line contained very many addresses (the
6164 report of this problem mentioned 10 000) and each of them was rewritten,
6165 Exim could use up a very large amount of memory. (It kept on making new
6166 copies of the header line as it rewrote, and never released the old ones.)
6167 At the expense of a bit more processing, the header rewriting function has
6168 been changed so that it no longer eats memory in this way.
6170 66. The generation of the Received: header has been moved from the time that a
6171 message starts to be received, to the time that it finishes. The timestamp
6172 in the Received: header should now be very close to that of the <= log
6173 line. There are two side-effects of this change:
6175 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6176 logged header lines no longer include the local Received: line, because
6177 it has not yet been created. The same applies to a copy of the message
6178 that is returned to a non-SMTP sender when a message is rejected.
6180 (b) When a filter file is tested using -bf, no additional Received: header
6181 is added to the test message. After some thought, I decided that this
6184 This change does not affect the value of $received_for. It is still set
6185 after address rewriting, but before local_scan() is called.
6187 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6189 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6190 gave an unhelpful panic error message, and a defer error. I have managed to
6191 change this behaviour so that it now rejects any supplied certificate,
6192 which seems right, as the list of acceptable certificates is empty.
6194 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6195 gave an unhelpful defer error. I have not managed to make this reject any
6196 supplied certificates, but the error message it gives is "no certificate
6197 supplied", which is not helpful.
6199 70. exigrep's output now also includes lines that are not associated with any
6200 message, but which match the given pattern. Implemented by a patch from
6201 Martin Sluka, which also tidied up the Perl a bit.
6203 71. Recipient callout verification, like sender verification, was using <> in
6204 the MAIL FROM command. This isn't really the right thing, since the actual
6205 sender may affect whether the remote host accepts the recipient or not. I
6206 have changed it to use the actual sender in the callout; this means that
6207 the cache record is now keyed on a recipient/sender pair, not just the
6208 recipient address. There doesn't seem to be a real danger of callout loops,
6209 since a callout by the remote host to check the sender would use <>.
6210 [SEE ABOVE: changed after hitting problems.]
6212 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6213 temporary errors. However, in the case of such a code being given after
6214 the end of a data transmission (i.e. after ".") Exim was failing to write
6215 a retry record for the message. (Yes, there was some broken host that was
6216 actually sending 8xx at this point.)
6218 73. An unknown lookup type in a host list could cause Exim to panic-die when
6219 the list was checked. (An example that provoked this was putting <; in the
6220 middle of a list instead of at the start.) If this happened during a DATA
6221 ACL check, a -D file could be left lying around. This kind of configuration
6222 error no longer causes Exim to die; instead it causes a defer error. The
6223 incident is still logged to the main and panic logs.
6225 74. Buglet left over from Exim 3 conversion. The message "too many messages
6226 in one connection" was written to the rejectlog but not the mainlog, except
6227 when address rewriting (yes!) was being logged.
6229 75. Added write_rejectlog option.
6231 76. When a system filter was run not as root (that is, when system_filter_user
6232 was set), the values of the $n variables were not being returned to the
6233 main process; thus, they were not subsequently available in the $sn
6236 77. Added +return_path_on_delivery log selector.
6238 78. A connection timeout was being treated differently from recipients deferred
6239 when testing hosts_max_try with a message that was older than the host's
6240 retry timeout. (The host should not be counted, thus allowing all hosts to
6241 be tried at least once before bouncing.) This may have been the cause of an
6242 occasionally reported bug whereby a message would remain on the queue
6243 longer than the retry timeout, but would be bounced if a delivery was
6244 forced. I say "may" because I never totally pinned down the problem;
6245 setting up timeout/retry tests is difficult. See also the next item.
6247 79. The ultimate address timeout was not being applied to errors that involved
6248 a combination of host plus message (for example, a timeout on a MAIL
6249 command). When an address resolved to a number of possible hosts, and they
6250 were not all tried for each delivery (e.g. because of hosts_max_try), a
6251 message could remain on the queue longer than the retry timeout.
6253 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6256 81. Fixed an obscure SMTP outgoing bug which required at least the following
6257 conditions: (a) there was another message waiting for the same server;
6258 (b) the server returned 5xx to all RCPT commands in the first message so
6259 that the message was not completed; (c) the server dropped the connection
6260 or gave a negative response to the RSET that Exim sends to abort the
6261 transaction. The observed case was a dropped connection after DATA that had
6262 been sent in pipelining mode. That is, the server had advertised PIPELINING
6263 but was not implementing it correctly. The effect of the bug was incorrect
6264 behaviour, such as trying another host, and this could lead to a crash.
6270 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6271 and daemon.c were passed as pointers to ints; they should have been
6272 pointers to socklen_t variables (which are typically unsigned ints).
6274 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6277 3. Fixed a really odd bug that affected only the testing scheme; patching a
6278 certain fixed string in the binary changed the value of another string that
6279 happened to be identical to the end of the original first string.
6281 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6282 name", it returns that address as the IP address. On some operating
6283 systems (e.g. Solaris), it also passes back the IP address string as the
6284 "host name". However, on others (e.g. Linux), it passes back an empty
6285 string. Exim wasn't checking for this, and was changing the host name to an
6286 empty string, assuming it had been canonicalized.
6288 5. Although rare, it is permitted to have more than one PTR record for a given
6289 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6290 all the names associated with an address, because they do in Solaris.
6291 However, it seems that they do not in Linux for data that comes from the
6292 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6293 I found this out when I moved to a new Linux workstation and tried to run
6294 the Exim test suite.
6296 To get round this problem I have changed the code so that it now does its
6297 own call to the DNS to look up PTR records when searching for a host name.
6298 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6299 addresses that are only in /etc/hosts are still found.
6301 This behaviour is, however, controlled by an option called host_lookup_
6302 order, which defaults to "bydns:byaddr". If people want to use the other
6303 order, or indeed, just use one or the other means of lookup, they can
6304 specify it in this variable.
6306 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6307 some operating systems, this comes back from gethostbyaddr() as an empty
6308 string, and this is what Exim used to test for. However, it seems that in
6309 other systems, "." is yielded. Exim now tests for this case too.
6311 7. The values of check_spool_space and check_log_space are now held internally
6312 as a number of kilobytes instead of an absolute number of bytes. If a
6313 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6314 kilobyte. This means that much larger values can be stored.
6316 8. Exim monitor: an attempt to get the action menu when not actually pointing
6317 at a message produces an empty menu entitled "No message selected". This
6318 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6319 no entries in it ("Shell widget menu has zero width and/or height"). So I
6320 have added a single, blank menu entry in this case.
6322 9. Added ${quote_local_part.
6324 10. MIME decoding is now applied to the contents of Subject: header lines when
6327 11. Now that a reference to $sender_host_address automatically causes a reverse
6328 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6329 host lookup before query-style lookups in lists that might use this
6330 variable. This has therefore been abolished, and the "net-" prefix is no
6331 longer necessary for query-style lookups.
6333 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6334 have been a typo for LFLAGS, so it has been changed.
6336 13. The install script calls Exim with "-C /dev/null" in order to find the
6337 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6338 to be output. However, since Exim outputs its version number before the
6339 error, it didn't break the script. It just looked ugly. I fixed this by
6340 always allowing "-C /dev/null" if the caller is root.
6342 14. Ignore overlarge ACL variable number when reading spool file - insurance
6343 against a later release with more variables having written the file.
6345 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6346 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6347 this, as well as the form without the "IPv6" on the front.
6349 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6350 OS/Makefile-Darwin file.
6352 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6353 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6355 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6356 variable. However, some OS use ints for inodes. I've added cast to long int
6357 to get rid of the compiler warning.
6359 19. I had forgotten to lock out "/../" in configuration file names when
6360 ALT_CONFIG_PREFIX was set.
6362 20. Routers used for verification do not need to specify transports. However,
6363 if such a router generated a host list, and callout was configured, Exim
6364 crashed, because it could not find a port number from the (non-existent)
6365 transport. It now assumes port 25 in this circumstance.
6367 21. Added the -t option to exigrep.
6369 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6370 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6371 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6372 them is set, LOOKUP_LSEARCH is forced.
6374 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6376 24. Added sender and host information to the "rejected by local_scan()" log
6377 line; previously there was no indication of these.
6379 25. Added .include_if_exists.
6381 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6382 Linux. It turns out that not all file systems support this. Apparently some
6383 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6384 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6385 not supported on the file descriptor, is now ignored when Exim is trying to
6386 sync a directory. This applies only to Linux.
6388 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6390 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6391 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6394 29. Added support for maildirsize files from supplied patch (modified a bit).
6396 30. The use of :fail: followed by an empty string could lead Exim to respond to
6397 sender verification failures with (e.g.):
6399 550 Verification failed for <xxx>
6400 550 Sender verify failed
6402 where the first response line was missing the '-' that indicates it is not
6403 the final line of the response.
6405 31. The loop for finding the name of the user that called Exim had a hardwired
6406 limit of 10; it now uses the value of finduser_retries, which is used for
6407 all other user lookups.
6409 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6412 33. Exim was neglecting to zero errno before one call of strtol() when
6413 expanding a string and expecting an integer value. On some systems this
6414 resulted in spurious "integer overflow" errors. Also, it was casting the
6415 result into an int without checking.
6417 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6418 did not work. The code looks as if it has *never* worked, though it appears
6419 to have been documented since at least release 1.62. I have made it work.
6421 35. The "timeout_DNS" error in retry rules, also documented since at least
6422 1.62, also never worked. As it isn't clear exactly what this means, and
6423 clearly it isn't a major issue, I have abolished the feature by treating it
6424 as "timeout", and writing a warning to the main and panic logs.
6426 36. The display of retry rules for -brt wasn't always showing the error code
6429 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6430 timeout_connect_A, timeout_connect_MX.
6432 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6433 to the empty sender.
6435 39. The daemon was not analysing the content of -oX till after it had closed
6436 stderr and disconnected from the controlling terminal. This meant that any
6437 syntax errors were only noted on the panic log, and the return code from
6438 the command was 0. By re-arranging the code a little, I've made the
6439 decoding happen first, so such errors now appear on stderr, and the return
6440 code is 1. However, the actual setting up of the sockets still happens in
6441 the disconnected process, so errors there are still only recorded on the
6444 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6445 connections (as happens on some IP stacks) was logged at start up time as
6446 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6447 it from "IPv6 and IPv4", which means that two separate sockets are being
6450 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6451 says whether AF_INET or AF_INET6 was passed as an argument.
6453 42. Exiwhat output was messed up when time zones were included in log
6456 43. Exiwhat now gives more information about the daemon's listening ports,
6457 and whether -tls-on-connect was used.
6459 44. The "port" option of the smtp transport is now expanded.
6461 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6462 silently ignored. Now an error message is written to the main and panic
6465 46. There's a new ACL modifier called "logwrite" which writes to a log file
6466 as soon as it is encountered.
6468 47. Added $local_user_uid and $local_user_gid at routing time.
6470 48. Exim crashed when trying to verify a sender address that was being
6473 49. Exim was recognizing only a space character after ".include". It now also
6474 recognizes a tab character.
6476 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6477 extracting the relevant information from the specification. The man page no
6478 longer contains scrambled data for the -d option, and I've added a section
6479 at the front about calling Exim under different names.
6481 51. Added "extra_headers" argument to the "mail" command in filter files.
6483 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6486 53. Installed eximstats 1.29.
6488 54. Added transport_filter_timeout as a generic transport option.
6490 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6491 Cc: header lines. This was required by RFC 822, but it not required by RFC
6494 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6495 incoming messages that did not have them. Now it does so only if the
6496 message originates locally, that is, if there is no associated remote host
6497 address. When Resent- header lines are present, this applies to the Resent-
6498 lines rather than the non-Resent- lines.
6500 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6501 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6503 58. Messages for configuration errors now include the name of the main
6504 configuration file - useful now that there may be more than one file in a
6505 list (.included file names were always shown).
6507 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6508 for those rare installations that do not start the daemon as root or run it
6509 setuid root. I've cut out the call to initgroups() if the daemon is not
6512 60. The Exim user and group can now be bound into the binary as text strings
6513 that are looked up at the start of Exim's processing.
6515 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6517 62. Added $mailstore_basename variable.
6519 63. Installed patch to sieve.c from Michael Haardt.
6521 64. When Exim failed to open the panic log after failing to open the main log,
6522 the original message it was trying to log was written to stderr and debug
6523 output, but if they were not available (the usual case in production), it
6524 was lost. Now it is written to syslog before the two lines that record the
6525 failures to open the logs.
6527 65. Users' Exim filters run in subprocesses under the user's uid. It is
6528 possible for a "deliver" command or an alias in a "personal" command to
6529 provoke an address rewrite. If logging of address rewriting is configured,
6530 this fails because the process is not running as root or exim. There may be
6531 a better way of dealing with this, but for the moment (because 4.30 needs
6532 to be released), I have disabled address rewrite logging when running a
6533 filter in a non-root, non-exim process.
6539 1. The buildconfig auxiliary program wasn't quoting the value set for
6540 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6541 not defined. This bug was masked in 4.22 by the effect that was fixed in
6544 2. Some messages that were rejected after a message id was allocated were
6545 shown as "incomplete" by exigrep. It no longer does this for messages that
6546 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6548 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6549 have allow_domain_literals set, the ID did not get logged in the <= line.
6550 Domain literals are now always recognized in Message-ID: header lines.
6552 4. The first argument for a ${extract expansion item is the key name or field
6553 number. Leading and trailing spaces in this item were not being ignored,
6554 causing some misleading effects.
6556 5. When deliver_drop_privilege was set, single queue runner processes started
6557 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6558 same command in the process it spins off) were not dropping privilege.
6560 6. When the daemon running as "exim" started a queue runner, it always
6561 re-executed Exim in the spun-off process. This is a waste of effort when
6562 deliver_drop_privilege is set. The new process now just calls the
6563 queue-runner function directly.
6569 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6572 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6573 ignored. Though the use of -oP was forcing the writing of a pid file, it
6574 was always written to the default place.
6576 3. If the message "no IP address found for host xxxx" is generated during
6577 incoming verification, it is now followed by identification of the incoming
6578 connection (so you can more easily find what provoked it).
6580 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6582 5. Added some features to "harden" Exim a bit more against certain attacks:
6584 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6585 be put in Local/Makefile. This is like the never_users runtime option,
6586 but it cannot be overridden. The default setting is "root".
6588 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6589 prefix string with which any file named in a -C command line option
6592 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6593 is retained for -C and -D only if the caller of Exim is root. Without
6594 it, the exim user may also use -C and -D and retain privilege.
6596 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6597 command line option is disabled.
6599 6. Macro names set by the -D option must start with an upper case letter, just
6600 like macro names defined in the configuration file.
6602 7. Added "dereference=" facility to LDAP.
6604 8. Two instances of the typo "uknown" in the source files are fixed.
6606 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6607 the Configure-Makefile script screwed up while processing it.
6609 10. Incorporated PCRE 4.4.
6611 11. The SMTP synchronization check was not operating right at the start of an
6612 SMTP session. For example, it could not catch a HELO sent before the client
6613 waited for the greeting. There is now a check for outstanding input at the
6614 point when the greeting is written. Because of the duplex, asynchronous
6615 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6616 way, but not yet received, when the check is performed.
6618 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6619 on TCP/IP sockets, because this apparently causes some broken clients to
6622 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6623 unchanged) from the Cygwin maintainer.
6625 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6626 was supported instead of testing for mbx. Effectively a typo.
6628 15. The spa authenticator server code was not checking that the input it
6629 received was valid base64.
6631 16. The debug output line for the "set" modifier in ACLs was not showing the
6632 name of the variable that was being set.
6634 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6636 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6637 lookup on its own. Something else was needed to trigger the lookup. For
6638 example, a match in host_lookup or the need for a host name in a host list.
6639 Now, if $sender_host_name is referenced and the host name has not yet been
6640 looked up, a lookup is performed. If the lookup fails, the variable remains
6641 empty, and $host_lookup_failed is set to "1".
6643 19. Added "eqi" as a case-independent comparison operator.
6645 20. The saslauthd authentication condition could segfault if neither service
6646 nor realm was specified.
6648 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6649 error message that was logged was misleading, and incoming SMTP
6650 connections were dropped. The message is now more accurate, and temporary
6651 errors are given to SMTP connections.
6653 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6654 (except RSET) with a 421 error, until QUIT is received. However, it was
6655 failing to send a response to QUIT.
6657 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6658 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6659 ACL is now run after the helo_try_verify_hosts code.
6661 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6662 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6663 case-independent, so other case variants are also recognized). Apparently
6664 some systems use these upper case variants.
6666 25. If more than two messages were waiting for the same host, and a transport
6667 filter was specified for the transport, Exim sent two messages over the
6668 same TCP/IP connection, and then failed with "socket operation on non-
6669 socket" when it tried to send the third.
6671 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6673 27. The extern definition of crypt16() in expand.c was not being excluded when
6674 the OS had its own crypt16() function.
6676 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6677 as a preferred synonym for return_size_limit, both as an option and as an
6680 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6682 30. Changed the default configuration ACL to relax the local part checking rule
6683 for addresses that are not in any local domains. For these addresses,
6684 slashes and pipe symbols are allowed within local parts, but the sequence
6685 /../ is explicitly forbidden.
6687 31. SPA server authentication was not clearing the challenge buffer before
6690 32. log_message in a "warn" ACL statement was writing to the reject log as
6691 well as to the main log, which contradicts the documentation and doesn't
6692 seem right (because no rejection is happening). So I have stopped it.
6694 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6695 However, I am unable to do any testing of this.
6697 34. Fixed an infelicity in the appendfile transport. When checking directories
6698 for a mailbox, to see if any needed to be created, it was accidentally
6699 using path names with one or more superfluous leading slashes; tracing
6700 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6702 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6703 contents are added to the log line that is written for every discarded
6704 recipient. (Previously a log_message setting was ignored.)
6706 36. The ${quote: operator now quotes the string if it is empty.
6708 37. The install script runs exim in order to find its version number. If for
6709 some reason other than non-existence or emptiness, which it checks, it
6710 could not run './exim', it was installing it with an empty version number,
6711 i.e. as "exim-". This error state is now caught, and the installation is
6714 38. An argument was missing from the function that creates an error message
6715 when Exim fails to connect to the socket for saslauthd authentication.
6716 This could cause Exim to crash, or give a corrupted message.
6718 39. Added isip, isip4, and isip6 to ${if conditions.
6720 40. The ACL variables $acl_xx are now saved with the message, and can be
6721 accessed later in routers, transports, and filters.
6723 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6724 strings in the file are not string-expanded.
6726 42. If a MAIL command specified a SIZE value that was too large to fit into an
6727 int variable, the check against message_size_limit failed. Such values are
6728 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6729 day this will have to be increased, but I don't think I want to be around
6730 when emails are that large.
6737 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6738 iconv() is not standard in FreeBSD.
6740 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6741 IPv6 enabled. The observed symptom was a segmentation fault on return from
6742 the function os_common_find_running_interfaces() in src/os.c.
6744 3. In the check_special_case() function in daemon.c I had used "errno" as an
6745 argument name, which causes warnings on some systems. This was basically a
6746 typo, since it was named "eno" in the comments!
6748 4. The code that waits for the clock to tick (at a resolution of some fraction
6749 of a second) so as to ensure message-id uniqueness was always waiting for
6750 at least one whole tick, when it could have waited for less. [This is
6751 almost certainly not relevant at current processor speeds, where it is
6752 unlikely to ever wait at all. But we try to future-proof.]
6754 5. The function that sleeps for a time interval that includes fractions of a
6755 second contained a race. It did not block SIGALRM between setting the
6756 timer, and suspending (a couple of lines later). If the interval was short
6757 and the sigsuspend() was delayed until after it had expired, the suspension
6758 never ended. On busy systems this could lead to processes getting stuck for
6761 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6762 process, before it forks any delivery processes. The open lookup caching
6763 mechanism meant that the open file or database connection was passed into
6764 the delivery process. The problem was that delivery processes always tidy
6765 up cached lookup data. This could cause a problem for the next delivery
6766 process started by the queue runner, because the external queue runner
6767 process does not know about the closure. So the next delivery process
6768 still has data in the lookup cache. In the case of a file lookup, there was
6769 no problem because closing a file descriptor in a subprocess doesn't affect
6770 the parent. However, if the lookup was caching a connection to a database,
6771 the connection was closed, and the second delivery process was likely to
6772 see errors such as "PGSQL: query failed: server closed the connection
6773 unexpectedly". The problem has been fixed by closing all cached lookups
6774 in a queue runner before running a delivery process.
6776 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6777 seem to have the "const" qualifier which it has on other OS. I've
6780 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6781 *of the same type* (client or server) with the same public name that an
6782 error should be diagnosed.
6784 9. When Exim looked up a host name for an IP address, but failed to find the
6785 original IP address when looking up the host name (a safety check), it
6786 output the message "<ip address> does not match any IP for NULL", which was
6787 confusing, to say the least. The bug was that the host name should have
6788 appeared instead of "NULL".
6790 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6791 user that is built into the binary, and the -C or -D options is used, root
6792 privilege is dropped before the configuration file is read. In addition,
6793 logging is switched to stderr instead of the normal log files. If the
6794 configuration then re-defines the Exim user, the unprivileged environment
6795 is probably not what is expected, so Exim logs a panic warning message (but
6798 However, if deliver_drop_privilege is set, the unprivileged state may well
6799 be exactly what is intended, so the warning has been cut out in that case,
6800 and Exim is allowed to try to write to its normal log files.
6806 1. smtp_return_error_details was not giving details for temporary sender
6807 or receiver verification errors.
6809 2. Diagnose a configuration error if two authenticators have the same public
6812 3. Exim used not to create the message log file for a message until the first
6813 delivery attempt. This could be confusing when incoming messages were held
6814 for policy or load reasons. The message log file is now created at the time
6815 the message is received, and an initial "Received" line is written to it.
6817 4. The automatically generated man page for command line options had a minor
6818 bug that caused no ill effects; however, a more serious problem was that
6819 the procedure for building the man page automatically didn't always
6820 operate. Consequently, release 4.20 contains an out-of-date version. This
6821 shouldn't happen again.
6823 5. When building Exim with embedded Perl support, the script that builds the
6824 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6825 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6827 6. The freeze_tell option was not being used for messages that were frozen on
6828 arrival, either by an ACL or by local_scan().
6830 7. Added the smtp_incomplete_transaction log selector.
6832 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6833 was accepting AUTH without a new EHLO.
6835 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6836 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6839 10. Logging of TCP/IP connections (when configured) now happens in the main
6840 daemon process instead of the child process, so that the TCP/IP connection
6841 count is more accurate (but it can never be perfect).
6843 11. The use of "drop" in a nested ACL was not being handled correctly in the
6844 outer ACL. Now, if condition failure induced by the nested "drop" causes
6845 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6846 or "require"), the connection is dropped.
6848 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6849 that yield "discard" can now be used with an "accept" or a "discard" verb,
6850 but an error is generated for any others (because I can't see a useful way
6851 to define what should happen).
6853 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6854 are now processed in the same way as lines in the Exim configuration file.
6855 In particular, continuation lines are supported.
6857 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6859 15. Added -ti meaning -t -i.
6861 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6862 domains, and warn by logging if others are found.
6864 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6865 returned from GIFCONF when Exim is trying to find the list of interfaces on
6866 a host. The code in os.c has been modified to copy each ifreq to an aligned
6867 structure in all cases.
6869 Also, in some cases, the returned ifreq's were being copied to a 'struct
6870 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6871 means the last couple of bytes of an IPv6 address could be chopped if the
6872 ifreq contained only a normal sockaddr (14 bytes storage).
6874 18. Named domain lists were not supported in the hosts_treat_as_local option.
6875 An entry such as +xxxx was not recognized, and was treated as a literal
6878 19. Ensure that header lines added by a DATA ACL are included in the reject log
6879 if the ACL subsequently rejects the message.
6881 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6882 MD5 (which is deprecated).
6884 21. When testing a filter file using -bf, Exim was writing a message when it
6885 took the sender from a "From " line in the message, but it was not doing so
6886 when it took $return_path from a Return-Path: header line. It now does.
6888 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6889 with a valid header line field name (a series of printing characters
6890 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6892 23. Changed "disc" in the source to "disk" to conform to the documentation and
6893 the book and for uniformity.
6895 24. Ignore Sendmail's -Ooption=value command line item.
6897 25. When execve() failed while trying to run a command in a pipe transport,
6898 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6899 could be confused with a return value of 69 from the command itself. This
6900 has been changed to 127, the value the shell returns if it is asked to run
6901 a non-existent command. The wording for the related log line suggests a
6902 non-existent command as the problem.
6904 26. If received_header_text expands to an empty string, do not add a Received:
6905 header line to the message. (Well, it adds a token one on the spool, but
6906 marks it "old" so that it doesn't get used or transmitted.)
6908 27. Installed eximstats 1.28 (addition of -nt option).
6910 28. There was no check for failure on the call to getsockname() in the daemon
6911 code. This can fail if there is a shortage of resources on the system, with
6912 ENOMEM, for example. A temporary error is now given on failure.
6914 29. Contrary to the C standard, it seems that in some environments, the
6915 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6916 program. Exim now does this explicitly; it affects the formatting of
6917 timestamps using strftime().
6919 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6920 complaints. I've now initialized all the variables, to avoid this.
6922 32. Header lines added by a system filter were not being "seen" during
6923 transport-time rewrites.
6925 33. The info_callback() function passed to OpenSSL is set up with type void
6926 (*)(SSL *, int, int), as described somewhere. However, when calling the
6927 function (actually a macro) that sets it up, the type void(*)() is
6928 expected. I've put in a cast to prevent warnings from picky compilers.
6930 34. If a DNS black list lookup found a CNAME record, but there were no A
6931 records associated with the domain it pointed at, Exim crashed.
6933 35. If a DNS black list lookup returned more than one A record, Exim ignored
6934 all but the first. It now scans all returned addresses if a particular IP
6935 value is being sought. In this situation, the contents of the
6936 $dnslist_value variable are a list of all the addresses, separated by a
6939 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6940 to accept a host name and all its aliases if the forward lookup for any of
6941 them yielded the IP address of the incoming connection. Now it accepts only
6942 those names whose forward lookup yields the correct IP address. Any other
6943 names are discarded. This closes a loophole whereby a rogue DNS
6944 administrator could create reverse DNS records to break through a
6945 wildcarded host restriction in an ACL.
6947 37. If a user filter or a system filter that ran in a subprocess used any of
6948 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6949 the wrong values were passed to the pipe command ($thisaddress had the
6950 value of $0, $0 had the value of $1, etc). This bug was introduced by
6951 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6953 38. Improved the line breaking for long SMTP error messages from ACLs.
6954 Previously, if there was no break point between 40 and 75 characters, Exim
6955 left the rest of the message alone. Two changes have been made: (a) I've
6956 reduced the minimum length to 35 characters; (b) if it can't find a break
6957 point between 35 and 75 characters, it looks ahead and uses the first one
6958 that it finds. This may give the occasional overlong line, but at least the
6959 remaining text gets split now.
6961 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6962 file descriptors might be low, and that setting 1000 would always raise it.
6963 It turns out that in some environments, the limit is already over 1000 and
6964 that lowering it causes trouble. So now Exim takes care not to decrease it.
6966 40. When delivering a message, the value of $return_path is set to $sender_
6967 address at the start of routing (routers may change the value). By an
6968 oversight, this default was not being set up when an address was tested by
6969 -bt or -bv, which affected the outcome if any router or filter referred to
6972 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6973 log only when "message" or "log_message" are set. However, if one of the
6974 conditions was an address verification, or a call to a nested ACL, the
6975 messages generated by the underlying test were being passed through. This
6976 no longer happens. The underlying message is available in $acl_verify_
6977 message for both "message" and "log_message" expansions, so it can be
6978 passed through if needed.
6980 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6981 new expansion $bh_ to give the encoded byte string without charset
6982 translation. Translation happens only if iconv() is available; HAVE_ICONV
6983 indicates this at build time. HEADERS_CHARSET gives the charset to
6984 translate to; headers_charset can change it in the configuration, and
6985 "headers charset" can change it in an individual filter file.
6987 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6988 that creates RFC 2047 encoded "words" labels them as that charset instead
6989 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6990 expansion operator; (ii) when Exim creates a From: line for a local
6991 message; (iii) when a header line is rewritten to include a "phrase" part.
6993 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6994 buggy, causing it to skip the first lines of messages whose message ID
6995 ended in 'D'. This would not have bitten before Exim release 4.14, because
6996 message IDs were unlikely to end in 'D' before then. The effect was to have
6997 incorrect size information for certain domains.
6999 45. #include "config.h" was missing at the start of the crypt16.c module. This
7000 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7003 46. If there was a timeout during a "random" callout check, Exim treated it as
7004 a failure of the random address, and carried on sending RSET and the real
7005 address. If the delay was just some slowness somewhere, the response to the
7006 original RCPT would be taken as a response to RSET and so on, causing
7007 mayhem of various kinds.
7009 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7010 when I implemented it. It didn't allow for the fact that some option values
7011 may legitimately be negative (e.g. size_addition), and it didn't even do
7012 the right test for positive values.
7014 48. Domain names in DNS records are case-independent. Exim always looks them up
7015 in lower case. Some resolvers return domain names in exactly the case they
7016 appear in the zone file, that is, they may contain uppercase letters. Not
7017 all resolvers do this - some return always lower case. Exim was treating a
7018 change of case by a resolver as a change of domain, similar to a widening
7019 of a domain abbreviation. This triggered its re-routing code and so it was
7020 trying to route what was effectively the same domain again. This normally
7021 caused routing to fail (because the router wouldn't handle the domain
7022 twice). Now Exim checks for this case specially, and just changes the
7023 casing of the domain that it ultimately uses when it transmits the message
7026 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7029 50. If a filter generated a file delivery with a non-absolute name (possible if
7030 no home directory exists for the router), the forbid_file option was not
7033 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7034 the existing equality matching.
7036 52. Exim was using ints instead of ino_t variables in some places where it was
7037 dealing with inode numbers.
7039 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7040 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7041 called TMPDIR, and if it finds it is different, it changes its value.
7043 54. The smtp_printf() function is now made available to local_scan() so
7044 additional output lines can be written before returning. There is also an
7045 smtp_fflush() function to enable the detection of a dropped connection.
7046 The variables smtp_input and smtp_batched_input are exported to
7049 55. Changed the default runtime configuration: the message "Unknown user"
7050 has been removed from the ACL, and instead placed on the localuser router,
7051 using the cannot_route_message feature. This means that any verification
7052 failures that generate their own messages won't get overridden. Similarly,
7053 the "Unrouteable address" message that was in the ACL for unverifiable
7054 relay addresses has also been removed.
7056 56. Added hosts_avoid_esmtp to the smtp transport.
7058 57. The exicyclog script was not checking for the esoteric option
7059 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7060 will work only if exicyclog is run under the appropriate euid.
7062 58. Following a discussion on the list, the rules by which Exim recognises line
7063 endings on incoming messages have been changed. The -dropcr and drop_cr
7064 options are now no-ops, retained only for backwards compatibility. The
7065 following line terminators are recognized: LF CRLF CR. However, special
7066 processing applies to CR:
7068 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7069 nor a local message in the state where . is a terminator.
7071 (ii) If a bare CR is encountered in a header line, an extra space is added
7072 after the line terminator so as not to end the header. The reasoning
7073 behind this is that bare CRs in header lines are most likely either
7074 to be mistakes, or people trying to play silly games.
7076 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7077 was being incorrectly given as 18 bytes larger than it should have been.
7078 This is a VOB (very old bug).
7080 60. This may never have affected anything current, but just in case it has:
7081 When the local host is found other than at the start of a list of hosts,
7082 the local host, those with the same MX, and any that follow, are discarded.
7083 When the list in question was part of a longer list of hosts, the following
7084 hosts (not currently being processed) were also being discarded. This no
7085 longer happens. I'm not sure if this situation could ever has previously
7088 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7091 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7092 header. This is something that is recommended in a new Internet Draft, and
7093 is something that is documented as being done by Sendmail. There are two
7094 possible values. For messages generated by the autoreply transport, Exim
7097 Auto-Submitted: auto-replied
7099 whereas for all other generated messages (e.g. bounces) it adds
7101 Auto-Submitted: auto-generated
7103 63. The "personal" condition in filters now includes a test for the
7104 Auto-Submitted: header. If it contains the string "auto-" the message it
7105 not considered personal.
7107 64. Added rcpt_include_affixes as a generic transport option.
7109 65. Added queue_only_override (default true).
7111 66. Added the syslog_duplication option.
7113 67. If what should have been the first header line of a message consisted of
7114 a space followed by a colon, Exim was mis-interpreting it as a header line.
7115 It isn't of course - it is syntactically invalid and should therefore be
7116 treated as the start of the message body. The misbehaviour could have
7117 caused a number of strange effects, including loss of data in subsequent
7118 header lines, and spool format errors.
7120 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7121 client host had authenticated. This control can now be exercised by an ACL
7122 for more flexibility.
7124 69. By default, callouts do not happen when testing with -bh. There is now a
7125 variant, -bhc, which does actually run the callout code, including
7126 consulting and updating the callout cache.
7128 70. Added support for saslauthd authentication, courtesy of Alexander
7131 71. If statvfs() failed on the spool or log directories while checking their
7132 size for availability, Exim confusingly gave the error "space shortage".
7133 Furthermore, in debugging mode it crashed with a floating point exception.
7134 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7135 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7136 serious problem, Exim now writes to the main and panic logs when this
7137 happens, with details of the failure. It then refuses to accept the
7138 incoming message, giving the message "spool directory problem" or "log
7139 directory problem" with a 421 code for SMTP messages.
7141 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7142 0, 1, and 2 exist, because some OS complain for execs without them (see
7143 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7144 descriptors. However, the code omitted to check that the open succeeded,
7145 causing mysterious errors if for some reason the permissions on /dev/null
7146 got screwed. Now Exim writes a message to the main and panic logs, and
7147 bombs out if it can't open /dev/null.
7149 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7150 interact so that it is all more flexible. It is supposed to remain
7151 backwards compatible. Also added extra_local_interfaces.
7153 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7154 to bomb out with an assertion failure - to the client this appears as a
7155 connection drop. This problem occurs in the part of the code that was taken
7156 from the Samba project. Fortunately, the assertion is in a very simple
7157 function, so I have fixed this by reproducing the function inline in the
7158 one place where it is called, and arranging for authentication to fail
7159 instead of killing the process with assert().
7161 75. The SPA client code was not working when the server requested OEM rather
7162 than Unicode encoding.
7164 76. Added code to make require_files with a specific uid setting more usable in
7165 the case where statting the file as root fails - usually a non-root-mounted
7166 NFS file system. When this happens and the failure is EACCES, Exim now
7167 forks a subprocess and does the per-uid checking as the relevant uid.
7169 77. Added process_log_path.
7171 78. If log_file_path was not explicitly set, a setting of check_log_space or
7172 check_log_inodes was ignored.
7174 79. If a space check for the spool or log partitions fails, the incident is now
7175 logged. Of course, in the latter case the data may get lost...
7177 80. Added the %p formatting code to string_format() so that it can be used to
7178 print addresses in debug_print(). Adjusted all the address printing in the
7179 debugging in store.c to use %p rather than %d.
7181 81. There was a concern that a line of code in smtp_in.c could overflow a
7182 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7183 initially expressed, the concern was not well-founded, because trailing
7184 spaces are removed early. However, if the trailing spaces were followed by
7185 a NULL, they did not get removed, so the overflow was possible. Two fixes
7188 (a) I re-wrote the offending code in a cleaner fashion.
7189 (b) If an incoming SMTP command contains a NULL character, it is rejected
7192 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7193 runs initgroups(), so that if the Exim user is in any additional groups,
7194 they will be used during message reception.
7200 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob