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.
8 Since Exim version 4.90
11 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
12 Previously only the last row was returned.
14 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
15 we assumed that tags in the header were well-formed, and parsed the
16 element content after inspecting only the first char of the tag.
17 Assumptions at that stage could crash the receive process on malformed
20 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
21 While running the DKIM ACL we operate on the Permanent memory pool so that
22 variables created with "set" persist to the DATA ACL. Also (at any time)
23 DNS lookups that fail create cache records using the Permanent pool. But
24 expansions release any allocations made on the current pool - so a dnsdb
25 lookup expansion done in the DKIM ACL releases the memory used for the
26 DNS negative-cache, and bad things result. Solution is to switch to the
27 Main pool for expansions.
28 While we're in that code, add checks on the DNS cache during store_reset,
29 active in the testsuite.
30 Problem spotted, and debugging aided, by Wolfgang Breyha.
32 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
33 When none of the hosts presented to a transport match an already-open
34 connection, close it and proceed with the list. Previously we would
35 queue the message. Spotted by Lena with Yahoo, probably involving
38 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
39 Previously a spurious "250 OK id=" response was appended to the proper
42 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
43 rows affected is given instead).
49 JH/01 Rework error string handling in TLS interface so that the caller in
50 more cases is responsible for logging. This permits library-sourced
51 string to be attached to addresses during delivery, and collapses
52 pairs of long lines into single ones.
54 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
55 during configuration. Wildcards are allowed and expanded.
57 JH/02 Rework error string handling in DKIM to pass more info back to callers.
58 This permits better logging.
60 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
61 do not close it down and have the child transport start it up again on
62 the passed-on TCP connection. Instead, proxy the child (and any
63 subsequent ones) for TLS via a unix-domain socket channel. Logging is
64 affected: the continued delivery log lines do not have any DNSSEC, TLS
65 Certificate or OCSP information. TLS cipher information is still logged.
67 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
68 identical IP addresses on different listening ports. Will also affect
71 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
72 add noisy ifdef guards to special-case this sillyness.
73 Patch from Bernd Kuhls.
75 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
76 than 255 are no longer allowed.
78 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
79 Disable the session-cache too, which might reduce our load. Since we
80 currrectly use a new context for every connection, both as server and
81 client, there is no benefit for these.
82 GnuTLS appears to not support tickets server-side by default (we don't
83 call gnutls_session_ticket_enable_server()) but client side is enabled
84 by default on recent versions (3.1.3 +) unless the PFS priority string
87 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
88 <https://reproducible-builds.org/specs/source-date-epoch/>.
90 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
91 the check for any unsuccessful recipients did not notice the limit, and
92 erroneously found still-pending ones.
94 JH/08 Pipeline CHUNKING command and data together, on kernels that support
95 MSG_MORE. Only in-clear (not on TLS connections).
97 JH/09 Avoid using a temporary file during transport using dkim. Unless a
98 transport-filter is involved we can buffer the headers in memory for
99 creating the signature, and read the spool data file once for the
100 signature and again for transmission.
102 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
103 4.77 as the kernel support then wasn't solid, having issues in 64bit
104 mode. Now, it's been long enough. Add support for FreeBSD also.
106 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
107 case where the routing stage had gathered several addresses to send to
108 a host before calling the transport for the first, we previously failed
109 to close down TLS in the old transport process before passing the TCP
110 connection to the new process. The new one sent a STARTTLS command
111 which naturally failed, giving a failed delivery and bloating the retry
112 database. Investigation and fix prototype from Wolfgang Breyha.
114 JH/12 Fix check on SMTP command input synchronisation. Previously there were
115 false-negatives in the check that the sender had not preempted a response
116 or prompt from Exim (running as a server), due to that code's lack of
117 awareness of the SMTP input buffering.
119 PP/04 Add commandline_checks_require_admin option.
120 Exim drops privileges sanely, various checks such as -be aren't a
121 security problem, as long as you trust local users with access to their
122 own account. When invoked by services which pass untrusted data to
123 Exim, this might be an issue. Set this option in main configuration
124 AND make fixes to the calling application, such as using `--` to stop
127 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
128 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
129 responses to those, into a single TLS record each way (this usually means
130 a single packet). As a side issue, smtp_enforce_sync now works on TLS
133 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
134 affects you only if you're dancing at the edge of the param size limits.
135 If you are, and this message makes sense to you, then: raise the
136 configured limit or use OpenSSL 1.1. Nothing we can do for older
139 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
140 cmdline element to get the documented default one. Previously it was
143 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
144 are unsuable for non-message-specific reasons - eg. connection timeout,
145 banner-time rejection.
147 JH/16 Fix logging of delivery remote port, when specified by router, under
150 PP/06 Repair manualroute's ability to take options in any order, even if one
151 is the name of a transport.
154 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
156 JH/17 Change the list-building routines interface to use the expanding-string
157 triplet model, for better allocation and copying behaviour.
159 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
160 Previously it was constructed the first time a possibly-matching string
161 was met in the configuration file input during startup; now it is done
164 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
165 compatible one, to avoid the (poorly documented) possibility of a config
166 file in the working directory redirecting the DB files, possibly correpting
167 some existing file. CVE-2017-10140 assigned for BDB.
169 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
170 cache-hot. Previously, although the result was properly cached, the
171 initial verify call returned a defer.
173 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
174 the main verify for receipient in uncached-mode.
176 JH/22 Retire historical build files to an "unsupported" subdir. These are
177 defined as "ones for which we have no current evidence of testing".
179 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
180 if present. Previously it was ignored.
182 JH/24 Start using specified-initialisers in C structure init coding. This is
183 a C99 feature (it's 2017, so now considered safe).
185 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
186 if was a fixed-sized field and bitmask ops via macros; it is now more
189 PP/07 GitHub PR 56: Apply MariaDB build fix.
190 Patch provided by Jaroslav Škarvada.
192 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
193 during Coverity cleanups [4.87 JH/47]
194 Diagnosis and fix provided by Michael Fischer v. Mollard.
196 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
197 the right size to place the terminating semicolon on its own folded
198 line, the header hash was calculated to an incorrect value thanks to
199 the (relaxed) space the fold became.
201 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
202 and confused the parent.
204 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
205 which could crash as a result. This could lead to undeliverable messages.
207 JH/28 Logging: "next input sent too soon" now shows where input was truncated
210 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
211 matters on fast-turnover and PID-randomising systems, which were getting
212 out-of-order delivery.
214 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
215 a possibly-overlapping copy. The symptom was that "Remote host closed
216 connection in response to HELO" was logged instead of the actual 4xx
219 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
220 Previously only that bufferd was discarded, resulting in SYMTP command
223 JH/32 DKIM: when a message has multiple signatures matching an identity given
224 in dkim_verify_signers, run the dkim acl once for each. Previously only
225 one run was done. Bug 2189.
227 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
228 "panic the current process" to "deliberately defer". The panic log is
229 still written with the problem list name; the mail and reject logs now
230 get a temp-reject line for the message that was being handled, saying
231 something like "domains check lookup or other defer". The SMTP 451
232 message is still "Temporary local problem".
234 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
235 A crafted sequence of BDAT commands could result in in-use memory beeing
236 freed. CVE-2017-16943.
238 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
239 from SMTP input. Previously it was always done; now only done for DATA
240 and not BDAT commands. CVE-2017-16944.
242 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
243 to the message (such as an overlong header line). Previously this was
244 not done and we did not exit BDAT mode. Followon from the previous item
245 though a different problem.
251 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
252 than -2003 did; needs libidn2 in addition to libidn.
254 JH/02 The path option on a pipe transport is now expanded before use.
256 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
257 Patch provided by "Björn", documentation fix added too.
259 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
260 missing a wire-to-host endian conversion.
262 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
263 close after a BDAT command line could be taken as a following command,
264 giving a synch failure. Fix by only checking for synch immediately
265 before acknowledging the chunk.
267 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
268 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
269 macro. Patches provided by Josh Soref.
271 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
272 Previously we did not; the RFC seems ambiguous and VRFY is not listed
273 by IANA as a service extension. However, John Klensin suggests that we
276 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
277 the dkim code may be unix-mode line endings rather than smtp wire-format
278 CRLF, so prepend a CR to any bare LF.
280 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
281 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
283 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
284 the first were themselves being wrongly included in the feed into dkim
285 processing; with most chunk sizes in use this resulted in an incorrect
286 body hash calculated value.
288 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
289 DKIM signature block, for verification. Although advised against by
290 standards it is specifically not ruled illegal.
292 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
294 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
295 missing a body hash (the bh= tag).
297 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
298 It seems that HAProxy sends the Proxy Protocol information in clear and
299 only then does a TLS startup, so do the same.
301 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
302 TCP connections (such as for Spamd) unless the daemon successfully set
303 Fast Open mode on its listening sockets. This fixes breakage seen on
304 too-old kernels or those not configured for Fast Open, at the cost of
305 requiring both directions being enabled for TFO, and TFO never being used
306 by non-daemon-related Exim processes.
308 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
309 endings, at least on the first header line. Try to canonify any that get
310 past that check, despite the cost.
312 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
313 now limited to an arbitrary five deep, while parsing addresses with the
314 strip_excess_angle_brackets option enabled.
316 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
317 instead leave the unprompted TLS handshake in socket buffer for the
318 TLS library to consume.
320 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
322 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
324 JH/16 Drop variables when they go out of scope. Memory management drops a whole
325 region in one operation, for speed, and this leaves assigned pointers
326 dangling. Add checks run only under the testsuite which checks all
327 variables at a store-reset and panics on a dangling pointer; add code
328 explicitly nulling out all the variables discovered. Fixes one known
329 bug: a transport crash, where a dangling pointer for $sending_ip_address
330 originally assigned in a verify callout, is re-used.
332 PP/06 Drop '.' from @INC in various Perl scripts.
334 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
336 PP/08 Reduce a number of compilation warnings under clang; building with
337 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
338 should be warning-free.
340 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
342 HS/01 Fix portability problems introduced by PP/08 for platforms where
343 realloc(NULL) is not equivalent to malloc() [SunOS et al].
345 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
346 chunk. This allows us to accept broken chunked messages. We need a more
347 general solution here.
349 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
350 already-broken messages in the queue.
352 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
354 JH/19 Fix reference counting bug in routing-generated-address tracking.
360 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
361 supports it and a size is available (ie. the sending peer gave us one).
363 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
364 years of being deprecated). The replacements are the ACLs
365 acl_smtp_mime and acl_not_smtp_mime.
367 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
368 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
369 or even an in-clear connection were permitted. Now, if the host lookup
370 was dnssec and dane was requested then the host is only used if the
371 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
372 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
373 if one fails this test.
374 This means that a poorly-configured remote DNS will make it incommunicado;
375 but it protects against a DNS-interception attack on it.
377 JH/04 Bug 1810: make continued-use of an open smtp transport connection
378 non-noisy when a race steals the message being considered.
380 JH/05 If main configuration option tls_certificate is unset, generate a
381 self-signed certificate for inbound TLS connections.
383 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
384 in rewrites and routers.
386 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
387 and logged a warning sing 4.83; now they are a configuration file error.
389 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
390 (lacking @domain). Apply the same qualification processing as RCPT.
392 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
394 JH/10 Support ${sha256:} applied to a string (as well as the previous
397 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
398 a cutthrough deliver is pending, as we always want to make a connection.
399 This also avoids re-routing the message when later placing the cutthrough
400 connection after a verify cache hit.
401 Do not update it with the verify result either.
403 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
404 when routing results in more than one destination address.
406 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
407 signing (which inhibits the cutthrough capability). Previously only
408 the presence of an option was tested; now an expansion evaluating as
409 empty is permissible (obviously it should depend only on data available
410 when the cutthrough connection is made).
412 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
413 the relevant preceding SMTP command did not note the pipelining mode.
415 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
416 Previously they were not counted.
418 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
419 as one having no matching records. Previously we deferred the message
420 that needed the lookup.
422 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
423 distinguished as "(=".
425 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
426 for missing MX records. Previously it only worked for missing A records.
428 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
430 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
431 after the data-go-ahead and data-ack. Patch from Jason Betts.
433 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
434 even for a "none" policy. Patch from Tony Meyer.
436 JH/22 Fix continued use of a connection for further deliveries. If a port was
437 specified by a router, it must also match for the delivery to be
440 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
441 When one of the recipients of a message was unsuitable for the connection
442 (has no matching addresses), we lost track of needing to mark it
443 deferred. As a result mail would be lost.
445 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
447 JH/25 Decoding ACL controls is now done using a binary search; the source code
448 takes up less space and should be simpler to maintain. Merge the ACL
449 condition decode tables also, with similar effect.
451 JH/26 Fix problem with one_time used on a redirect router which returned the
452 parent address unchanged. A retry would see the parent address marked as
453 delivered, so not attempt the (identical) child. As a result mail would
456 JH/27 Fix a possible security hole, wherein a process operating with the Exim
457 UID can gain a root shell. Credit to http://www.halfdog.net/ for
458 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
461 JH/28 Enable {spool,log} filesystem space and inode checks as default.
462 Main config options check_{log,spool}_{inodes,space} are now
463 100 inodes, 10MB unless set otherwise in the configuration.
465 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
466 Previously it only applied to the main-section connection policy
469 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
471 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
472 by me. Added RFC7919 DH primes as an alternative.
474 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
475 are not in the system include path.
477 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
478 GnuTLS, when a session startup failed (eg because the client disconnected)
479 Exim did stdio operations after fclose. This was exposed by a recent
480 change which nulled out the file handle after the fclose.
482 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
483 signed directly by the cert-signing cert, rather than an intermediate
484 OCSP-signing cert. This is the model used by LetsEncrypt.
486 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
488 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
489 an incoming connection.
491 HS/02 Bug 1802: Do not half-close the connection after sending a request
494 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
495 fallback to "prime256v1".
497 JH/34 SECURITY: Use proper copy of DATA command in error message.
498 Could leak key material. Remotely exploitable. CVE-2016-9963.
504 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
505 and 3.4.4 - once the server is enabled to respond to an OCSP request
506 it does even when not requested, resulting in a stapling non-aware
507 client dropping the TLS connection.
509 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
510 support variable-length bit vectors. No functional change.
512 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
513 The I= interface field on outgoing lines is now after the H= remote
514 host field, same as incoming lines. There is a separate
515 outgoing_interface log selector which allows you to disable the
518 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
519 If not running log_selector +smtp_connection the mainlog would be held
520 open indefinitely after a "too many connections" event, including to a
521 deleted file after a log rotate. Leave the per net connection logging
522 leaving it open for efficiency as that will be quickly detected by the
523 check on the next write.
525 HS/01 Bug 1671: Fix post transport crash.
526 Processing the wait-<transport> messages could crash the delivery
527 process if the message IDs didn't exist for some reason. When
528 using 'split_spool_directory=yes' the construction of the spool
529 file name failed already, exposing the same netto behaviour.
531 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
532 mime_regex ACL conditions.
534 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
535 to DSN fail messages (bounces): remote IP, remote greeting, remote response
536 to HELO, local diagnostic string.
538 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
539 log line to debug. Even when configured with a tls authenticator many
540 client connections are expected to not authenticate in this way, so
541 an authenticate fail is not an error.
543 HS/02 Add the Exim version string to the process info. This way exiwhat
544 gives some more detail about the running daemon.
546 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
547 matter for fast-change records such as DNSBLs.
549 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
550 retry record, even if constant. There may be multiple transports with
551 different interface settings and the retry behaviour needs to be kept
554 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
556 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
558 JH/10 Bug 840: fix log_defer_output option of pipe transport
560 JH/11 Bug 830: use same host for all RCPTS of a message, even under
561 hosts_randomize. This matters a lot when combined with mua_wrapper.
563 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
564 ${quote_pgsql:<string>} operator.
566 JH/13 Bug 1708: avoid misaligned access in cached lookup.
568 JH/14 Change header file name for freeradius-client. Relevant if compiling
569 with Radius support; from the Gentoo tree and checked under Fedora.
571 JH/15 Bug 1712: Introduce $prdr_requested flag variable
573 JH/16 Bug 1714: Permit an empty string as expansion result for transport
574 option transport_filter, meaning no filtering.
576 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
578 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
579 defaults to "*" (all hosts). The variable is now available when not built
580 with TLS, default unset, mainly to enable keeping the testsuite sane.
581 If a server certificate is not supplied (via tls_certificate) an error is
582 logged, and clients will find TLS connections fail on startup. Presumably
583 they will retry in-clear.
584 Packagers of Exim are strongly encouraged to create a server certificate
585 at installation time.
587 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
588 with the $config_file variable.
590 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
591 in transport context, after the attempt, and per-recipient. The latter type
592 is per host attempted. The event data is the error message, and the errno
593 information encodes the lookup type (A vs. MX) used for the (first) host,
594 and the trailing two digits of the smtp 4xx response.
596 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
597 to write to mainlog (or rejectlog, paniclog) in the window between file
598 creation and permissions/ownership being changed. Particularly affects
599 installations where exicyclog is run as root, rather than exim user;
600 result is that the running daemon panics and dies.
602 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
604 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
605 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
606 "pri" and "weight". Note that the previous implicit priority given by the
607 list order is no longer honoured.
609 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
612 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
613 by defining SUPPORT_SOCKS.
615 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
616 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
617 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
618 variables are renamed to proxy_{local,external}_{address,port}.
620 JH/27 Move Internationalisation support from Experimental to mainline, enabled
621 for a build by defining SUPPORT_I18N
623 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
624 of the query string, and make ${quote_redis:} do that quoting.
626 JH/29 Move Events support from Experimental to mainline, enabled by default
627 and removable for a build by defining DISABLE_EVENT.
629 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
631 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
632 cached by the daemon.
634 JH/32 Move Redis support from Experimental to mainline, enabled for a build
635 by defining LOOKUP_REDIS. The libhiredis library is required.
637 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
638 keys are given for lookup.
640 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
641 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
642 only supported when built with TLS support. The PolarSSL SHA routines
643 are still used when the TLS library is too old for convenient support.
645 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
646 openssl_options), for security. OpenSSL forces this from version 1.1.0
647 server-side so match that on older versions.
649 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
650 allocation for $value could be released as the expansion processing
651 concluded, but leaving the global pointer active for it.
653 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
654 and to use the domains and local_parts ACL conditions.
656 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
657 incorrectly not doubled on cutthrough transmission, hence seen as a
658 body-termination at the receiving system - resulting in truncated mails.
659 Commonly the sender saw a TCP-level error, and retransmitted the message
660 via the normal store-and-forward channel. This could result in duplicates
661 received - but deduplicating mailstores were liable to retain only the
662 initial truncated version.
664 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
666 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
668 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
669 we're in there, support oversigning also; bug 1309.
671 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
673 HS/04 Add support for keep_environment and add_environment options.
675 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
676 either intentional arithmetic overflow during PRNG, or testing config-
679 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
680 delivery resulted in actual delivery. Cancel cutthrough before DATA
683 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
684 rejects a recipient: pass the reject to the originator.
686 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
687 Many were false-positives and ignorable, but it's worth fixing the
690 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
691 for the new environment-manipulation done at startup. Move the routines
692 from being local to tls.c to being global via the os.c file.
694 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
695 an extract embedded as result-arg for a map, the first arg for extract
696 is unavailable so we cannot tell if this is a numbered or keyed
697 extraction. Accept either.
703 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
706 JH/02 The smtp transport option "multi_domain" is now expanded.
708 JH/03 The smtp transport now requests PRDR by default, if the server offers
711 JH/04 Certificate name checking on server certificates, when exim is a client,
712 is now done by default. The transport option tls_verify_cert_hostnames
713 can be used to disable this per-host. The build option
714 EXPERIMENTAL_CERTNAMES is withdrawn.
716 JH/05 The value of the tls_verify_certificates smtp transport and main options
717 default to the word "system" to access the system default CA bundle.
718 For GnuTLS, only version 3.0.20 or later.
720 JH/06 Verification of the server certificate for a TLS connection is now tried
721 (but not required) by default. The verification status is now logged by
722 default, for both outbound TLS and client-certificate supplying inbound
725 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
728 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
729 Status Notification (bounce) messages are now MIME format per RFC 3464.
730 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
731 under the control of the dsn_advertise_hosts option, and routers may
732 have a dsn_lasthop option.
734 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
735 default, modifiable by a malware= option. The list separator for
736 the options can now be changed in the usual way. Bug 68.
738 JH/10 The smtp_receive_timeout main option is now expanded before use.
740 JH/11 The incoming_interface log option now also enables logging of the
741 local interface on delivery outgoing connections.
743 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
744 if the interface and destination host and port all match.
746 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
749 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
750 Patch from Andrew Lewis.
752 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
753 now supports optional time-restrictions, weighting, and priority
754 modifiers per server. Patch originally by <rommer@active.by>.
756 JH/16 The spamd_address main option now supports a mixed list of local
757 and remote servers. Remote servers can be IPv6 addresses, and
758 specify a port-range.
760 JH/17 Bug 68: The spamd_address main option now supports an optional
761 timeout value per server.
763 JH/18 Bug 1581: Router and transport options headers_add/remove can
764 now have the list separator specified.
766 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
769 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
772 JH/21 Support for the A6 type of dns record is withdrawn.
774 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
775 rather than the verbs used.
777 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
778 from 255 to 1024 chars.
780 JH/24 Verification callouts now attempt to use TLS by default.
782 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
783 are generic router options now. The defaults didn't change.
785 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
786 Original patch from Alexander Shikoff, worked over by JH.
788 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
789 files if ZCAT_COMMAND is not executable.
791 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
793 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
795 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
796 Normally benign, it bites when the pair was led to by a CNAME;
797 modern usage is to not canonicalize the domain to a CNAME target
798 (and we were inconsistent anyway for A-only vs AAAA+A).
800 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
802 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
803 when evaluating $sender_host_dnssec.
805 JH/31 Check the HELO verification lookup for DNSSEC, adding new
806 $sender_helo_dnssec variable.
808 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
810 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
812 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
814 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
815 documented as working, but never had. Support all but $spam_report.
817 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
818 added for tls authenticator.
820 HS/03 Add perl_taintmode main config option
826 TL/01 When running the test suite, the README says that variables such as
827 no_msglog_check are global and can be placed anywhere in a specific
828 test's script, however it was observed that placement needed to be near
829 the beginning for it to behave that way. Changed the runtest perl
830 script to read through the entire script once to detect and set these
831 variables, reset to the beginning of the script, and then run through
832 the script parsing/test process like normal.
834 TL/02 The BSD's have an arc4random API. One of the functions to induce
835 adding randomness was arc4random_stir(), but it has been removed in
836 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
837 function when detected.
839 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
840 cause callback expansion.
842 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
843 syntax errors in an expansion can be treated as a string instead of
844 logging or causing an error, due to the internal use of bool_lax
845 instead of bool when processing it.
847 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
848 server certificates when making smtp deliveries.
850 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
852 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
854 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
856 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
857 Merged patch from Sebastian Wiedenroth.
859 JH/05 Fix results-pipe from transport process. Several recipients, combined
860 with certificate use, exposed issues where response data items split
861 over buffer boundaries were not parsed properly. This eventually
862 resulted in duplicates being sent. This issue only became common enough
863 to notice due to the introduction of connection certificate information,
864 the item size being so much larger. Found and fixed by Wolfgang Breyha.
866 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
867 size buffer was used, resulting in syntax errors when an expansion
870 JH/07 Add support for directories of certificates when compiled with a GnuTLS
871 version 3.3.6 or later.
873 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
874 is EXPERIMENTAL_EVENT, the main-configuration and transport options
875 both become "event_action", the variables become $event_name, $event_data
876 and $event_defer_errno. There is a new variable $verify_mode, usable in
877 routers, transports and related events. The tls:cert event is now also
878 raised for inbound connections, if the main configuration event_action
881 TL/06 In test suite, disable OCSP for old versions of openssl which contained
882 early OCSP support, but no stapling (appears to be less than 1.0.0).
884 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
885 server certificate names available under the smtp transport option
886 "tls_verify_cert_hostname" now do not permit multi-component wildcard
889 JH/10 Time-related extraction expansions from certificates now use the main
890 option "timezone" setting for output formatting, and are consistent
891 between OpenSSL and GnuTLS compilations. Bug 1541.
893 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
894 encoded parameter in the incoming message. Bug 1558.
896 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
897 include certificate info, eximon was claiming there were spoolfile
900 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
902 JH/14 Log delivery-related information more consistently, using the sequence
903 "H=<name> [<ip>]" wherever possible.
905 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
906 are problematic for Debian distribution, omit them from the release
909 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
911 JH/16 Fix string representation of time values on 64bit time_t architectures.
914 JH/17 Fix a null-indirection in certextract expansions when a nondefault
915 output list separator was used.
920 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
921 checkers that were complaining about end of non-void function with no
924 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
925 This was a regression introduced in 4.83 by another bugfix.
927 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
929 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
930 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
936 TF/01 Correctly close the server side of TLS when forking for delivery.
938 When a message was received over SMTP with TLS, Exim failed to clear up
939 the incoming connection properly after forking off the child process to
940 deliver the message. In some situations the subsequent outgoing
941 delivery connection happened to have the same fd number as the incoming
942 connection previously had. Exim would try to use TLS and fail, logging
943 a "Bad file descriptor" error.
945 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
946 utilities have not been installed.
948 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
949 temporary space as the ACL may create new global variables.
951 TL/01 LDAP support uses per connection or global context settings, depending
952 upon the detected version of the libraries at build time.
954 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
955 to extract and use the src ip:port in logging and expansions as if it
956 were a direct connection from the outside internet. PPv2 support was
957 updated based on HAProxy spec change in May 2014.
959 JH/02 Add ${listextract {number}{list}{success}{fail}}.
961 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
962 Properly escape header and check for NULL return.
964 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
967 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
969 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
970 characters in header names, implemented as a verify condition.
971 Contributed by Michael Fischer v. Mollard.
973 TL/05 Rename SPF condition results err_perm and err_temp to standardized
974 results permerror and temperror. Previous values are deprecated but
975 still accepted. In a future release, err_perm and err_temp will be
976 completely removed, which will be a backward incompatibility if the
977 ACL tests for either of these two old results. Patch contributed by
978 user bes-internal on the mailing list.
980 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
982 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
983 selectors, in both main and reject logs.
985 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
988 JH/07 Add malware type "sock" for talking to simple daemon.
990 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
992 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
993 routers/transports under cutthrough routing.
995 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
996 numbers. Touch up "bool" conditional to keep the same definition.
998 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1000 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1002 JH/12 Expand items in router/transport headers_add or headers_remove lists
1003 individually rather than the list as a whole. Bug 1452.
1005 Required for reasonable handling of multiple headers_ options when
1006 they may be empty; requires that headers_remove items with embedded
1007 colons must have them doubled (or the list-separator changed).
1009 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1010 view the policy declared in the DMARC record. Currently, $dmarc_status
1011 is a combined value of both the record presence and the result of the
1014 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1016 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1017 dnslookup router and the smtp transport (applying to the forward
1020 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1021 of ldap servers used for a specific lookup. Patch provided by Heiko
1024 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1025 New variable $lookup_dnssec_authenticated for observability.
1027 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1028 Patch submitted by Lars Timman.
1030 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1032 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1033 Requires trusted mode and valid format message id, aborts otherwise.
1034 Patch contributed by Heiko Schlichting.
1036 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1037 certextract with support for various fields. Bug 1358.
1039 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1040 is requested by default, modifiable by smtp transport option
1043 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1044 operate on certificate variables to give certificate fingerprints
1045 Also new ${sha256:cert_variable}.
1047 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1049 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1050 Christian Aistleitner.
1052 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1054 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1055 file. Patch from Wolfgang Breyha.
1057 JH/25 Expand the coverage of the delivery $host and $host_address to
1058 client authenticators run in verify callout. Bug 1476.
1060 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1061 align with daemon_smtp_ports. Bug 72.
1063 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1064 support and error reporting did not work properly.
1066 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1067 and is readable. Patch from Andrew Colin Kissa.
1069 TL/14 Enhance documentation of ${run expansion and how it parses the
1070 commandline after expansion, particularly in the case when an
1071 unquoted variable expansion results in an empty value.
1073 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1075 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1076 by renaming away from it. Addresses GH issue 15, reported by
1079 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1081 TL/15 SECURITY: prevent double expansion in math comparison functions
1082 (can expand unsanitized data). Not remotely exploitable.
1089 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1091 PP/02 Make -n do something, by making it not do something.
1092 When combined with -bP, the name of an option is not output.
1094 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1097 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1098 $sender_host_name and config options to manage this, and basic check
1101 PP/05 DSCP support for outbound connections and control modifier for inbound.
1103 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1104 (Only plugin which currently uses this is kerberos4, which nobody should
1105 be using, but we should make it available and other future plugins might
1106 conceivably use it, even though it would break NAT; stuff *should* be
1107 using channel bindings instead).
1109 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1110 name; added for Sendmail compatibility; requires admin caller.
1111 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1112 just ignore it); requires trusted caller.
1113 Also parse but ignore: -Ac -Am -X<logfile>
1116 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1118 TL/02 Add +smtp_confirmation as a default logging option.
1120 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1121 Patch by Magnus Holmgren from 2007-02-20.
1123 TL/04 Bugzilla 1281 - Spec typo.
1124 Bugzilla 1283 - Spec typo.
1125 Bugzilla 1290 - Spec grammar fixes.
1127 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1129 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1131 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1132 mailing list by by Dmitry Isaikin.
1134 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1136 JH/02 Support "G" suffix to numbers in ${if comparisons.
1138 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1140 NM/01 Bugzilla 1197 - Spec typo
1141 Bugzilla 1196 - Spec examples corrections
1143 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1145 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1146 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1149 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1150 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1152 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1153 "acl {{name}{arg}...}", and optional args on acl condition
1156 JH/05 Permit multiple router/transport headers_add/remove lines.
1158 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1160 JH/07 Avoid using a waiting database for a single-message-only transport.
1161 Performance patch from Paul Fisher. Bugzilla 1262.
1163 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1166 JH/09 Add $headers_added variable, with content from use of ACL modifier
1167 add_header (but not yet added to the message). Bugzilla 199.
1169 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1170 Pulled from Bugzilla 817 by Wolfgang Breyha.
1172 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1174 (nb: this is the same fix as in Exim 4.80.1)
1176 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1179 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1182 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1184 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1185 advertise SMTP AUTH mechanism to us, instead of a generic
1186 protocol violation error. Also, make Exim more robust to bad
1187 data from the Dovecot auth socket.
1189 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1191 When a queue runner is handling a message, Exim first routes the
1192 recipient addresses, during which it prunes them based on the retry
1193 hints database. After that it attempts to deliver the message to
1194 any remaining recipients. It then updates the hints database using
1197 So if a recipient address works intermittently, it can get repeatedly
1198 deferred at routing time. The retry hints record remains fresh so the
1199 address never reaches the final cutoff time.
1201 This is a fairly common occurrence when a user is bumping up against
1202 their storage quota. Exim had some logic in its local delivery code
1203 to deal with this. However it did not apply to per-recipient defers
1204 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1206 This change adds a proper retry rule check during routing so that the
1207 final cutoff time is checked against the message's age. We only do
1208 this check if there is an address retry record and there is not a
1209 domain retry record; this implies that previous attempts to handle
1210 the address had the retry_use_local_parts option turned on. We use
1211 this as an approximation for the destination being like a local
1212 delivery, as in LMTP.
1214 I suspect this new check makes the old local delivery cutoff check
1215 redundant, but I have not verified this so I left the code in place.
1217 TF/02 Correct gecos expansion when From: is a prefix of the username.
1219 Test 0254 submits a message to Exim with the header
1223 When I ran the test suite under the user fanf2, Exim expanded
1224 the header to contain my full name, whereas it should have added
1225 a Resent-Sender: header. It erroneously treats any prefix of the
1226 username as equal to the username.
1228 This change corrects that bug.
1230 GF/01 DCC debug and logging tidyup
1231 Error conditions log to paniclog rather than rejectlog.
1232 Debug lines prefixed by "DCC: " to remove any ambiguity.
1234 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1236 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1237 Bug spotted by Jeremy Harris; was flawed since initial commit.
1238 Would have resulted in OCSP responses post-SNI triggering an Exim
1239 NULL dereference and crash.
1241 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1243 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1244 Bug detection, analysis and fix by Samuel Thibault.
1245 Bugzilla 1331, Debian bug #698092.
1247 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1249 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1250 Server implementation by Todd Lyons, client by JH.
1251 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1252 config variable "prdr_enable" controls whether the server
1253 advertises the facility. If the client requests PRDR a new
1254 acl_data_smtp_prdr ACL is called once for each recipient, after
1255 the body content is received and before the acl_smtp_data ACL.
1256 The client is controlled by both of: a hosts_try_prdr option
1257 on the smtp transport, and the server advertisement.
1258 Default client logging of deliveries and rejections involving
1259 PRDR are flagged with the string "PRDR".
1261 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1262 fclose(). Diagnosis by Todd Lyons.
1264 PP/17 Update configure.default to handle IPv6 localhost better.
1265 Patch by Alain Williams (plus minor tweaks).
1268 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1269 This is now consistent with GnuTLS, and is now documented: the
1270 previous undocumented portable approach to treating the option as
1271 unset was to force an expansion failure. That still works, and
1272 an empty string is now equivalent.
1274 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1275 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1276 not performing validation itself.
1278 PP/20 Added force_command boolean option to pipe transport.
1279 Patch from Nick Koston, of cPanel Inc.
1281 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1284 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1286 PP/21 Fix eximon continuous updating with timestamped log-files.
1287 Broken in a format-string cleanup in 4.80, missed when I repaired the
1288 other false fix of the same issue.
1289 Report and fix from Heiko Schlichting.
1292 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1293 Report from Prashanth Katuri.
1295 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1296 It's SecureTransport, so affects any MacOS clients which use the
1297 system-integrated TLS libraries, including email clients.
1299 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1300 using a MIME ACL for non-SMTP local injection.
1301 Report and assistance in diagnosis by Warren Baker.
1303 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1305 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1307 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1308 last id that failed so it may be referenced in subsequent ACL's.
1310 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1313 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1314 ldap library initialization, allowing self-signed CA's to be
1315 used. Also properly sets require_cert option later in code by
1316 using NULL (global ldap config) instead of ldap handle (per
1317 session). Bug diagnosis and testing by alxgomz.
1319 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1320 the src/util/ subdirectory.
1322 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1323 renamed to Transport Post Delivery Action by Jeremy Harris, as
1326 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1327 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1328 redis_servers = needs to be configured which will be used by the redis
1329 lookup. Patch from Warren Baker, of The Packet Hub.
1331 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1333 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1334 hostname or reverse DNS when processing a host list. Used suggestions
1335 from multiple comments on this bug.
1337 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1339 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1340 Missed a few lines, added it to make the runtest require no keyboard
1343 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1344 contains upper case chars. Make router use caseful_local_part.
1346 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1347 support when GnuTLS has been built with p11-kit.
1353 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1355 This, or similar/improved, will also be change PP/11 of 4.82.
1361 PP/01 Handle short writes when writing local log-files.
1362 In practice, only affects FreeBSD (8 onwards).
1363 Bugzilla 1053, with thanks to Dmitry Isaikin.
1365 NM/01 Bugzilla 949 - Documentation tweak
1367 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1370 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1372 PP/02 Implemented gsasl authenticator.
1374 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1376 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1377 `pkg-config foo` for cflags/libs.
1379 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1380 with rest of GSASL and with heimdal_gssapi.
1382 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1383 `pkg-config foo` for cflags/libs for the TLS implementation.
1385 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1386 properties get this fed in as external SSF. A number of robustness
1387 and debugging improvements to the cyrus_sasl authenticator.
1389 PP/08 cyrus_sasl server now expands the server_realm option.
1391 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1392 Patch by Jeremy Harris.
1394 PP/10 Added dbmjz lookup type.
1396 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1398 PP/12 MAIL args handles TAB as well as SP, for better interop with
1399 non-compliant senders.
1400 Analysis and variant patch by Todd Lyons.
1402 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1403 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1404 Patch from Dirk Mueller <dmueller@suse.com>
1406 PP/13 tls_peerdn now print-escaped for spool files.
1407 Observed some $tls_peerdn in wild which contained \n, which resulted
1408 in spool file corruption.
1410 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1411 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1412 or write after TLS renegotiation, which otherwise led to messages
1415 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1416 as a tracking header (ie: a signed header comes before the signature).
1417 Patch from Wolfgang Breyha.
1419 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1420 comma-sep list; embedded commas doubled.
1422 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1424 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1426 Report and patch from Dmitry Banschikov.
1428 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1429 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1430 failures. We appear to now support TLS1.1+ with Exim.
1432 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1433 lets Exim select keys and certificates based upon TLS SNI from client.
1434 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1435 before an outbound SMTP session. New log_selector, +tls_sni.
1437 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1438 NULL dereference. Report and patch from Alun Jones.
1440 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1441 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1442 Not seeing resolver debug output on NetBSD, but suspect this is a
1443 resolver implementation change.
1445 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1446 Left warnings. Added "eximon gdb" invocation mode.
1448 PP/21 Defaulting "accept_8bitmime" to true, not false.
1450 PP/22 Added -bw for inetd wait mode support.
1452 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1453 locate the relevant includes and libraries. Made this the default.
1455 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1456 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1458 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1459 This may cause build issues on older platforms.
1461 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1462 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1463 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1464 Added SNI support via GnuTLS too.
1465 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1467 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1469 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1470 Applied second patch from Janne, implementing suggestion to default
1471 multiple-strings-in-record handling to match SPF spec.
1473 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1475 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1476 read-only, out of scope).
1477 Patch from Wolfgang Breyha, report from Stuart Northfield.
1479 PP/29 Fix three issues highlighted by clang analyser static analysis.
1480 Only crash-plausible issue would require the Cambridge-specific
1481 iplookup router and a misconfiguration.
1482 Report from Marcin Mirosław.
1484 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1486 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1487 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1488 As part of this, removing so much warning spew let me fix some minor
1489 real issues in debug logging.
1491 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1492 assignment on my part. Fixed.
1494 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1495 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1496 Janne Snabb (who went above and beyond: thank you).
1498 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1499 string otherwise requires a connection and a bunch more work and it's
1500 relatively easy to get wrong. Should also expose TLS library linkage
1503 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1504 64-bit ${eval} (JH/03).
1506 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1507 GNU libc to support some of the 64-bit stuff, should not lead to
1508 conflicts. Defined before os.h is pulled in, so if a given platform
1509 needs to override this, it can.
1511 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1512 protection layer was required, which is not implemented.
1513 Bugzilla 1254, patch from Wolfgang Breyha.
1515 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1516 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1517 tls_dhparam take prime identifiers. Also unbreak combination of
1518 OpenSSL+DH_params+TLSSNI.
1520 PP/39 Disable SSLv2 by default in OpenSSL support.
1526 PP/01 Solaris build fix for Oracle's LDAP libraries.
1527 Bugzilla 1109, patch from Stephen Usher.
1529 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1531 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1534 TF/02 Fix a couple more cases where we did not log the error message
1535 when unlink() failed. See also change 4.74-TF/03.
1537 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1538 lock up or crash if it happened to be inside a call to libc when it
1539 got a SIGUSR1 from exiwhat.
1541 The SIGUSR1 handler appends the current process status to the process
1542 log which is later printed by exiwhat. It used to use the general
1543 purpose logging code to do this, but several functions it calls are
1544 not safe for signals.
1546 The new output code in the SIGUSR1 handler is specific to the process
1547 log, and simple enough that it's easy to inspect for signal safety.
1548 Removing some special cases also simplifies the general logging code.
1549 Removing the spurious timestamps from the process log simplifies
1552 TF/04 Improved ratelimit ACL condition.
1554 The /noupdate option has been deprecated in favour of /readonly which
1555 has clearer semantics. The /leaky, /strict, and /readonly update modes
1556 are mutually exclusive. The update mode is no longer included in the
1557 database key; it just determines when the database is updated. (This
1558 means that when you upgrade Exim will forget old rate measurements.)
1560 Exim now checks that the per_* options are used with an update mode that
1561 makes sense for the current ACL. For example, when Exim is processing a
1562 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1563 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1564 must specify per_mail/readonly. If you omit the update mode it defaults to
1565 /leaky where that makes sense (as before) or /readonly where required.
1567 The /noupdate option is now undocumented but still supported for
1568 backwards compatibility. It is equivalent to /readonly except that in
1569 ACLs where /readonly is required you may specify /leaky/noupdate or
1570 /strict/noupdate which are treated the same as /readonly.
1572 A useful new feature is the /count= option. This is a generalization
1573 of the per_byte option, so that you can measure the throughput of other
1574 aggregate values. For example, the per_byte option is now equivalent
1575 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1577 The per_rcpt option has been generalized using the /count= mechanism
1578 (though it's more complicated than the per_byte equivalence). When it is
1579 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1580 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1581 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1582 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1583 Note that using per_rcpt with a non-readonly update mode in more than
1584 one ACL will cause the recipients to be double-counted. (The per_mail
1585 and per_byte options don't have this problem.)
1587 The handling of very low rates has changed slightly. If the computed rate
1588 is less than the event's count (usually one) then this event is the first
1589 after a long gap. In this case the rate is set to the same as this event's
1590 count, so that the first message of a spam run is counted properly.
1592 The major new feature is a mechanism for counting the rate of unique
1593 events. The new per_addr option counts the number of different
1594 recipients that someone has sent messages to in the last time period. It
1595 behaves like per_rcpt if all the recipient addresses are different, but
1596 duplicate recipient addresses do not increase the measured rate. Like
1597 the /count= option this is a general mechanism, so the per_addr option
1598 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1599 example, measure the rate that a client uses different sender addresses
1600 with the options per_mail/unique=$sender_address. There are further
1601 details in the main documentation.
1603 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1605 TF/06 Removed a few PCRE remnants.
1607 TF/07 Automatically extract Exim's version number from tags in the git
1608 repository when doing development or release builds.
1610 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1611 Bugzilla 879. Patch from Paul Fisher.
1613 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1614 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1617 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1619 PP/05 Variable $av_failed, true if the AV scanner deferred.
1620 Bugzilla 1078. Patch from John Horne.
1622 PP/06 Stop make process more reliably on build failure.
1623 Bugzilla 1087. Patch from Heiko Schlittermann.
1625 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1626 Bugzilla 1089. Patch from Heiko Schlittermann.
1628 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1629 Bugzilla 1131. Patch from Holger Weiß.
1631 PP/09 Handle IPv6 addresses with SPF.
1632 Bugzilla 860. Patch from Wolfgang Breyha.
1634 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1636 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1639 PP/11 match_* no longer expand right-hand-side by default.
1640 New compile-time build option, EXPAND_LISTMATCH_RHS.
1641 New expansion conditions, "inlist", "inlisti".
1643 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1645 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1647 PP/14 fix log_write() format string regression from TF/03.
1648 Bugzilla 1152. Patch from Dmitry Isaikin.
1654 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1656 PP/02 Harmonised TLS library version reporting; only show if debugging.
1657 Layout now matches that introduced for other libraries in 4.74 PP/03.
1659 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1661 PP/04 New "dns_use_edns0" global option.
1663 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1666 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1667 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1669 TK/01 Updated PolarSSL code to 0.14.2.
1670 Bugzilla 1097. Patch from Andreas Metzler.
1672 PP/07 Catch divide-by-zero in ${eval:...}.
1673 Fixes bugzilla 1102.
1675 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1678 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1679 format-string attack -- SECURITY: remote arbitrary code execution.
1681 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1682 time unintentionally subject to list matching rules, letting the header
1683 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1684 arbitrary string expansion). This allowed for information disclosure.
1686 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1687 INT_MIN/-1 -- value coerced to INT_MAX.
1693 NM/01 Workaround for PCRE version dependency in version reporting
1696 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1697 This fixes portability to compilers other than gcc, notably
1698 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1700 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1701 makefiles for portability to HP-UX and POSIX correctness.
1703 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1704 Also via indented variable definition in the Makefile.
1705 (Debugging by Oliver Heesakkers).
1707 PP/02 Restore caching of spamd results with expanded spamd_address.
1708 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1710 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1711 Improves build reliability. Fix from: Frank Elsner
1713 NM/02 Fix wide character breakage in the rfc2047 coding
1714 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1716 NM/03 Allow underscore in dnslist lookups
1717 Fixes bug 1026. Patch from Graeme Fowler
1719 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1720 Code patches from Adam Ciarcinski of NetBSD.
1722 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1725 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1726 is logged, to avoid truncation. Patch from John Horne.
1728 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1729 Patch from Jakob Hirsch.
1731 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1732 SQL string expansion failure details.
1733 Patch from Andrey Oktyabrski.
1735 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1736 Patch from Simon Arlott.
1738 PP/09 New lookups functionality failed to compile on old gcc which rejects
1739 extern declarations in function scope.
1740 Patch from Oliver Fleischmann
1742 PP/10 Use sig_atomic_t for flags set from signal handlers.
1743 Check getgroups() return and improve debugging.
1744 Fixed developed for diagnosis in bug 927 (which turned out to be
1747 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1748 Patch from Mark Zealey.
1750 PP/12 Bugzilla 1056: Improved spamd server selection.
1751 Patch from Mark Zealey.
1753 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1754 Based on patch from Heiko Schlittermann.
1756 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1757 Patch from Uwe Doering, sign-off by Michael Haardt.
1759 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1760 variable declaration deep within a block. Bug and patch from
1763 PP/15 lookups-Makefile IRIX compatibility coercion.
1765 PP/16 Make DISABLE_DKIM build knob functional.
1767 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1768 Patch by Simon Arlott
1770 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1771 variable argument macros. Our copy now differs from upstream.
1777 TF/01 Failure to get a lock on a hints database can have serious
1778 consequences so log it to the panic log.
1780 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1781 controlled using the smtp_confirmation log selector.
1783 TF/03 Include the error message when we fail to unlink a spool file.
1785 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1786 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1787 for maintaining out-of-tree patches for some time.
1789 PP/01 Bugzilla 139: Documentation and portability issues.
1790 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1791 Handle per-OS dynamic-module compilation flags.
1793 PP/02 Let /dev/null have normal permissions.
1794 The 4.73 fixes were a little too stringent and complained about the
1795 permissions on /dev/null. Exempt it from some checks.
1796 Reported by Andreas M. Kirchwitz.
1798 PP/03 Report version information for many libraries, including
1799 Exim version information for dynamically loaded libraries. Created
1800 version.h, now support a version extension string for distributors
1801 who patch heavily. Dynamic module ABI change.
1803 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1804 privilege escalation vulnerability whereby the Exim run-time user
1805 can cause root to append content of the attacker's choosing to
1808 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1811 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1812 If dropping privileges for untrusted macros, we disabled normal logging
1813 on the basis that it would fail; for the Exim run-time user, this is not
1814 the case, and it resulted in successful deliveries going unlogged.
1815 Fixed. Reported by Andreas Metzler.
1821 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1822 only prepend for the Resent-* case. Fixes regression introduced in
1823 Exim 4.70 by NM/22 for Bugzilla 607.
1825 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1826 increasing numbers of administrators be bitten by this.
1828 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1830 PP/03 Bugzilla 994: added openssl_options main configuration option.
1832 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1834 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1836 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1838 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1839 without a peer certificate, leading to a segfault because of an
1840 assumption that peers always have certificates. Be a little more
1841 paranoid. Problem reported by Martin Tscholak.
1843 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1844 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1845 NB: ClamAV planning to remove STREAM in "middle of 2010".
1846 CL also introduces -bmalware, various -d+acl logging additions and
1847 more caution in buffer sizes.
1849 PP/09 Implemented reverse_ip expansion operator.
1851 PP/10 Bugzilla 937: provide a "debug" ACL control.
1853 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1855 PP/12 Bugzilla 973: Implement --version.
1857 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1859 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1861 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1863 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1864 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1865 ignore trailing whitespace.
1867 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1869 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1872 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1873 Notification from Dr Andrew Aitchison.
1875 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1876 ExtendedDetectionInfo response format.
1877 Notification from John Horne.
1879 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1882 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1883 XSL and documented dependency on system catalogs, with examples of how
1886 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1889 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1890 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1891 configuration file which is writeable by the Exim user or group.
1893 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1894 of configuration files to cover files specified with the -C option if
1895 they are going to be used with root privileges, not just the default
1898 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1899 option (effectively making it always true).
1901 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1902 files to be used while preserving root privileges.
1904 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1905 that rogue child processes cannot use them.
1907 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1908 run-time user, instead of root.
1910 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1911 Exim run-time user without dropping privileges.
1913 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1914 result string, instead of calling string_vformat() twice with the same
1917 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1918 for other users. Others should always drop root privileges if they use
1919 -C on the command line, even for a whitelisted configure file.
1921 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1923 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1929 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1930 $data_path, and $header_path variables; fixed documentation bugs and
1933 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1934 exipick to access non-standard spools, including the "frozen" queue
1937 NM/01 Bugzilla 965: Support mysql stored procedures.
1938 Patch from Alain Williams
1940 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1942 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1943 Patch from Andreas Metzler
1945 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1946 Patch from Kirill Miazine
1948 NM/05 Bugzilla 671: Added umask to procmail example.
1950 JJ/03 installed exipick 20100323.0, fixing doc bug
1952 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1953 directory. Notification and patch from Dan Rosenberg.
1955 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1957 TK/02 Improve log output when DKIM signing operation fails.
1959 MH/01 Treat the transport option dkim_domain as a colon separated
1960 list, not as a single string, and sign the message with each element,
1961 omitting multiple occurences of the same signer.
1963 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1964 Bugzilla 985, 986. Patch by Simon Arlott
1966 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1967 Patch by Simon Arlott
1969 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1970 MBX locking. Notification from Dan Rosenberg.
1976 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1978 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1980 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1982 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1984 NM/04 Bugzilla 915: Flag broken perl installation during build.
1990 TK/01 Added patch by Johannes Berg that expands the main option
1991 "spamd_address" if it starts with a dollar sign.
1993 TK/02 Write list of recipients to X-Envelope-Sender header when building
1994 the mbox-format spool file for content scanning (suggested by Jakob
1997 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1998 (http://www.dcc-servers.net/) support via dccifd. Activated by
1999 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2001 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2002 by Mark Daniel Reidel <mr@df.eu>.
2004 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2005 When building exim an external PCRE library is now needed -
2006 PCRE is a system library on the majority of modern systems.
2007 See entry on PCRE_LIBS in EDITME file.
2009 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2010 conversation. Added nologin parameter to request.
2011 Patch contributed by Kirill Miazine.
2013 TF/01 Do not log submission mode rewrites if they do not change the address.
2015 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2017 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2018 log files in place. Contributed by Roberto Lima.
2020 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2022 TF/03 Bugzilla 615: When checking the local_parts router precondition
2023 after a local_part_suffix or local_part_prefix option, Exim now
2024 does not use the address's named list lookup cache, since this
2025 contains cached lookups for the whole local part.
2027 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2028 Robert Millan. Documentation is in experimental-spec.txt.
2030 TF/04 Bugzilla 668: Fix parallel build (make -j).
2032 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2034 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2035 Patch provided by Jan Srzednicki.
2037 TF/05 Leading white space used to be stripped from $spam_report which
2038 wrecked the formatting. Now it is preserved.
2040 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2041 that they are available at delivery time.
2043 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2045 TF/08 TLS error reporting now respects the incoming_interface and
2046 incoming_port log selectors.
2048 TF/09 Produce a more useful error message if an SMTP transport's hosts
2049 setting expands to an empty string.
2051 NM/06 Bugzilla 744: EXPN did not work under TLS.
2052 Patch provided by Phil Pennock.
2054 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2055 Patch provided by Richard Godbee.
2057 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2058 acl_smtp_notquit, added index entry.
2060 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2061 Patch provided by Eugene Bujak.
2063 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2064 accept(). Patch provided by Maxim Dounin.
2066 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2067 Patch provided by Phil Pennock.
2069 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2071 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2072 Patch provided by Brad "anomie" Jorsch.
2074 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2075 Patch provided by Dean Brooks.
2077 TK/05 Add native DKIM support (does not depend on external libraries).
2079 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2080 Patch provided by Graeme Fowler.
2082 NM/16 Bugzilla 851: Documentation example syntax fix.
2084 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2086 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2089 NM/19 Bugzilla 745: TLS version reporting.
2090 Patch provided by Phil Pennock.
2092 NM/20 Bugzilla 167: bool: condition support.
2093 Patch provided by Phil Pennock.
2095 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2096 clients. Patch provided by Phil Pennock.
2098 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2099 Patch provided by Brad "anomie" Jorsch.
2101 NM/23 Bugzilla 687: Fix misparses in eximstats.
2102 Patch provided by Heiko Schlittermann.
2104 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2105 Patch provided by Heiko Schlittermann.
2107 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2108 plus update to original patch.
2110 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2112 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2113 Patch provided by David Brownlee.
2115 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2117 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2119 NM/30 Bugzilla 888: TLS documentation bugfixes.
2121 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2123 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2124 Unlike the original bugzilla I have changed all shell scripts in src tree.
2126 NM/33 Bugzilla 898: Transport filter timeout fix.
2127 Patch by Todd Rinaldo.
2129 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2130 Patch by Serge Demonchaux.
2132 NM/35 Bugzilla 39: Base64 decode bug fixes.
2133 Patch by Jakob Hirsch.
2135 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2137 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2139 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2141 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2147 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2148 ALT-N's libdkim that I have put here:
2149 http://duncanthrax.net/exim-experimental/
2151 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2152 were called 'true' and it seems that C99 defines that as a reserved
2153 keyword to be used with 'bool' variable types. That means you could
2154 not include C99-style headers which use bools without triggering
2155 build errors in sieve.c.
2157 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2158 as mailq or other aliases. Changed the --help handling significantly
2159 to do whats expected. exim_usage() emits usage/help information.
2161 SC/01 Added the -bylocaldomain option to eximstats.
2163 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2165 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2167 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2173 PH/01 Another patch from the Sieve maintainer.
2175 PH/02 When an IPv6 address is converted to a string for single-key lookup
2176 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2177 dots are used instead of colons so that keys in lsearch files need not
2178 contain colons. This was done some time before quoting was made available
2179 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2180 (notated using the quote facility) so as to distinguish them from IPv4
2181 keys. This meant that lookups for IP addresses in host lists did not work
2182 for iplsearch lookups.
2184 This has been fixed by arranging for IPv6 addresses to be expressed with
2185 colons if the lookup type is iplsearch. This is not incompatible, because
2186 previously such lookups could never work.
2188 The situation is now rather anomalous, since one *can* have colons in
2189 ordinary lsearch keys. However, making the change in all cases is
2190 incompatible and would probably break a number of configurations.
2192 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2195 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2196 conversion specification without a maximum field width, thereby enabling
2197 a rogue spamd server to cause a buffer overflow. While nobody in their
2198 right mind would setup Exim to query an untrusted spamd server, an
2199 attacker that gains access to a server running spamd could potentially
2200 exploit this vulnerability to run arbitrary code as the Exim user.
2202 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2203 $primary_hostname instead of what libspf2 thinks the hosts name is.
2205 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2206 a directory entry by the name of the lookup key. Previously, if a
2207 symlink pointed to a non-existing file or a file in a directory that
2208 Exim lacked permissions to read, a lookup for a key matching that
2209 symlink would fail. Now it is enough that a matching directory entry
2210 exists, symlink or not. (Bugzilla 503.)
2212 PH/03 The body_linecount and body_zerocount variables are now exported in the
2215 PH/04 Added the $dnslist_matched variable.
2217 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2218 This means they are set thereafter only if the connection becomes
2221 PH/06 Added the client_condition to authenticators so that some can be skipped
2222 by clients under certain conditions.
2224 PH/07 The error message for a badly-placed control=no_multiline_responses left
2225 "_responses" off the end of the name.
2227 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2229 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2230 (without spaces) instead of just copying the configuration text.
2232 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2234 PH/11 Added $max_received_linelength.
2236 PH/12 Added +ignore_defer and +include_defer to host lists.
2238 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2239 way in which PCRE > 7.0 is built.
2241 PH/14 Implemented queue_only_load_latch.
2243 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2244 MAIL command. The effect was to mangle the value on 64-bit systems.
2246 PH/16 Another patch from the Sieve maintainer.
2248 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2250 PH/18 If a system quota error occurred while trying to create the file for
2251 a maildir delivery, the message "Mailbox is full" was not appended to the
2252 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2253 only to a quota excession during the actual writing of the file.
2255 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2256 characters?) which causes problems in log lines. The DN values are now
2257 passed through string_printing() before being added to log lines.
2259 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2260 and InterBase are left for another time.)
2262 PH/21 Added message_body_newlines option.
2264 PH/22 Guard against possible overflow in moan_check_errorcopy().
2266 PH/23 POSIX allows open() to be a macro; guard against that.
2268 PH/24 If the recipient of an error message contained an @ in the local part
2269 (suitably quoted, of course), incorrect values were put in $domain and
2270 $local_part during the evaluation of errors_copy.
2276 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2277 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2280 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2281 issue a MAIL command.
2283 PH/02 In an ACL statement such as
2285 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2287 if a client was not listed at all, or was listed with a value other than
2288 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2289 the condition was not true (as it should be), so access was not denied.
2290 The bug was that the ! inversion was incorrectly passed on to the second
2291 item. This has been fixed.
2293 PH/03 Added additional dnslists conditions == and =& which are different from
2294 = and & when the dns lookup returns more than one IP address.
2296 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2297 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2299 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2300 FSYNC, which compiles an option called disable_fsync that allows for
2301 bypassing fsync(). The documentation is heavily laced with warnings.
2303 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2305 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2306 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2307 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2308 including adding "make clean"; (3) Added -fPIC when compiling the test
2309 dynamically loaded module, to get rid of a warning.
2311 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2312 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2313 = 0s. The bug is otherwise harmless.
2315 PH/07 There was a bug in the dovecot authenticator such that the value of
2316 $auth1 could be overwritten, and so not correctly preserved, after a
2317 successful authentication. This usually meant that the value preserved by
2318 the server_setid option was incorrect.
2320 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2322 PH/09 Installed PCRE release 7.0.
2324 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2325 run for batched SMTP input. It is now run at the start of every message
2326 in the batch. While fixing this I discovered that the process information
2327 (output by running exiwhat) was not always getting set for -bs and -bS
2328 input. This is fixed, and it now also says "batched" for BSMTP.
2330 PH/11 Added control=no_pipelining.
2332 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2333 patch, slightly modified), and move the expansion of helo_data till after
2334 the connection is made in the smtp transport (so it can use these
2337 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2339 PH/14 Added log_selector = +pid.
2341 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2343 PH/16 Add ${if forany and ${if forall.
2345 PH/17 Added dsn_from option to vary the From: line in DSNs.
2347 PH/18 Flush SMTP output before performing a callout, unless control =
2348 no_callout_flush is set.
2350 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2351 was true (the default) a successful delivery failed to delete the retry
2352 item, thus causing premature timeout of the address. The bug is now
2355 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2357 PH/21 Long custom messages for fakedefer and fakereject are now split up
2358 into multiline responses in the same way that messages for "deny" and
2359 other ACL rejections are.
2361 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2362 with slight modification.
2364 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2365 draft, changing the syntax and factoring some duplicate code".
2367 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2368 for deliveries of the second and subsequent messages over the same SMTP
2371 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2372 ${reduce, with only minor "tidies".
2374 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2376 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2377 expansion side effects.
2379 PH/27 When a message times out after an over-quota error from an Exim-imposed
2380 quota, the bounce message says "mailbox is full". This message was not
2381 being given when it was a system quota that was exceeded. It now should
2384 MH/03 Made $recipients available in local_scan(). local_scan() already has
2385 better access to the recipient list through recipients_list[], but
2386 $recipients can be useful in postmaster-provided expansion strings.
2388 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2389 in the case of a MAIL command with additional options following the
2390 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2391 were accidentally chopped off.
2393 PH/29 SMTP synchronization checks are implemented when a command is read -
2394 there is a check that no more input is waiting when there shouldn't be
2395 any. However, for some commands, a delay in an ACL can mean that it is
2396 some time before the response is written. In this time, more input might
2397 arrive, invalidly. So now there are extra checks after an ACL has run for
2398 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2399 pipelining has not been advertised.
2401 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2403 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2404 returned in the SMTP response when smtp_return_error_details was false.
2405 This has been fixed.
2407 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2408 instead of the C I/O that was originally supplied, because problems were
2409 reported on Solaris.
2411 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2412 Exim which did not show up earlier: it was assuming that a call to
2413 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2414 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2415 was a macro that became an assignment, so it seemed to work. This has
2416 changed to a proper function call with a void return, hence the compile
2417 error. Exim's code has been fixed.
2419 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2422 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2424 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2426 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2427 because it turns out to be too big or there is a timeout) there is no
2428 "Completed" line in the log. When some messages of this type were
2429 selected by exigrep, they were listed as "not completed". Others were
2430 picked up by some special patterns. I have improved the selection
2431 criteria to be more general.
2433 PH/38 The host_find_failed option in the manualroute router can now be set
2434 to "ignore", to completely ignore a host whose IP address cannot be
2435 found. If all hosts are ignored, the behaviour is controlled by the new
2436 host_all_ignored option.
2438 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2439 homing or because of multiple MX records with /mx) generated more than
2440 one IP address, and the following item turned out to be the local host,
2441 all the secondary addresses of the first item were incorrectly removed
2442 from the list, along with the local host and any following hosts (which
2443 is what is supposed to happen).
2445 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2446 whoever called Exim into the -H file. In the case of the daemon it was
2447 behaving confusingly. When first started, it used values for whoever
2448 started the daemon, but after a SIGHUP it used the Exim user (because it
2449 calls itself on a restart). I have changed the code so that it now always
2452 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2453 message are rejected with the same error (e.g. no authentication or bad
2454 sender address), and a DATA command is nevertheless sent (as can happen
2455 with PIPELINING or a stupid MUA), the error message that was given to the
2456 RCPT commands is included in the rejection of the DATA command. This is
2457 intended to be helpful for MUAs that show only the final error to their
2460 PH/42 Another patch from the Sieve maintainer.
2462 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2463 Eximstats - Fixed some broken HTML links and added missing column headers
2465 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2466 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2468 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2470 PH/43 Yet another patch from the Sieve maintainer.
2472 PH/44 I found a way to check for a TCP/IP connection going away before sending
2473 the response to the final '.' that terminates a message, but only in the
2474 case where the client has not sent further data following the '.'
2475 (unfortunately, this is allowed). However, in many cases there won't be
2476 any further data because there won't be any more messages to send. A call
2477 to select() can be used: if it shows that the input is "ready", there is
2478 either input waiting, or the socket has been closed. An attempt to read
2479 the next input character can distinguish the two cases. Previously, Exim
2480 would have sent an OK response which the client would never have see.
2481 This could lead to message repetition. This fix should cure that, at
2482 least in a lot of common cases.
2484 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2485 advertised in response to EHLO.
2491 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2492 fixed by 4.65/MH/01 (is this a record?) are fixed:
2494 (i) An empty string was always treated as zero by the numeric comparison
2495 operators. This behaviour has been restored.
2497 (ii) It is documented that the numeric comparison operators always treat
2498 their arguments as decimal numbers. This was broken in that numbers
2499 starting with 0 were being interpreted as octal.
2501 While fixing these problems I realized that there was another issue that
2502 hadn't been noticed. Values of message_size_limit (both the global option
2503 and the transport option) were treated as octal if they started with 0.
2504 The documentation was vague. These values are now always treated as
2505 decimal, and I will make that clear in the documentation.
2511 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2512 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2515 MH/01 Don't check that the operands of numeric comparison operators are
2516 integers when their expansion is in "skipping" mode (fixes bug
2517 introduced by 4.64-PH/07).
2519 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2520 child addresses, Exim now panics and dies. Previously, because the count
2521 is held in a short int, deliveries were likely to be lost. As such a
2522 large number of recipients for a single message is ridiculous
2523 (performance will be very, very poor), I have chosen to impose a limit
2524 rather than extend the field.
2530 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2531 leftover -K file (the existence of which was triggered by #402).
2532 While we were at it, introduced process PID as part of the -K
2533 filename. This should rule out race conditions when creating
2536 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2537 processing considerably. Previous code took too long for large mails,
2538 triggering a timeout which in turn triggers #401.
2540 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2541 in the DK code in transports.c. sendfile() is not really portable,
2542 hence the _LINUX specificness.
2544 TF/01 In the add_headers option to the mail command in an Exim filter,
2545 there was a bug that Exim would claim a syntax error in any
2546 header after the first one which had an odd number of characters
2549 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2550 callout verification, Exim cached a "reject" for the entire domain. This
2551 is correct for most verifications, but it is not correct for a recipient
2552 verification with use_sender or use_postmaster set, because in that case
2553 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2554 case of MAIL FROM:<> rejection from other early rejections (e.g.
2555 rejection of HELO). When verifying a recipient using a non-null MAIL
2556 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2557 Whatever the result of the callout, the value of the domain cache is
2558 left unchanged (for any other kind of callout, getting as far as trying
2559 RCPT means that the domain itself is ok).
2561 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2564 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2565 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2566 session, the daemon ignores ECONNECT errors and logs others; it now
2567 ignores EPIPE as well.
2569 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2570 (quoted-printable decoding).
2572 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2573 later the small subsequent patch to fix an introduced bug.
2575 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2577 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2579 PH/08 An error is now given if message_size_limit is specified negative.
2581 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2582 to be given (somewhat) arbitrary names.
2584 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2587 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2588 miscellaneous code fixes
2590 PH/10 Added the log_reject_target ACL modifier to specify where to log
2593 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2594 hostname. This is wrong, because it relates to the incoming message (and
2595 probably the interface on which it is arriving) and not to the outgoing
2596 callout (which could be using a different interface). This has been
2597 changed to use the value of the helo_data option from the smtp transport
2598 instead - this is what is used when a message is actually being sent. If
2599 there is no remote transport (possible with a router that sets up host
2600 addresses), $smtp_active_hostname is used.
2602 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2603 tweaks were necessary in order to get it to work (see also 21 below):
2604 (a) The code assumed that strncpy() returns a negative number on buffer
2605 overflow, which isn't the case. Replaced with Exim's string_format()
2607 (b) There were several signed/unsigned issues. I just did the minimum
2608 hacking in of casts. There is scope for a larger refactoring.
2609 (c) The code used strcasecmp() which is not a standard C function.
2610 Replaced with Exim's strcmpic() function.
2611 (d) The code set only $1; it now sets $auth1 as well.
2612 (e) A simple test gave the error "authentication client didn't specify
2613 service in request". It would seem that Dovecot has changed its
2614 interface. Fortunately there's a specification; I followed it and
2615 changed what the client sends and it appears to be working now.
2617 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2620 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2621 address A is aliased to B and C, where B exists and C does not. Without
2622 -v the output is "A verified" because verification stops after a
2623 successful redirection if more than one address is generated. However,
2624 with -v the child addresses are also verified. Exim was outputting "A
2625 failed to verify" and then showing the successful verification for C,
2626 with its parentage. It now outputs "B failed to verify", showing B's
2627 parentage before showing the successful verification of C.
2629 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2630 look up a TXT record in a specific list after matching in a combined
2633 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2634 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2635 they consult the DNS. I had assumed they would set it the way they
2636 wanted; and indeed my experiments on Linux seem to show that in some
2637 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2638 To be on the safe side, however, I have now made the interface to
2639 host_find_byname() similar to host_find_bydns(), with an argument
2640 containing the DNS resolver options. The host_find_byname() function now
2641 sets these options at its start, just as host_find_bydns() does. The smtp
2642 transport options dns_qualify_single and dns_search_parents are passed to
2643 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2644 of host_find_byname() use the default settings of RES_DEFNAMES
2645 (qualify_single) but not RES_DNSRCH (search_parents).
2647 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2648 spool_read_header() do less string testing, by means of a preliminary
2649 switch on the second character of optional "-foo" lines. (This is
2650 overdue, caused by the large number of possibilities that now exist.
2651 Originally there were few.) While I was there, I also converted the
2652 str(n)cmp tests so they don't re-test the leading "-" and the first
2653 character, in the hope this might squeeze out yet more improvement.
2655 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2656 flag allowing group syntax was set by the header_syntax check but not
2657 turned off, possible causing trouble later; (2) The flag was not being
2658 set at all for the header_verify test, causing "group"-style headers to
2659 be rejected. I have now set it in this case, and also caused header_
2660 verify to ignore an empty address taken from a group. While doing this, I
2661 came across some other cases where the code for allowing group syntax
2662 while scanning a header line wasn't quite right (mostly, not resetting
2663 the flag correctly in the right place). These bugs could have caused
2664 trouble for malformed header lines. I hope it is now all correct.
2666 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2667 with the "reply" argument non-NULL. The code, however (which originally
2668 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2669 but it didn't always do it. This confused somebody who was copying the
2670 code for some other use. I have removed all the tests.
2672 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2673 feature that was used to support insecure browsers during the U.S. crypto
2674 embargo. It requires special client support, and Exim is probably the
2675 only MTA that supported it -- and would never use it because real RSA is
2676 always available. This code has been removed, because it had the bad
2677 effect of slowing Exim down by computing (never used) parameters for the
2678 RSA_EXPORT functionality.
2680 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2681 authenticator to fail if there's a tab character in the incoming data
2682 (there should never be unless someone is messing about, as it's supposed
2683 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2684 if the connection is using TLS or if the remote IP is the same as the
2685 local IP, and the "valid-client-cert option" if a client certificate has
2688 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2689 authenticators. This can be used for authorization after authentication
2690 succeeds. (In the case of plaintext, it servers for both authentication
2693 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2694 if any retry times were supplied.
2696 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2697 connection, where there is no client IP address to check. In this
2698 situation, the verify now always succeeds.
2700 PH/25 Applied John Jetmore's -Mset patch.
2702 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2704 PH/27 In a string expansion for a processed (not raw) header when multiple
2705 headers of the same name were present, leading whitespace was being
2706 removed from all of them, but trailing whitespace was being removed only
2707 from the last one. Now trailing whitespace is removed from each header
2708 before concatenation. Completely empty headers in a concatenation (as
2709 before) are ignored.
2711 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2712 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2714 PH/29 [Removed. This was a change that I later backed out, and forgot to
2715 correct the ChangeLog entry (that I had efficiently created) before
2716 committing the later change.]
2718 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2719 address errors (4xx response to RCPT) over the same connection as other
2720 messages routed to the same hosts. Such deliveries are always "forced",
2721 so retry times are not inspected. This resulted in far too many retries
2722 for the affected addresses. The effect occurred only when there were more
2723 hosts than the hosts_max_try setting in the smtp transport when it had
2724 the 4xx errors. Those hosts that it had tried were not added to the list
2725 of hosts for which the message was waiting, so if all were tried, there
2726 was no problem. Two fixes have been applied:
2728 (i) If there are any address or message errors in an SMTP delivery, none
2729 of the hosts (tried or untried) are now added to the list of hosts
2730 for which the message is waiting, so the message should not be a
2731 candidate for sending over the same connection that was used for a
2732 successful delivery of some other message. This seems entirely
2733 reasonable: after all the message is NOT "waiting for some host".
2734 This is so "obvious" that I'm not sure why it wasn't done
2735 previously. Hope I haven't missed anything, but it can't do any
2736 harm, as the worst effect is to miss an optimization.
2738 (ii) If, despite (i), such a delivery is accidentally attempted, the
2739 routing retry time is respected, so at least it doesn't keep
2740 hammering the server.
2742 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2743 in ${readsocket because some servers need this prod.
2745 PH/32 Added some extra debug output when updating a wait-xxx database.
2747 PH/33 The hint "could be header name not terminated by colon", which has been
2748 given for certain expansion errors for a long time, was not being given
2749 for the ${if def:h_colon_omitted{... case.
2751 PH/34 The spec says: "With one important exception, whenever a domain list is
2752 being scanned, $domain contains the subject domain." There was at least
2753 one case where this was not true.
2755 PH/35 The error "getsockname() failed: connection reset by peer" was being
2756 written to the panic log as well as the main log, but it isn't really
2757 panic-worthy as it just means the connection died rather early on. I have
2758 removed the panic log writing for the ECONNRESET error when getsockname()
2761 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2762 runs only) independently of the message's sender address. This meant
2763 that, if the 4xx error was in fact related to the sender, a different
2764 message to the same recipient with a different sender could confuse
2765 things. In particular, this can happen when sending to a greylisting
2766 server, but other circumstances could also provoke similar problems.
2767 I have changed the default so that the retry time for these errors is now
2768 based a combination of the sender and recipient addresses. This change
2769 can be overridden by setting address_retry_include_sender=false in the
2772 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2773 remote server are returned as part of bounce messages. This was not
2774 happening for LMTP over a pipe (the lmtp transport), but now it is the
2775 same for both kinds of LMTP.
2777 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2778 in header lines that were in fact CNAMEs. This is no longer the case.
2780 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2781 and queue runs started by the daemon processed all messages. This has
2782 been fixed so that -R and -S can now usefully be given with -q<time>.
2784 PH/40 Import PCRE release 6.7 (fixes some bugs).
2786 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2788 PH/42 Give an error if -q is specified more than once.
2790 PH/43 Renamed the variables $interface_address and $interface_port as
2791 $received_ip_address and $received_port, to make it clear that these
2792 values apply to message reception, and not to the outgoing interface when
2793 a message is delivered. (The old names remain recognized, of course.)
2795 PH/44 There was no timeout on the connect() call when using a Unix domain
2796 socket in the ${readsocket expansion. There now is.
2798 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2799 be meaningful with "accept".
2801 SC/01 Eximstats V1.43
2802 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2804 SC/02 Eximstats V1.44
2805 Use a glob alias rather than an array ref in the generated
2806 parser. This improves both readability and performance.
2808 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2809 Collect SpamAssassin and rejection statistics.
2810 Don't display local sender or destination tables unless
2811 there is data to show.
2812 Added average volumes into the top table text output.
2814 SC/04 Eximstats V1.46
2815 Collect data on the number of addresses (recipients)
2816 as well as the number of messages.
2818 SC/05 Eximstats V1.47
2819 Added 'Message too big' to the list of mail rejection
2820 reasons (thanks to Marco Gaiarin).
2822 SC/06 Eximstats V1.48
2823 Mainlog lines which have GMT offsets and are too short to
2824 have a flag are now skipped.
2826 SC/07 Eximstats V1.49 (Alain Williams)
2827 Added the -emptyok flag.
2829 SC/08 Eximstats V1.50
2830 Fixes for obtaining the IP address from reject messages.
2832 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2833 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2834 whitespace changes from 4.64-PH/27
2836 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2839 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2840 are found, allow negative numbers in numeric criteria)
2842 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2844 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2845 to match changes made in 4.64-PH/43
2847 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2849 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2850 "domain" is actually an IP address, and force a failure. This locks out
2851 those revolvers/nameservers that support "A-for-A" lookups, in
2852 contravention of the specifications.
2854 PH/48 When a host name was looked up from an IP address, and the subsequent
2855 forward lookup of the name timed out, the host name was left in
2856 $sender_host_name, contrary to the specification.
2858 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2859 restricted to single-key lookups, Exim was not diagnosing an error if
2860 * or *@ was used with a query-style lookup.
2862 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2864 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2865 long ago, but noone interested enough thought of it. Let's just say that
2866 the "1.1" means that there are some new functions that weren't there at
2867 some point in the past.
2869 PH/51 Error processing for expansion failure of helo_data from an smtp
2870 transport during callout processing was broken.
2872 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2873 tested/used via the -bh/-bhc/-bs options.
2875 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2876 bug, fixed in subsequent PCRE releases).
2878 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2879 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2881 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2887 SC/01 Use a glob alias rather than an array ref in eximstats generated
2888 parser. This improves both readability and performance.
2890 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2891 Don't display local sender or destination tables in eximstats unless
2892 there is data to show.
2893 Added average volumes into the eximstats top table text output.
2895 SC/03 Collect data on the number of addresses (recipients) as well
2896 as the number of messages in eximstats.
2898 TF/01 Correct an error in the documentation for the redirect router. Exim
2899 does (usually) call initgroups() when daemonizing.
2901 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2902 with consistent privilege compared to when running as a daemon.
2904 TF/03 Note in the spec that $authenticated_id is not set for local
2905 submissions from trusted users.
2907 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2908 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2910 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2911 by adding some example configuration directives to the default
2912 configuration file. A little bit of work is required to uncomment the
2913 directives and define how usernames and passwords are checked, but
2914 there is now a framework to start from.
2916 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2917 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2918 without this. I don't know how relevant this is to other LDAP libraries.
2920 PH/02 Add the verb name to the "unknown ACL verb" error.
2922 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2924 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2926 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2927 directory not expanded when it should be if an expanded home directory
2928 was set for the address (which is overridden by the transport).
2930 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2933 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2934 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2935 because it is too late at that time, and has no effect.
2937 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2938 security issue with \' (bugzilla #107). I could not use the
2939 PQescapeStringConn() function, because it needs a PGconn value as one of
2942 PH/08 When testing addresses using -bt, indicate those final addresses that
2943 are duplicates that would not cause an additional delivery. At least one
2944 person was confused, thinking that -bt output corresponded to deliveries.
2945 (Suppressing duplicates isn't a good idea as you lose the information
2946 about possibly different redirections that led to the duplicates.)
2948 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2949 systems where poll() doesn't work, in particular OS X.
2951 PH/10 Added more information to debugging output for retry time not reached.
2953 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2954 operations in malware.c.
2956 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2959 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2960 syslog_duplication set false, log lines that would normally be written
2961 both the the main log and to the reject log were not written to syslog at
2964 PH/14 In the default configuration, change the use of "message" in ACL warn
2965 statements to "add_header".
2967 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2968 not followed by a command (e.g. "seen endif").
2970 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2971 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2974 PH/17 Added extra conditions to the default value of delay_warning_condition
2978 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2979 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2980 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2983 The Auto-Submitted: and various List- headers are standardised, whereas I
2984 don't think Precedence: ever was.
2986 PH/18 Refactored debugging code in route_finduser() to show more information,
2987 in particular, the error code if getpwnam() issues one.
2989 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2990 This is apparently needed in addition to the PH/07 change above to avoid
2991 any possible encoding problems.
2993 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2994 but not after initializing Perl.
2996 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2997 output them only if debugging. By default they are written stderr,
2998 apparently, which is not desirable.
3000 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3003 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3006 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3008 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3009 authenticated or an ident call has been made. Suppress the default
3010 values for $authenticated_id and $authenticated_sender (but permit -oMai
3011 and -oMas) when testing with -bh.
3013 PH/24 Re-jigged the order of the tests in the default configuration so that the
3014 tests for valid domains and recipients precede the DNS black list and CSA
3015 tests, on the grounds that those ones are more expensive.
3017 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3018 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3019 This bug exists in every version of Exim that I still have, right back to
3022 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3023 However, an attempt to turn on case-sensitivity in a regex key by
3024 including (?-i) didn't work because the subject string was already
3025 lowercased, and the effects were non-intuitive. It turns out that a
3026 one-line patch can be used to allow (?-i) to work as expected.
3032 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3033 other effects) broke the use of negated acl sub-conditions.
3035 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3038 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3039 "Deny" causes Exim to reject the incoming connection with a 554 error.
3040 Unfortunately, if there is a major crisis, such as a disk failure,
3041 tcp-wrappers gives "deny", whereas what one would like would be some
3042 kind of temporary error. A kludge has been added to help with this.
3043 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3044 554 error is used if errno is still zero or contains ENOENT (which occurs
3045 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3048 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3050 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3051 errors. Otherwise a message that provokes a temporary error (when other
3052 messages do not) can cause a whole host to time out.
3054 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3055 addresses were routed directly to files or pipes from a redirect router.
3056 File deliveries just didn't batch; pipe deliveries might have suffered
3059 PH/06 A failure to get a lock for a hints database would erroneously always say
3060 "Failed to get write lock", even when it was really a read lock.
3062 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3063 of 0600. This has been changed to use the value of the lockfile_mode
3064 option (which defaults to 0600).
3066 PH/08 Applied small patch from the Sieve maintainer.
3068 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3069 folder from quota calculations, a direct delivery into this folder messed
3070 up the contents of the maildirsize file. This was because the regex was
3071 used only to exclude .Trash (or whatever) when the size of the mailbox
3072 was calculated. There was no check that a delivery was happening into an
3073 excluded directory. This bug has been fixed by ignoring all quota
3074 processing for deliveries into excluded directories.
3076 PH/10 Added the maildirfolder_create_regex option to appendfile.
3082 PH/01 The code for finding all the local interface addresses on a FreeBSD
3083 system running IPv6 was broken. This may well have applied to all BSD
3084 systems, as well as to others that have similar system calls. The broken
3085 code found IPv4 interfaces correctly, but gave incorrect values for the
3086 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3087 that it would not match correctly against @[] and not recognize the IPv6
3090 PH/02 The ipliteral router was not recognizing addresses of the form user@
3091 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3093 PH/03 Added disable_ipv6.
3095 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3096 decoded form, because it is most often used to construct To: headers
3097 lines in autoreplies, and the decoded form may well be syntactically
3098 invalid. However, $reply_address has leading white space removed, and all
3099 newlines turned into spaces so that the autoreply transport does not
3102 PH/05 If group was specified without a user on a router, and no group or user
3103 was specified on a transport, the group from the router was ignored.
3105 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3106 for visible compile-time settings that can be used to change these
3107 numbers, for those that want even more. Backwards compatibility with old
3108 spool files has been maintained. However, going back to a previous Exim
3109 release will lost any variables that are in spool files.
3111 PH/07 Two small changes when running in the test harness: increase delay when
3112 passing a TCP/IP connection to a new process, in case the original
3113 process has to generate a bounce, and remove special handling of
3114 127.0.0.2 (sic), which is no longer necessary.
3116 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3117 be the same on different OS.
3119 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3122 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3123 whether --show-vars was specified or not
3125 JJ/02 exipick: Added support for new ACL variable spool format introduced
3128 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3129 syntactically invalid From: or Reply-to: line, and a filter used this to
3130 generate an autoreply, and therefore failed to obtain an address for the
3131 autoreply, Exim could try to deliver to a non-existent relative file
3132 name, causing unrelated and misleading errors. What now happens is that
3133 it logs this as a hard delivery error, but does not attempt to create a
3136 PH/11 The exinext utility has a -C option for testing purposes, but although
3137 the given file was scanned by exinext itself; it wasn't being passed on
3138 when Exim was called.
3140 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3141 an end-of-file indication when reading a command response.
3143 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3144 compiled. In many other places in Exim, IPv6 addresses are always
3145 recognized, so I have changed this. It also means that IPv4 domain
3146 literals of the form [IPV4:n.n.n.n] are now always recognized.
3148 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3149 used if the router is not running as root, for example, when verifying at
3150 ACL time, or when using -bh. The debugging output from this situation was
3151 non-existent - all you got was a failure to exec. I have made two
3154 (a) Failures to set uid/gid, the current directory, or a process leader
3155 in a subprocess such as that created by queryprogram now generate
3156 suitable debugging output when -d is set.
3158 (b) The queryprogram router detects when it is not running as root,
3159 outputs suitable debugging information if -d is set, and then runs
3160 the subprocess without attempting to change uid/gid.
3162 PH/15 Minor change to Makefile for building test_host (undocumented testing
3165 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3166 additional section of a DNS packet that returns MX or SRV records.
3167 Instead, it always explicitly searches for A/AAAA records. This avoids
3168 major problems that occur when a DNS server includes only records of one
3169 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3170 fixed another bug: if SRV records were looked up and the corresponding
3171 address records were *not* found in the additional section, the port
3172 values from the SRV records were lost.
3174 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3175 using the correct key (the original address) when searching the retry
3176 rules in order to find which one to use for generating the retry hint.
3178 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3179 adding the default one. Similarly, if it contains a Reply-To: header, the
3180 errors_reply_to option, if set, is not used.
3182 PH/19 When calculating a retry time, Exim used to measure the "time since
3183 failure" by looking at the "first failed" field in the retry record. Now
3184 it does not use this if it is later than than the arrival time of the
3185 message. Instead it uses the arrival time. This makes for better
3186 behaviour in cases where some deliveries succeed, thus re-setting the
3187 "first failed" field. An example is a quota failure for a huge message
3188 when small messages continue to be delivered. Without this change, the
3189 "time since failure" will always be short, possible causing more frequent
3190 delivery attempts for the huge message than are intended.
3191 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3193 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3194 $1, $2, $3) because the numerical variables can be reset during some
3195 expansion items (e.g. "match"), thereby losing the authentication data.
3197 PH/21 Make -bV show the size of off_t variables so that the test suite can
3198 decide whether to run tests for quotas > 2G.
3200 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3201 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3202 filecount value is greater than 2G or if a quota value is greater than 2G
3203 on a system where the size of off_t is not greater than 4, a panic error
3206 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3207 never match. The debug and -bh output now contains an explicit error
3208 message indicating a malformed IPv4 address or mask.
3210 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3211 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3212 PH/23 above applies.
3214 PH/25 Do not write to syslog when running in the test harness. The only
3215 occasion when this arises is a failure to open the main or panic logs
3216 (for which there is an explicit test).
3218 PH/26 Added the /no_tell option to "control=freeze".
3220 PH/27 If a host name lookup failed very early in a connection, for example, if
3221 the IP address matched host_lookup and the reverse lookup yielded a name
3222 that did not have a forward lookup, an error message of the form "no IP
3223 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3224 could be logged. Now it outputs the IP address instead of "NULL".
3226 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3227 allows the sender and the authenticated sender to be set when
3228 submitting a message from within Exim. Since child_open_exim() is
3229 documented for local_scan(), the new function should be too.
3231 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3232 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3233 results in an empty string is now treated as unset.
3235 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3237 PH/31 Added sender_verify_fail logging option.
3239 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3240 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3241 tidied the source and removed it altogether.
3243 PH/33 When a queue run was abandoned because the load average was too high, a
3244 log line was always written; now it is written only if the queue_run log
3245 selector is set. In addition, the log line for abandonment now contains
3246 information about the queue run such as the pid. This is always present
3247 in "start" and "stop" lines but was omitted from the "abandon" line.
3249 PH/34 Omit spaces between a header name and the colon in the error message that
3250 is given when verify = headers_syntax fails (if there are lots of them,
3251 the message gets confusing).
3253 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3254 names, as there are now some PTR records that contain slashes. This check
3255 is only to protect against broken name servers that fall over on strange
3256 characters, so the fact that it applies to all lookups doesn't matter.
3258 PH/36 Now that the new test suite is complete, we can remove some of the
3259 special code in Exim that was needed for the old test suite. For example,
3260 sorting DNS records because real resolvers return them in an arbitrary
3261 order. The new test suite's fake resolver always returns records in the
3264 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3265 bounces) except when queue_only is set, to avoid logging races between
3266 the different processes.
3268 PH/38 Panic-die if .include specifies a non-absolute path.
3270 PH/39 A tweak to the "H" retry rule from its user.
3272 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3273 a label. They prevented compilation on older perls.
3275 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3276 a warning to be raised on newish perls.
3278 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3279 on queue. Changes to match documented behaviour of showing count of
3280 messages matching specified criteria.
3282 PH/40 Changed the default ident timeout from 30s to 5s.
3284 PH/41 Added support for the use of login_cap features, on those BSD systems
3285 that have them, for controlling the resources used by pipe deliveries.
3287 PH/42 The content-scanning code uses fopen() to create files in which to put
3288 message data. Previously it was not paying any attention to the mode of
3289 the files. Exim runs with umask(0) because the rest of the code creates
3290 files with open(), and sets the required mode explicitly. Thus, these
3291 files were ending up world-writeable. This was not a big issue, because,
3292 being within the spool directory, they were not world-accessible. I have
3293 created a function called modefopen, which takes an additional mode
3294 argument. It sets umask(777), creates the file, chmods it to the required
3295 mode, then resets the umask. All the relevant calls to fopen() in the
3296 content scanning code have been changed to use this function.
3298 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3299 to 24 hours. This avoids potential overflow problems when processing G
3300 and H retry rules. I suspect nobody ever tinkers with this value.
3302 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3304 PH/45 When the plaintext authenticator is running as a client, the server's
3305 challenges are checked to ensure they are valid base64 strings. By
3306 default, the authentication attempt is cancelled if an invalid string is
3307 received. Setting client_ignore_invalid_base64 true ignores these errors.
3308 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3309 they are received. Thus, the responses can be made to depend on the
3310 challenges. If an invalid string is ignored, an empty string is placed in
3313 PH/46 Messages that are created by the autoreply transport now contains a
3314 References: header, in accordance with RFCs 2822 and 3834.
3316 PH/47 Added authenticated_sender_force to the smtp transport.
3318 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3320 PH/49 Installed latest patch from the Sieve maintainer.
3322 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3323 was also set, the appendfile transport was unnecessarily scanning a
3324 directory of message files (e.g. for maildir delivery) to find the count
3325 of files (along with the size), even though it did not need this
3326 information. It now does the scan only if it needs to find either the
3327 size of the count of files.
3329 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3331 PH/52 Two bugs concerned with error handling when the smtp transport is
3334 (i) Exim was not creating retry information for temporary errors given
3335 for individual recipients after the DATA command when the smtp transport
3336 was used in LMTP mode. This meant that they could be retried too
3337 frequently, and not timed out correctly.
3339 (ii) Exim was setting the flag that allows error details to be returned
3340 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3341 recipients that were returned after the DATA command.
3343 PH/53 This is related to PH/52, but is more general: for any failing address,
3344 when detailed error information was permitted to be returned to the
3345 sender, but the error was temporary, then after the final timeout, only
3346 "retry timeout exceeded" was returned. Now it returns the full error as
3347 well as "retry timeout exceeded".
3349 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3350 do this, and (what is worse) MTAs that accept it.
3352 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3353 will now be deprecated.
3355 PH/56 New os.c-cygwin from the Cygwin maintainer.
3357 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3358 formats (previously only available in exim formats via -bpr, -bpru,
3359 and -bpra. Now also available in native and exiqgrep formats)
3361 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3362 with very large, slow to parse queues
3364 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3366 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3368 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3369 responses to authentication challenges, though it was showing the
3370 challenges; (ii) I've removed the CR characters from the debug output for
3373 PH/58 Allow for the insertion of a newline as well as a space when a string
3374 is turned into more than one encoded-word during RFC 2047 encoding. The
3375 Sieve code now uses this.
3377 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3378 data_4xx, lost_connection, tls_required.
3380 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3381 message was being sent as an SMTP response.
3383 PH/61 Add -l and -k options to exicyclog.
3385 PH/62 When verifying, if an address was redirected to one new address, so that
3386 verification continued, and the new address failed or deferred after
3387 having set something in $address_data, the value of $address_data was not
3388 passed back to the ACL. This was different to the case when no
3389 redirection occurred. The value is now passed back in both cases.
3391 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3392 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3393 use login_cap.h, so on its own it isn't the distinguishing feature. The
3394 new name refers directly to the setclassresources() function.
3396 PH/65 Added configuration files for NetBSD3.
3398 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3400 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3401 is preferred over IPv4.
3403 PH/68 The bounce_return_message and bounce_return_body options were not being
3404 honoured for bounces generated during the reception of non-SMTP messages.
3405 In particular, this applied to messages rejected by the ACL. This bug has
3406 been fixed. However, if bounce_return_message is true and bounce_return_
3407 body is false, the headers that are returned for a non-SMTP message
3408 include only those that have been read before the error was detected.
3409 (In the case of an ACL rejection, they have all been read.)
3411 PH/69 The HTML version of the specification is now built in a directory called
3412 spec_html instead of spec.html, because the latter looks like a path with
3413 a MIME-type, and this confuses some software.
3415 PH/70 Catch two compiler warnings in sieve.c.
3417 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3418 function verify_get_ident() calls ip_connect() to connect a socket, but
3419 if the "connect()" function timed out, ip_connect() used to close the
3420 socket. However, verify_get_ident() also closes the socket later, and in
3421 between Exim writes to the log, which may get opened at this point. When
3422 the socket was closed in ip_connect(), the log could get the same file
3423 descriptor number as the socket. This naturally causes chaos. The fix is
3424 not to close the socket in ip_connect(); the socket should be closed by
3425 the function that creates it. There was only one place in the code where
3426 this was missing, in the iplookup router, which I don't think anybody now
3427 uses, but I've fixed it anyway.
3429 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3430 well as to direct DNS lookups. Otherwise the handling of names in host
3431 lists is inconsistent and therefore confusing.
3437 PH/01 Two changes to the default runtime configuration:
3439 (1) Move the checks for relay_from_hosts and authenticated clients from
3440 after to before the (commented out) DNS black list checks.
3442 (2) Add control=submission to the relay_from_hosts and authenticated
3443 clients checks, on the grounds that messages accepted by these
3444 statements are most likely to be submissions.
3446 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3448 (1) Generate an error if the third argument for the ${prvs expansion is
3451 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3454 (3) Reset the variables that are obtained from the first argument of
3455 ${prvscheck and used in the second argument before leaving the code,
3456 because their memory is reclaimed, so using them afterwards may do
3459 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3460 one (it's much easier than Tom thought :-).
3462 (5) Because of (4), we can now allow for the use of $prvscheck_result
3463 inside the third argument.
3465 PH/03 For some reason, the default setting of PATH when running a command from
3466 a pipe transport was just "/usr/bin". I have changed it to
3469 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3470 anything to be listed in the output from -bV.
3472 PH/05 When a filter generated an autoreply, the entire To: header line was
3473 quoted in the delivery log line, like this:
3475 => >A.N.Other <ano@some.domain> <original@ddress> ...
3477 This has been changed so that it extracts the operative address. There
3478 may be more than one such address. If so, they are comma-separated, like
3481 => >ano@some.domain,ona@other.domain <original@ddress> ...
3483 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3484 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3485 not being looked up in the rDNS to get a host name, Exim was showing the
3486 IP address twice in Received: lines, even though the IP addresses were
3487 identical. For example:
3489 Received: from [1.2.3.4] (helo=[1.2.3.4])
3491 However, if the real host name was known, it was omitting the HELO data
3492 if it matched the actual IP address. This has been tidied up so that it
3493 doesn't show the same IP address twice.
3495 PH/07 When both +timestamp and +memory debugging was on, the value given by
3496 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3497 called by the debug printing, thereby overwriting the timestamp buffer.
3498 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3500 PH/08 When the original message was included in an autoreply transport, it
3501 always said "this is a copy of the message, including all the headers",
3502 even if body_only or headers_only was set. It now gives an appropriate
3505 PH/09 Applied a patch from the Sieve maintainer which:
3507 o fixes some comments
3508 o adds the (disabled) notify extension core
3509 o adds some debug output for the result of if/elsif tests
3510 o points to the current vacation draft in the documentation
3511 and documents the missing references header update
3515 o fixes a bug in processing the envelope test (when testing
3516 multiple envelope elements, the last element determined the
3519 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3520 Electronic Mail") by including:
3522 Auto-submitted: auto-generated
3524 in the messages that it generates (bounce messages and others, such as
3525 warnings). In the case of bounce messages for non-SMTP messages, there was
3526 also a typo: it was using "Auto_submitted" (underscore instead of
3527 hyphen). Since every message generated by Exim is necessarily in response
3528 to another message, thes have all been changed to:
3530 Auto-Submitted: auto-replied
3532 in accordance with these statements in the RFC:
3534 The auto-replied keyword:
3536 - SHOULD be used on messages sent in direct response to another
3537 message by an automatic process,
3539 - MUST NOT be used on manually-generated messages,
3541 - MAY be used on Delivery Status Notifications (DSNs) and Message
3542 Disposition Notifications (MDNs),
3544 - MUST NOT be used on messages generated by automatic or periodic
3545 processes, except for messages which are automatic responses to
3548 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3549 to the default Received: header definition.
3551 PH/12 Added log selector acl_warn_skipped (default on).
3553 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3554 variables because (a) they are in the wrong storage pool and (b) even if
3555 they were copied, it wouldn't work properly because of the caching.
3557 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3558 checking when decoding. Apparently there are clients that generate
3559 overlong encoded strings. Why am I not surprised?
3561 PH/15 If the first argument of "${if match_address" was not empty, but did not
3562 contain an "@" character, Exim crashed. Now it writes a panic log message
3563 and treats the condition as false.
3565 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3567 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3568 extension "envelope-auth". The code is finished and in agreement with
3569 other implementations, but there is no documentation so far and in fact,
3570 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3571 not changing the active code.
3573 Print executed "if" and "elsif" statements when debugging is used. This
3574 helps a great deal to understand what a filter does.
3576 Document more things not specified clearly in RFC3028. I had all this
3577 sorted out, when out of a sudden new issues came to my mind. Oops."
3579 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3582 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3583 canonical form (as documented). However, after a host name lookup from
3584 the IP address, check_host() was doing a simple string comparison with
3585 addresses acquired from the DNS when checking that the found name did
3586 have the original IP as one of its addresses. Since any found IPv6
3587 addresses are likely to be in abbreviated form, the comparison could
3588 fail. Luckily, there already exists a function for doing the comparison
3589 by converting both addresses to binary, so now that is used instead of
3590 the text comparison.
3592 PH/20 There was another similar case to PH/19, when a complete host name was
3593 given in a host list; looking up its IP address could give an abbreviated
3594 form, whereas the current host's name might or might not be abbreviated.
3595 The same fix has been applied.
3601 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3602 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3605 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3606 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3608 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3610 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3611 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3612 Also, for "Auto-Submitted", treat anything other than "no" as
3613 non-personal, in accordance with RFC 3834. (Previously it treated
3614 anything starting "auto-" as non-personal.)
3616 TF/01 The control=submission/name=... option had a problem with syntax
3617 errors if the name included a slash character. The /name= option
3618 now slurps the rest of the string, so it can include any characters
3619 but it must come last in the list of options (after /sender_retain
3622 PH/05 Some modifications to the interface to the fake nameserver for the new
3630 TK/01 Added the "success_on_redirect" address verification option. See
3631 NewStuff for rationale and an example.
3633 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3635 PH/02 Patch to exigrep to allow it to work on syslog lines.
3637 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3638 fread() to skip over the body file's header line, because in Cygwin the
3639 header line is locked and is inaccessible.
3641 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3642 co-exist for some time) to make it clear that it is the Exim ID that is
3643 referenced, not the Message-ID: header line.
3645 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3646 string_format() function, because snprintf() does not exist on all
3649 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3650 ${stat: expansion item.
3652 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3653 protocol synchronization error", to keep the pedants happy.
3655 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3656 well as for IRIX systems, when gcc is being used. See the host.c source
3659 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3661 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3664 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3665 transport and to the smtp transport in LMTP mode.
3667 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3669 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3670 run a filter in a subprocess. This could lead to confusion in subsequent
3671 lookups in the parent process. There should also be a search_tidyup() at
3672 the end of the subprocess.
3674 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3675 only if the host matched helo_try_verify_hosts, which caused the
3676 verification to occur when the EHLO/HELO command was issued. The ACL just
3677 tested the remembered result. Now, if a previous verification attempt has
3678 not happened, "verify = helo" does it there and then.
3680 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3682 TK/03 Fix log output including CR from clamd.
3684 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3685 exist provoked a memory error which could cause a segfault.
3687 PH/15 Installed PCRE 6.2
3689 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3691 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3692 of the problem. Specifically, suggested +O2 rather than +O1 for the
3695 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3697 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3698 record, the port was not correctly logged when the outgoing_port log
3699 selector was set (it logged the transort's default port).
3701 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3702 fallback_hosts, and "hosts" in the smtp transport.
3704 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3705 host errors such as "Connection refused".
3707 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3708 authentication with radiusclient 0.4.9:
3710 - Error returned from rc_read_config was caught wrongly
3711 - Username/password not passed on to radius server due to wrong length.
3713 The presumption is that some radiusclient API changes for 4.51/PH/17
3714 were not taken care of correctly. The code is still untested by me (my
3715 Linux distribution still has 0.3.2 of radiusclient), but it was
3716 contributed by a Radius user.
3718 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3719 expanding the "port" option of the smtp transport.
3721 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3722 while reading a MIME header. Thanks to Tom Hughes for a patch.
3724 PH/24 Include config.h inside local_scan.h so that configuration settings are
3727 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3728 that in an ACL for RCPT (for example), you can examine exactly what was
3731 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3732 commands, but it was not correctly comparing the address with the actual
3733 client host address. Thus, it would show the EHLO address in Received:
3734 header lines when this was not necessary.
3736 PH/27 Added the % operator to ${eval:}.
3738 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3739 it should be ignoring failures (because with -C, for example, it has lost
3740 privilege). It wasn't ignoring creation failures other than "already
3743 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3746 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3747 because an input line was too long, either on its own, or by virtue of
3748 too many continuations, the temporary file was not being removed, and the
3749 return code was incorrect.
3751 PH/31 Missing "BOOL" in function definition in filtertest.c.
3753 PH/32 Applied Sieve patches from the maintainer.
3755 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3757 PH/33 Added "verify = not_blind".
3759 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3760 Local/Makefile (with some defaults set). These are used in built scripts
3761 such as exicyclog, but they have never been used in the exim_install
3762 script (though there are many overriding facilities there). I have
3763 arranged that the exim_install script now takes note of these two
3766 PH/35 Installed configuration files for Dragonfly.
3768 PH/36 When a locally submitted message by a trusted user did not contain a
3769 From: header, and the sender address was obtained from -f or from an SMTP
3770 MAIL command, and the trusted user did not use -F to supply a sender
3771 name, $originator_name was incorrectly used when constructing a From:
3772 header. Furthermore, $originator_name was used for submission mode
3773 messages from external hosts without From: headers in a similar way,
3774 which is clearly wrong.
3776 PH/37 Added control=suppress_local_fixups.
3778 PH/38 When log_selector = +received_sender was set, and the addition of the
3779 sender made the log line's construction buffer exactly full, or one byte
3780 less than full, an overflow happened when the terminating "\n" was
3783 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3784 when the result of a list match is failure because a DNS lookup failed.
3786 PH/40 RM_COMMAND is now used in the building process.
3788 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3789 the "build-* directories that it finds.
3791 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3792 address in a domain literal was a prefix of an interface address.
3794 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3795 when verifying a sender address, unless rewrite_headers is false.
3797 PH/44 Wrote a long comment about why errors_to addresses are verified as
3798 recipients, not senders.
3800 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3801 the ratelimit ACL was added.
3803 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3805 PH/46 Added extra information about PostgreSQL errors to the error string.
3807 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3808 avoiding the need to install special zones in a real server. This is
3809 backwards compatible; if it can't find the fake resolver, it drops back.
3810 Thus, both old and new test suites can be run.
3812 TF/02 Added util/ratelimit.pl
3814 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3815 clock is set back in time.
3817 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3818 Candler <B.Candler@pobox.com>.
3820 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3821 OK for addresses that are the result of redirections.
3823 PH/48 A number of further additions for the benefit of the new test suite,
3824 including a fake gethostbyname() that interfaces to the fake DNS resolver
3827 TF/06 The fix for widen_domains has also been applied to qualify_single and
3828 search_parents which are the other dnslookup options that can cause
3831 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3834 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3836 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3837 service" if it encountered an MX record with an empty target hostname.
3838 The message is now "an MX or SRV record indicated no SMTP service".
3840 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3841 if the DNS of the sending site is misconfigured. This is quite a
3842 common situation. This change restores the behaviour of treating a
3843 helo verification defer as a failure.
3845 PH/51 If self=fail was set on a router, the bounce message did not include the
3846 actual error message.
3852 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3854 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3855 command itself ended in error, the underlying message about the transport
3856 filter timeout was being overwritten with the pipe command error. Now the
3857 underlying error message should be appended to the second error message.
3859 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3861 PH/02 Reduce the amount of output that "make" produces by default. Full output
3862 can still be requested.
3864 PH/03 The warning log line about a condition test deferring for a "warn" verb
3865 was being output only once per connection, rather than after each
3866 occurrence (because it was using the same function as for successful
3867 "warn" verbs). This seems wrong, so I have changed it.
3869 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3870 it should not have, which might have caused a crash in the right
3871 circumstances, but probably never did.
3873 PH/04 Installed a modified version of Tony Finch's patch to make submission
3874 mode fix the return path as well as the Sender: header line, and to
3875 add a /name= option so that you can make the user's friendly name appear
3878 TF/03 Added the control = fakedefer ACL modifier.
3880 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3881 Mark Lowes for thorough testing.
3883 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3885 TK/03 Merged latest SRS patch from Miles Wilton.
3887 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3888 with the definition in sysexits.h (which is #included earlier).
3889 Fortunately, Exim does not actually use EX_OK. The code used to try to
3890 preserve the sysexits.h value, by assuming that macro definitions were
3891 scanned for macro replacements. I have been disabused of this notion,
3892 so now the code just undefines EX_OK before #including unistd.h.
3894 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3895 in the smtp transport. When a block could not be written in a single
3896 write() function, the timeout was being re-applied to each part-write.
3897 This seems wrong - if the receiver was accepting one byte at a time it
3898 would take for ever. The timeout is now adjusted when this happens. It
3899 doesn't have to be particularly precise.
3901 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3902 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3904 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3905 without a domain if the check to <postmaster@domain> fails.
3907 SC/01 Eximstats: added -xls and the ability to specify output files
3908 (patch written by Frank Heydlauf).
3910 SC/02 Eximstats: use FileHandles for outputting results.
3912 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3914 SC/04 Eximstats: fixed display of large numbers with -nvr option
3916 SC/05 Eximstats: fixed merging of reports with empty tables.
3918 SC/06 Eximstats: added the -include_original_destination flag
3920 SC/07 Eximstats: removed tabs and trailing whitespace.
3922 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3924 TK/06 MBOX spool code: Add real "From " MBOX separator line
3925 so the .eml file is really in mbox format (even though
3926 most programs do not really care). Patch from Alex Miller.
3928 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3929 The latter is generated from $received_to and is only set if the
3930 message has one envelope recipient. SA can use these headers,
3931 obviously out-of-the-box. Patch from Alex Miller.
3933 PH/08 The ${def test on a variable was returning false if the variable's
3934 value was "0", contrary to what the specification has always said!
3935 The result should be true unless the variable is empty.
3937 PH/09 The syntax error of a character other than { following "${if
3938 def:variable_name" (after optional whitespace) was not being diagnosed.
3939 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3940 accidental colon was present, for example, could give incorrect results.
3942 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3943 result is used (not including appendfile, where other changes are about
3946 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3947 This involved changing a lot of size variables from int to off_t. It
3948 should work with maildirs and everything.
3950 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3951 spamd dying while we are connected to it.
3953 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3956 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3957 function for BDB 4.3.
3959 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3961 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3962 the sources for PCRE have been re-arranged and more files are now
3965 PH/15 The code I had for printing potentially long long variables in PH/11
3966 above was not the best (it lost precision). The length of off_t variables
3967 is now inspected at build time, and an appropriate printing format (%ld
3968 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3969 to be "long long int" or "long int". This is needed for the internal
3970 formatting function string_vformat().
3972 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3973 the configuration file to be ":syslog", then the script "guesses" where
3974 the logs files are, rather than using the compiled in default. In our
3975 case the guess is not the same as the compiled default, so the script
3976 suddenly stopped working when I started to use syslog. The patch checks
3977 to see if log_file_path is "". If so, it attempts to read it from exim
3978 with no configuration file to get the compiled in version, before it
3979 falls back to the previous guessing code."
3981 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3982 implementing BATV in an Exim configuration. See NewStuff for the gory
3985 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3986 Makefile that are specific to HP-UX.
3988 PH/18 If the "use_postmaster" option was set for a recipient callout together
3989 with the "random" option, the postmaster address was used as the MAIL
3990 FROM address for the random test, but not for the subsequent recipient
3991 test. It is now used for both.
3993 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3994 patch removes a few documentation additions to RFC 3028, because the
3995 latest draft now contains them. It adds the new en;ascii-case comparator
3996 and a new error check for 8bit text in MIME parts. Comparator and
3997 require names are now matched exactly. I enabled the subaddress
3998 extension, but it is not well tested yet (read: it works for me)."
4000 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4001 rework some of the code of TK/09 above to avoid the hardwired use of
4002 "%lld" and "long long". Replaced the call to snprintf() with a call to
4005 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4006 records point to non-existent hosts", "retry timeout exceeded", and
4007 "retry time not reached for any host after a long failure period".
4009 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4010 experimental DomainKeys support:
4012 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4013 (2) On an error such as an illegally used "control", the wrong name for
4014 the control was given.
4016 These problems did NOT occur unless DomainKeys support was compiled.
4018 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4020 PH/24 Added ${if match_ip condition.
4022 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4023 databases so that it will be absolutely obvious if a crash occurs in the
4024 DB library. This is a regular occurrence (often caused by mis-matched
4027 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4028 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4029 picked up on a user's system that detects such things. There doesn't seem
4030 to be a gcc warning option for this - only an attribute that has to be
4031 put on the function's prototype. It seems that in Fedora Core 4 they have
4032 set this on a number of new functions. No doubt there will be more in due
4035 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4036 specify a transport. However, if an address that was verified by such a
4037 router was the subject of a callout, Exim crashed because it tried to
4038 read the rcpt_include_affixes from the non-existent transport. Now it
4039 just assumes that the setting of that option is false. This bug was
4040 introduced by 4.51/PH/31.
4042 PH/28 Changed -d+all to exclude +memory, because that information is very
4043 rarely of interest, but it makes the output a lot bigger. People tend to
4044 do -d+all out of habit.
4046 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4047 code in os-type was giving problems when libc.so lives in lib64, like on
4050 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4051 aren't the modern standard, and it seems that some systems' include files
4052 don't always have them. Exim was already checking for some of the newer
4053 ones like T_AAAA, and defining it itself. I've added checks for all the
4054 record types that Exim uses.
4056 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4057 not automatically generating a new one, as it is supposed to. This
4058 prevented TLS from working. If the file did exist, but contained invalid
4059 data, a new version was generated, as expected. It was only the case of a
4060 non-existent file that was broken.
4062 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4063 with a change in libdomainkeys > 0.64.
4065 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4066 from DNS. If the selector record carries the flag, it now has
4067 precedence over the domain-wide flag.
4069 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4071 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4072 the use of a port name that isn't defined in /etc/services) occurred, the
4073 message was deferred as in a normal delivery, and thus remained on the
4074 spool, instead of being failed because of the mua_wrapper setting. This
4075 is now fixed, and I tidied up some of the mua_wrapper messages at the
4078 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4079 the messages in a hash of arrays rather than using individual hashes.
4080 This is a bit cleaner and results in dramatic memory savings, albeit
4081 at a slight CPU cost.
4083 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4084 as requested by Marc Sherman.
4086 SC/10 Eximstats: added histograms for user specified patterns as requested
4089 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4091 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4092 fopen() in the content-scanning modules that did not already have it.
4098 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4099 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4101 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4103 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4105 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4106 to test Sieve filters that use "vacation".
4108 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4109 that changes the way the GnuTLS parameters are stored in the cache file.
4110 The new format can be generated externally. For backward compatibility,
4111 if the data in the cache doesn't make sense, Exim assumes it has read an
4112 old-format file, and it generates new data and writes a new file. This
4113 means that you can't go back to an older release without removing the
4116 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4117 work if there are any delivery delays because "one_time" forces the
4118 parent to be marked "delivered", so its unseen clone is never tried
4119 again. For this reason, Exim now forbids the simultaneous setting of
4122 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4123 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4124 entry if you want to know the details. The fix, however, neglected to
4125 consider the case when local delivery batching is involved. The test for
4126 "previously delivered" was not happening when checking to see if an
4127 address could be batched with a previous (undelivered) one; under
4128 certain circumstances this could lead to multiple deliveries to the same
4131 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4132 in its include files, and this causes problems building Exim.
4134 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4135 header_syntax) but Exim was just ignoring anything given after a slash.
4136 In particular, this caused confusion with an attempt to use "verify =
4137 reverse_host_lookup/defer_ok". An error is now given when options are
4138 supplied for verify items that do not have them. (Maybe reverse_host_
4139 lookup should have a defer_ok option, but that's a different point.)
4141 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4142 defined by RFC 821) to 2048, because there were problems with some AUTH
4143 commands, and RFC 1869 says the size should be increased for extended
4144 SMTP commands that take arguments.
4146 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4149 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4150 "unknown" error; now it says that the functionality isn't in the binary.
4152 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4153 an address' error message when a string expansion fails (syntax or
4154 whatever). Otherwise the password may appear in the log. Following change
4155 PH/42 below, there is no longer a chance of it appearing in a bounce
4158 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4160 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4161 of its addresses was ever tried. (Bugzilla bug #2.)
4163 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4164 the result incorrectly in the debug output. (It correctly added a newline
4165 to what was transported.)
4167 TF/01 Added $received_time.
4169 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4170 commented out examples of how to interface to a virus scanner and to
4171 SpamAssassin. Also added commented examples of av_scanner and
4172 spamd_address settings.
4174 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4175 and controls are allowed in which ACLs. There were a couple of minor
4176 errors. Some of the entries in the conditions table (which is a table of
4177 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4178 negation of where the condition IS allowed.
4180 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4182 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4183 header file does not have a version number, so I've had to invent a new
4184 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4185 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4186 radiusclient), but it was contributed by a Radius user.
4188 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4189 files or directories, for OpenSSL.
4191 PH/19 When an Exim process that is running as root has to create an Exim log
4192 file, it does so in a subprocess that runs as exim:exim so as to get the
4193 ownership right at creation (otherwise, other Exim processes might see
4194 the file with the wrong ownership). There was no test for failure of this
4195 fork() call, which would lead to the process getting stuck as it waited
4196 for a non-existent subprocess. Forks do occasionally fail when resources
4197 run out. I reviewed all the other calls to fork(); they all seem to check
4200 PH/20 When checking for unexpected SMTP input at connect time (before writing
4201 the banner), Exim was not dealing correctly with a non-positive return
4202 from the read() function. If the client had disconnected by this time,
4203 the result was a log entry for a synchronization error with an empty
4204 string after "input=" when read() returned zero. If read() returned -1
4205 (an event I could not check), uninitialized data bytes were printed.
4206 There were reports of junk text (parts of files, etc) appearing after
4209 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4211 PH/22 Added support for macro redefinition, and (re)definition in between
4212 driver and ACL definitions.
4214 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4215 forgetting to use the resulting value; it was using the unexpanded value.
4217 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4218 hadn't been configured. The fix is from Juergen Kreileder, who
4219 understands it better than I do:
4221 "Here's what I see happening with three configured cyrus_sasl
4222 authenticators configured (plain, login, cram-md5):
4224 On startup auth_cyrus_sasl_init() gets called for each of these.
4225 This means three calls to sasl_listmech() without a specified mech_list.
4226 => SASL tests which mechs of all available mechs actually work
4227 => three warnings about OTP not working
4228 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4230 With the patch, sasl_listmech() also gets called three times. But now
4231 SASL's mech_list option is set to the server_mech specified in the the
4232 authenticator. Or in other words, the answer from sasl_listmech()
4233 gets limited to just the mech you're testing for (which is different
4235 => the return list contains just 'plain' or 'login', 'cram-md5' or
4236 nothing depending on the value of ob->server_mech.
4238 I've just tested the patch: Authentication still works fine,
4239 unavailable mechs specified in the exim configuration are still
4240 caught, and the auth.log warnings about OTP are gone."
4242 PH/25 When debugging is enabled, the contents of the command line are added
4243 to the debugging output, even when log_selector=+arguments is not
4246 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4247 answer is "GNU", and only if the return is "GNU/something" is the answer
4250 PH/27 $acl_verify_message is now set immediately after the failure of a
4251 verification in an ACL, and so is available in subsequent modifiers. In
4252 particular, the message can be preserved by coding like this:
4254 warn !verify = sender
4255 set acl_m0 = $acl_verify_message
4257 Previously, $acl_verify_message was set only while expanding "message"
4258 and "log_message" when a very denied access.
4260 PH/28 Modified OS/os.c-Linux with
4262 -#ifndef OS_LOAD_AVERAGE
4263 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4265 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4266 nomenclature these days.)
4268 PH/29 Installed patch from the Sieve maintainer that adds the options
4269 sieve_useraddress and sieve_subaddress to the redirect router.
4271 PH/30 In these circumstances:
4272 . Two addresses routed to the same list of hosts;
4273 . First host does not offer TLS;
4274 . First host accepts first address;
4275 . First host gives temporary error to second address;
4276 . Second host offers TLS and a TLS session is established;
4277 . Second host accepts second address.
4278 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4279 and peerdn, if requested) that were in fact used only for the second
4282 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4283 attention to any local part prefix or suffix that was matched by the
4284 router that accepted the address. It now behaves in the same way as it
4285 does for delivery: the affixes are removed from the local part unless
4286 rcpt_include_affixes is set on the transport.
4288 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4289 timeout during the DATA phase of an incoming message.
4291 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4292 applied a patch sanctioned by the Sieve maintainer.
4294 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4295 the uid or gid is negative. A case of a negative gid caused this to be
4296 noticed. The fix allows for either to be negative.
4298 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4299 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4302 PH/36 The change PH/12 above was broken. Fixed it.
4304 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4305 the grounds that routing the same address twice would always produce the
4306 same answer. This might have been true once, but it is certainly no
4307 longer true now. Routing a child address may depend on the previous
4308 routing that produced that child. Some complicated redirection strategies
4309 went wrong when messages had multiple recipients, and made Exim's
4310 behaviour dependent on the order in which the addresses were given.
4312 I have moved the duplicate checking until after the routing is complete.
4313 Exim scans the addresses that are assigned to local and remote
4314 transports, and removes any duplicates. This means that more work will be
4315 done, as duplicates will always all be routed, but duplicates are
4316 presumably rare, so I don't expect this is of any significance.
4318 For deliveries to pipes, files, and autoreplies, the duplicate checking
4319 still happens during the routing process, since they are not going to be
4322 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4323 It corrects a timeout issue with spamd. This is Ian's comment: "The
4324 background is that sometimes spamd either never reads data from a
4325 connection it has accepted, or it never writes response data. The exiscan
4326 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4327 blindly assumes that writes won't block so it may never time out."
4329 PH/39 Allow G after quota size as well as K and M.
4331 PH/40 The value set for $authenticated_id in an authenticator may not contain
4332 binary zeroes or newlines because the value is written to log lines and
4333 to spool files. There was no check on this. Now the value is run through
4334 the string_printing() function so that such characters are converted to
4335 printable escape sequences.
4337 PH/41 $message_linecount is a new variable that contains the total number of
4338 lines in the message. Compare $body_linecount, which is the count for the
4341 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4342 bounce and delay warning messages, except in certain special cases, which
4345 (a) An SMTP error message from a remote host;
4346 (b) A message specified in a :fail: redirection;
4347 (c) A message specified in a "fail" command in a system filter;
4348 (d) A message specified in a FAIL return from the queryprogram router;
4349 (e) A message specified by the cannot_route_message router option.
4351 In these cases only, Exim does include the error details in bounce and
4352 warning messages. There are also a few cases where bland messages such
4353 as "unrouteable address" or "local delivery error" are given.
4355 PH/43 $value is now also set for the "else" part of a ${run expansion.
4357 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4358 being worked on, but at least Exim now implements the latest version to
4361 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4362 process to complete was treated as a delivery failure, a timeout while
4363 writing the message to the pipe was logged, but erroneously treated as a
4364 successful delivery. Such timeouts include transport filter timeouts. For
4365 consistency with the overall process timeout, these timeouts are now
4366 treated as errors, giving rise to delivery failures by default. However,
4367 there is now a new Boolean option for the pipe transport called
4368 timeout_defer, which, if set TRUE, converts the failures into defers for
4369 both kinds of timeout. A transport filter timeout is now identified in
4372 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4373 systems where "make" and "gmake" are different, calling "gmake" at top
4374 level broke things. I've arranged for the value of $(MAKE) to be passed
4375 from the Makefile to this script so that it can call the same version of
4379 A note about Exim versions 4.44 and 4.50
4380 ----------------------------------------
4382 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4383 changes of various kinds. As a consequence, a big documentation update was
4384 needed. This delayed the release for rather longer than seemed good, especially
4385 in the light of a couple of (minor) security issues. Therefore, the changes
4386 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4387 release. So 4.44 and 4.50 are in effect two different branches that both start
4390 I have left the 4.50 change log unchanged; it contains all the changes since
4391 4.43. The change log for 4.44 is below; many of its items are identical to
4392 those for 4.50. This seems to be the most sensible way to preserve the
4393 historical information.
4399 1. Minor wording change to the doc/README.SIEVE file.
4401 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4402 computation of the current number of files was incorrect.
4404 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4405 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4408 4. Give more explanation in the error message when the command for a transport
4409 filter fails to execute.
4411 5. There are several places where Exim runs a non-Exim command in a
4412 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4413 was being done only for the command run by the queryprogram router. It is
4414 now done for all such subprocesses. The other cases are: ${run, transport
4415 filters, and the commands run by the lmtp and pipe transports.
4417 6. Added CONFIGURE_GROUP build-time option.
4419 7. Some older OS have a limit of 256 on the maximum number of file
4420 descriptors. Exim was using setrlimit() to set 1000 as a large value
4421 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4422 systems. I've change it so that if it can't get 1000, it tries for 256.
4424 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4425 was an oversight, and furthermore, ever since the addition of extra
4426 controls (e.g. 4.43/32), the checks on when to allow different forms of
4427 "control" were broken. There should now be diagnostics for all cases when a
4428 control that does not make sense is encountered.
4430 9. Added the /retain_sender option to "control=submission".
4432 10. $recipients is now available in the predata ACL (oversight).
4434 11. Tidy the search cache before the fork to do a delivery from a message
4435 received from the command line. Otherwise the child will trigger a lookup
4436 failure and thereby defer the delivery if it tries to use (for example) a
4437 cached ldap connection that the parent has called unbind on.
4439 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4440 of $address_data from the recipient verification was clobbered by the
4441 sender verification.
4443 13. The value of address_data from a sender verification is now available in
4444 $sender_address_data in subsequent conditions in the ACL statement.
4446 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4448 15. Added a new option "connect=<time>" to callout options, to set a different
4451 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4452 was its contents. (It was OK if the option was not defined at all.)
4454 17. A "Completed" log line is now written for messages that are removed from
4455 the spool by the -Mrm option.
4457 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4458 information about exactly what failed.
4460 19. Added -dd to debug only the daemon process.
4462 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4463 handles timeouts, both on the server side and network timeouts. Renamed the
4464 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4466 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4467 It is now set to "smtps".
4469 22. $host_address is now set to the target address during the checking of
4470 ignore_target_hosts.
4472 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4473 being passed; this would have caused $sender_host_name to have been used if
4474 matching the list had actually called for a host name (not very likely,
4475 since this list is usually IP addresses). A host name is now passed as
4478 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4479 code that specifies a non-restarting handler (typically sigaction() in
4480 modern systems) in an attempt to fix a rare and obscure crash bug.
4482 25. Narrowed the window for a race in the daemon that could cause it to ignore
4483 SIGCHLD signals. This is not a major problem, because they are used only to
4484 wake it up if nothing else does.
4486 26. A malformed maildirsize file could cause Exim to calculate negative values
4487 for the mailbox size or file count. Odd effects could occur as a result.
4488 The maildirsize information is now recalculated if the size or filecount
4491 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4492 support for a long time. Removed HAVE_SYS_VFS_H.
4494 28. Installed the latest version of exipick from John Jetmore.
4496 29. In an address list, if the pattern was not a regular expression, an empty
4497 subject address (from a bounce message) matched only if the pattern was an
4498 empty string. Non-empty patterns were not even tested. This was the wrong
4499 because it is perfectly reasonable to use an empty address as part of a
4500 database query. An empty address is now tested by patterns that are
4501 lookups. However, all the other forms of pattern expect the subject to
4502 contain a local part and a domain, and therefore, for them, an empty
4503 address still always fails if the pattern is not itself empty.
4505 30. Exim went into a mad DNS loop when attempting to do a callout where the
4506 host was specified on an smtp transport, and looking it up yielded more
4507 than one IP address.
4509 31. Re-factored the code for checking spool and log partition space into a
4510 function that finds that data and another that does the check. The former
4511 is then used to implement four new variables: $spool_space, $log_space,
4512 $spool_inodes, and $log_inodes.
4514 32. The RFC2047 encoding function was originally intended for short strings
4515 such as real names; it was not keeping to the 75-character limit for
4516 encoded words that the RFC imposes. It now respects the limit, and
4517 generates multiple encoded words if necessary. To be on the safe side, I
4518 have increased the buffer size for the ${rfc2047: expansion operator from
4521 33. It is now permitted to omit both strings after an "if" condition; if the
4522 condition is true, the result is "true". As before, when the second string
4523 is omitted, a false condition yields an empty string. This makes it less
4524 cumbersome to write custom ACL and router conditions.
4526 34. Failure to deliver a bounce message always caused it to be frozen, even if
4527 there was an errors_to setting on the router. The errors_to setting is now
4530 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4531 canonical form (fully expanded) before being placed in
4532 $sender_host_address.
4534 36. The table in the code that translates DNS record types into text (T_A to
4535 "A" for instance) was missing entries for NS and CNAME. It is just possible
4536 that this could have caused confusion if both these types were looked up
4537 for the same domain, because the text type is used as part of Exim's
4538 per-process caching. But the chance of anyone hitting this buglet seems
4541 37. The dnsdb lookup has been extended in a number of ways.
4543 (1) There is a new type, "zns", which walks up the domain tree until it
4544 finds some nameserver records. It should be used with care.
4546 (2) There is a new type, "mxh", which is like "mx" except that it returns
4547 just the host names, not the priorities.
4549 (3) It is now possible to give a list of domains (or IP addresses) to be
4550 looked up. The behaviour when one of the lookups defers can be
4551 controlled by a keyword.
4553 (4) It is now possible to specify the separator character for use when
4554 multiple records are returned.
4556 38. The dnslists ACL condition has been extended: it is now possible to supply
4557 a list of IP addresses and/or domains to be looked up in a particular DNS
4560 39. Added log_selector=+queue_time_overall.
4562 40. When running the queue in the test harness, wait just a tad after forking a
4563 delivery process, to get repeatability of debugging output.
4565 41. Include certificate and key file names in error message when GnuTLS fails
4566 to set them up, because the GnuTLS error message doesn't include the name
4567 of the failing file when there is a problem reading it.
4569 42. Allow both -bf and -bF in the same test run.
4571 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4573 44. The "Exiscan patch" is now merged into the mainline Exim source.
4575 45. Sometimes the final signoff response after QUIT could fail to get
4576 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4577 < 0 before doing a fflush(). This bug looks as though it goes back to the
4578 introduction of TLS in release 3.20, but "sometimes" must have been rare
4579 because the tests only now provoked it.
4581 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4582 (this can affect the format of dates).
4584 47. exim_tidydb, when checking for the continued existence of a message for
4585 which it has found a message-specific retry record, was not finding
4586 messages that were in split spool directories. Consequently, it was
4587 deleting retry records that should have stayed in existence.
4589 48. Steve fixed some bugs in eximstats.
4591 49. The SPA authentication driver was not abandoning authentication and moving
4592 on to the next authenticator when an expansion was forced to fail,
4593 contradicting the general specification for all authenticators. Instead it
4594 was generating a temporary error. It now behaves as specified.
4596 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4597 (the order specifies the preference for clients). The order is now AES256,
4598 AES128, 3DES, ARCFOUR128.
4600 51. Small patch to Sieve code - explicitly set From: when generating an
4603 52. Exim crashed if a remote delivery caused a very long error message to be
4604 recorded - for instance if somebody sent an entire SpamAssassin report back
4605 as a large number of 550 error lines. This bug was coincidentally fixed by
4606 increasing the size of one of Exim's internal buffers (big_buffer) that
4607 happened as part of the Exiscan merge. However, to be on the safe side, I
4608 have made the code more robust (and fixed the comments that describe what
4611 53. Now that there can be additional text after "Completed" in log lines (if
4612 the queue_time_overall log selector is set), a one-byte patch to exigrep
4613 was needed to allow it to recognize "Completed" as not the last thing in
4616 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4617 patch that reportedly fixes this has been added. I am not expert enough to
4618 create a test for it. This is what the patch creator wrote:
4620 "I found a little strange behaviour of ldap code when working with
4621 Windows 2003 AD Domain, where users was placed in more than one
4622 Organization Units. When I tried to give exim partial DN, the exit code
4623 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4624 But simultaneously result of request was absolutely normal ldap result,
4625 so I produce this patch..."
4627 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4628 REFERENCE, so I have modified the code to exclude the patch when that macro
4631 55. Some experimental protocols are using DNS PTR records for new purposes. The
4632 keys for these records are domain names, not reversed IP addresses. The
4633 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4634 leaves it alone. Component reversal etc. now happens only for IP addresses.
4637 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4639 57. Double the size of the debug message buffer (to 2048) so that more of very
4640 long debug lines gets shown.
4642 58. The exicyclog utility now does better if the number of log files to keep
4643 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4645 59. Two changes related to the smtp_active_hostname option:
4647 (1) $smtp_active_hostname is now available as a variable.
4648 (2) The default for smtp_banner uses $smtp_active_hostname instead
4649 of $primary_hostname.
4651 60. The host_aton() function is supposed to be passed a string that is known
4652 to be a valid IP address. However, in the case of IPv6 addresses, it was
4653 not checking this. This is a hostage to fortune. Exim now panics and dies
4654 if the condition is not met. A case was found where this could be provoked
4655 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4656 components; fortuitously, this particular loophole had already been fixed
4657 by change 4.50/55 above.
4659 If there are any other similar loopholes, the new check in host_aton()
4660 itself should stop them being exploited. The report I received stated that
4661 data on the command line could provoke the exploit when Exim was running as
4662 exim, but did not say which command line option was involved. All I could
4663 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4664 running as the user.
4667 61. There was a buffer overflow vulnerability in the SPA authentication code
4668 (which came originally from the Samba project). I have added a test to the
4669 spa_base64_to_bits() function which I hope fixes it.
4672 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4673 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4675 63. The daemon start-up calls getloadavg() while still root for those OS that
4676 need the first call to be done as root, but it missed one case: when
4677 deliver_queue_load_max is set with deliver_drop_privilege. This is
4678 necessary for the benefit of the queue runner, because there is no re-exec
4679 when deliver_drop_privilege is set.
4681 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4682 This has been fixed.
4684 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4685 was in use, was not putting the data itself into the right store pool;
4686 consequently, it could be overwritten for a subsequent message in the same
4687 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4690 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4692 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4693 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4694 the function were treating the return as a boolean value, which happened to
4695 work because 0=false and not-0=true, but is not correct code.
4697 68. The host_aton() function was not handling scoped IPv6 addresses (those
4698 with, for example, "%eth0" on the end) correctly.
4700 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4701 negated items (that is, ~something) in unsigned ints. Some compilers
4702 apparently mutter when there is no cast.
4704 70. If an address verification called from an ACL failed, and did not produce a
4705 user-specific message (i.e. there was only a "system" message), nothing was
4706 put in $acl_verify_message. In this situation, it now puts the system
4709 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4710 session; change 4.31/43 added the unwanted input to the log line - except
4711 that it did not do this in the start of session case. It now does.
4713 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4714 This is wrong and can cause the other end to generate a synchronization
4715 error if it is another Exim or anything else that does the synchronization
4716 check. A QUIT command is no longer sent after a timeout.
4718 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4719 during host lookups.
4721 74. The defer_ok option of callout verification was not working if it was used
4722 when verifying addresses in header lines, that is, for this case:
4724 verify = header_sender/callout=defer_ok
4726 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4727 those file descriptors could be used for SMTP connections. If anything
4728 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4729 could be sent to the SMTP client, causing chaos. The daemon now opens
4730 stdin, stdout, and stderr to /dev/null when it puts itself into the
4733 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4734 log by default. The user can override this with suitable Perl magic.
4736 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4737 the log message when discard triggers, was not working for the DATA ACL or
4738 for the non-SMTP ACL.
4740 78. Error message wording change in sieve.c.
4742 79. If smtp_accept_max_per_host was set, the number of connections could be
4743 restricted to fewer than expected, because the daemon was trying to set up
4744 a new connection before checking whether the processes handling previous
4745 connections had finished. The check for completed processes is now done
4746 earlier. On busy systems, this bug wouldn't be noticed because something
4747 else would have woken the daemon, and it would have reaped the completed
4750 80. If a message was submitted locally by a user whose login name contained one
4751 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4752 It caused a spool format error. I have fixed the spool reading code. A
4753 related problem was that the "from" clause in the Received: line became
4754 illegal because of the space(s). It is now covered by ${quote_local_part.
4756 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4759 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4761 83. Patch from Sieve maintainer to fix unterminated string problem in
4762 "vacation" handling.
4764 84. Some minor changes to the Linux configuration files to help with other
4765 OS variants using glibc.
4767 85. One more patch for Sieve to update vacation handling to latest spec.
4770 ----------------------------------------------------
4771 See the note above about the 4.44 and 4.50 releases.
4772 ----------------------------------------------------
4778 1. Change 4.43/35 introduced a bug that caused file counts to be
4779 incorrectly computed when quota_filecount was set in an appendfile
4782 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4783 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4786 3. Give more explanation in the error message when the command for a transport
4787 filter fails to execute.
4789 4. There are several places where Exim runs a non-Exim command in a
4790 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4791 was being done only for the command run by the queryprogram router. It is
4792 now done for all such subprocesses. The other cases are: ${run, transport
4793 filters, and the commands run by the lmtp and pipe transports.
4795 5. Some older OS have a limit of 256 on the maximum number of file
4796 descriptors. Exim was using setrlimit() to set 1000 as a large value
4797 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4798 systems. I've change it so that if it can't get 1000, it tries for 256.
4800 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4801 was an oversight, and furthermore, ever since the addition of extra
4802 controls (e.g. 4.43/32), the checks on when to allow different forms of
4803 "control" were broken. There should now be diagnostics for all cases when a
4804 control that does not make sense is encountered.
4806 7. $recipients is now available in the predata ACL (oversight).
4808 8. Tidy the search cache before the fork to do a delivery from a message
4809 received from the command line. Otherwise the child will trigger a lookup
4810 failure and thereby defer the delivery if it tries to use (for example) a
4811 cached ldap connection that the parent has called unbind on.
4813 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4814 of $address_data from the recipient verification was clobbered by the
4815 sender verification.
4817 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4818 was its contents. (It was OK if the option was not defined at all.)
4820 11. A "Completed" log line is now written for messages that are removed from
4821 the spool by the -Mrm option.
4823 12. $host_address is now set to the target address during the checking of
4824 ignore_target_hosts.
4826 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4827 being passed; this would have caused $sender_host_name to have been used if
4828 matching the list had actually called for a host name (not very likely,
4829 since this list is usually IP addresses). A host name is now passed as
4832 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4833 code that specifies a non-restarting handler (typically sigaction() in
4834 modern systems) in an attempt to fix a rare and obscure crash bug.
4836 15. Narrowed the window for a race in the daemon that could cause it to ignore
4837 SIGCHLD signals. This is not a major problem, because they are used only to
4838 wake it up if nothing else does.
4840 16. A malformed maildirsize file could cause Exim to calculate negative values
4841 for the mailbox size or file count. Odd effects could occur as a result.
4842 The maildirsize information is now recalculated if the size or filecount
4845 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4846 support for a long time. Removed HAVE_SYS_VFS_H.
4848 18. Updated exipick to current release from John Jetmore.
4850 19. Allow an empty sender to be matched against a lookup in an address list.
4851 Previously the only cases considered were a regular expression, or an
4854 20. Exim went into a mad DNS lookup loop when doing a callout where the
4855 host was specified on the transport, if the DNS lookup yielded more than
4858 21. The RFC2047 encoding function was originally intended for short strings
4859 such as real names; it was not keeping to the 75-character limit for
4860 encoded words that the RFC imposes. It now respects the limit, and
4861 generates multiple encoded words if necessary. To be on the safe side, I
4862 have increased the buffer size for the ${rfc2047: expansion operator from
4865 22. Failure to deliver a bounce message always caused it to be frozen, even if
4866 there was an errors_to setting on the router. The errors_to setting is now
4869 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4870 canonical form (fully expanded) before being placed in
4871 $sender_host_address.
4873 24. Updated eximstats to version 1.33
4875 25. Include certificate and key file names in error message when GnuTLS fails
4876 to set them up, because the GnuTLS error message doesn't include the name
4877 of the failing file when there is a problem reading it.
4879 26. Expand error message when OpenSSL has problems setting up cert/key files.
4882 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4883 (this can affect the format of dates).
4885 28. exim_tidydb, when checking for the continued existence of a message for
4886 which it has found a message-specific retry record, was not finding
4887 messages that were in split spool directories. Consequently, it was
4888 deleting retry records that should have stayed in existence.
4890 29. eximstats updated to version 1.35
4891 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4892 1.35 - bugfix such that pie charts by volume are generated correctly
4894 30. The SPA authentication driver was not abandoning authentication and moving
4895 on to the next authenticator when an expansion was forced to fail,
4896 contradicting the general specification for all authenticators. Instead it
4897 was generating a temporary error. It now behaves as specified.
4899 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4900 (the order specifies the preference for clients). The order is now AES256,
4901 AES128, 3DES, ARCFOUR128.
4903 31. Small patch to Sieve code - explicitly set From: when generating an
4906 32. Exim crashed if a remote delivery caused a very long error message to be
4907 recorded - for instance if somebody sent an entire SpamAssassin report back
4908 as a large number of 550 error lines. This bug was coincidentally fixed by
4909 increasing the size of one of Exim's internal buffers (big_buffer) that
4910 happened as part of the Exiscan merge. However, to be on the safe side, I
4911 have made the code more robust (and fixed the comments that describe what
4914 33. Some experimental protocols are using DNS PTR records for new purposes. The
4915 keys for these records are domain names, not reversed IP addresses. The
4916 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4917 leaves it alone. Component reversal etc. now happens only for IP addresses.
4920 34. The host_aton() function is supposed to be passed a string that is known
4921 to be a valid IP address. However, in the case of IPv6 addresses, it was
4922 not checking this. This is a hostage to fortune. Exim now panics and dies
4923 if the condition is not met. A case was found where this could be provoked
4924 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4925 components; fortuitously, this particular loophole had already been fixed
4926 by change 4.50/55 or 4.44/33 above.
4928 If there are any other similar loopholes, the new check in host_aton()
4929 itself should stop them being exploited. The report I received stated that
4930 data on the command line could provoke the exploit when Exim was running as
4931 exim, but did not say which command line option was involved. All I could
4932 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4933 running as the user.
4936 35. There was a buffer overflow vulnerability in the SPA authentication code
4937 (which came originally from the Samba project). I have added a test to the
4938 spa_base64_to_bits() function which I hope fixes it.
4941 36. The daemon start-up calls getloadavg() while still root for those OS that
4942 need the first call to be done as root, but it missed one case: when
4943 deliver_queue_load_max is set with deliver_drop_privilege. This is
4944 necessary for the benefit of the queue runner, because there is no re-exec
4945 when deliver_drop_privilege is set.
4947 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4948 was in use, was not putting the data itself into the right store pool;
4949 consequently, it could be overwritten for a subsequent message in the same
4950 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4953 38. Sometimes the final signoff response after QUIT could fail to get
4954 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4955 < 0 before doing a fflush(). This bug looks as though it goes back to the
4956 introduction of TLS in release 3.20, but "sometimes" must have been rare
4957 because the tests only now provoked it.
4963 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4964 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4965 Now it returns an integer. A number of calls were still expecting a T/F
4966 return. Fortuitously, in all cases, the tests worked in OK situations,
4967 which is the norm. However, things would have gone wrong on any write
4968 failures on the smtp file descriptor. This function is used when sending
4969 messages over SMTP and also when doing verify callouts.
4971 2. When Exim is called to do synchronous delivery of a locally submitted
4972 message (the -odf or -odi options), it no longer closes stderr before doing
4975 3. Implemented the mua_wrapper option.
4977 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4979 5. Implemented the functions header_remove(), header_testname(),
4980 header_add_at_position(), and receive_remove_recipient(), and exported them
4983 6. If an ACL "warn" statement specified the addition of headers, Exim already
4984 inserted X-ACL-Warn: at the start if there was no header name. However, it
4985 was not making this test for the second and subsequent header lines if
4986 there were newlines in the string. This meant that an invalid header could
4987 be inserted if Exim was badly configured.
4989 7. Allow an ACL "warn" statement to add header lines at the start or after all
4990 the Received: headers, as well as at the end.
4992 8. Added the rcpt_4xx retry error code.
4994 9. Added postmaster_mailfrom=xxx to callout verification option.
4996 10. Added mailfrom=xxxx to the callout verification option, for verify=
4999 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5000 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5002 12. Inserted some casts to stop certain compilers warning when using pointer
5003 differences as field lengths or precisions in printf-type calls (mostly
5004 affecting debugging statements).
5006 13. Added optional readline() support for -be (dynamically loaded).
5008 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5009 same clock tick as a message's arrival, so that its received time was the
5010 same as the "first fail" time on the retry record, and that message
5011 remained on the queue past the ultimate address timeout, every queue runner
5012 would try a delivery (because it was past the ultimate address timeout) but
5013 after another failure, the ultimate address timeout, which should have then
5014 bounced the address, did not kick in. This was a "< instead of <=" error;
5015 in most cases the first failure would have been in the next clock tick
5016 after the received time, and all would be well.
5018 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5019 being recognized when the domain list was tested by the match_domain
5020 condition in an expansion string.
5022 16. Added the ${str2b64: operator.
5024 17. Exim was always calling setrlimit() to set a large limit for the number of
5025 processes, without checking whether the existing limit was already
5026 adequate. (It did check for the limit on file descriptors.) Furthermore,
5027 errors from getrlimit() and setrlimit() were being ignored. Now they are
5028 logged to the main and panic logs, but Exim does carry on, to try to do its
5029 job under whatever limits there are.
5031 18. Imported PCRE 5.0.
5033 19. Trivial typo in log message " temporarily refused connection" (the leading
5036 20. If the log selector return_path_on_delivery was set and an address was
5037 redirected to /dev/null, the delivery process crashed because it assumed
5038 that a return path would always be set for a "successful" delivery. In this
5039 case, the whole delivery is bypassed as an optimization, and therefore no
5042 21. Internal re-arrangement: the function for sending a challenge and reading
5043 a response while authentication was assuming a zero-terminated challenge
5044 string. It's now changed to take a pointer and a length, to allow for
5045 binary data in such strings.
5047 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5049 23. Exim was not respecting finduser_retries when seeking the login of the
5050 uid under which it was called; it was always trying 10 times. (The default
5051 setting of finduser_retries is zero.) Also, it was sleeping after the final
5052 failure, which is pointless.
5054 24. Implemented tls_on_connect_ports.
5056 25. Implemented acl_smtp_predata.
5058 26. If the domain in control=submission is set empty, Exim assumes that the
5059 authenticated id is a complete email address when it generates From: or
5060 Sender: header lines.
5062 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5063 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5064 chown and chgrp in /bin and hostname in /usr/bin.
5066 28. Exim was keeping the "process log" file open after each use, just as it
5067 does for the main log. This opens the possibility of it remaining open for
5068 long periods when the USR1 signal hits a daemon. Occasional processlog
5069 errors were reported, that could have been caused by this. Anyway, it seems
5070 much more sensible not to leave this file open at all, so that is what now
5073 29. The long-running daemon process does not normally write to the log once it
5074 has entered its main loop, and it closes the log before doing so. This is
5075 so that log files can straightforwardly be renamed and moved. However,
5076 there are a couple of unusual error situations where the daemon does write
5077 log entries, and I had neglected to close the log afterwards.
5079 30. The text of an SMTP error response that was received during a remote
5080 delivery was being truncated at 512 bytes. This is too short for some of
5081 the long messages that one sometimes sees. I've increased the limit to
5084 31. It is now possible to make retry rules that apply only when a message has a
5085 specific sender, in particular, an empty sender.
5087 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5088 it possible to be selective about when SMTP synchronization is enforced.
5090 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5092 32. Implemented hosts_connection_nolog.
5094 33. Added an ACL for QUIT.
5096 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5099 35. Added mailbox_size and mailbox_filecount to appendfile.
5101 36. Added control = no_multiline_responses to ACLs.
5103 37. There was a bug in the logic of the code that waits for the clock to tick
5104 in the case where the clock went backwards by a substantial amount such
5105 that the microsecond fraction of "now" was more than the microsecond
5106 fraction of "then" (but the whole seconds number was less).
5108 38. Added support for the libradius Radius client library this is found on
5109 FreeBSD (previously only the radiusclient library was supported).
5115 1. When certain lookups returned multiple values in the form name=value, the
5116 quoting of the values was not always being done properly. Specifically:
5117 (a) If the value started with a double quote, but contained no whitespace,
5119 (b) If the value contained whitespace other than a space character (i.e.
5120 tabs or newlines or carriage returns) it was not quoted.
5121 This fix has been applied to the mysql and pgsql lookups by writing a
5122 separate quoting function and calling it from the lookup code. The fix
5123 should probably also be applied to nisplus, ibase and oracle lookups, but
5124 since I cannot test any of those, I have not disturbed their existing code.
5126 2. A hit in the callout cache for a specific address caused a log line with no
5127 reason for rejecting RCPT. Now it says "Previous (cached) callout
5128 verification failure".
5130 3. There was an off-by-one bug in the queryprogram router. An over-long
5131 return line was truncated at 256 instead of 255 characters, thereby
5132 overflowing its buffer with the terminating zero. As well as fixing this, I
5133 have increased the buffer size to 1024 (and made a note to document this).
5135 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5136 when Exim is waiting for an SMTP response from a remote server, Exim
5137 restarts its select() call on the socket, thereby resetting its timeout.
5138 This is not a problem when such interrupts are rare. Somebody set up a cron
5139 job to run exiwhat every 2 minutes, which is less than the normal select()
5140 timeout (5 or 10 minutes). This meant that the select() timeout never
5141 kicked in because it was always reset. I have fixed this by comparing the
5142 time when an interrupt arrives with the time at the start of the first call
5143 to select(). If more time than the timeout has elapsed, the interrupt is
5144 treated as a timeout.
5146 5. Some internal re-factoring in preparation for the addition of Sieve
5147 extensions (by MH). In particular, the "personal" test is moved to a
5148 separate function, and given an option for scanning Cc: and Bcc: (which is
5149 not set for Exim filters).
5151 6. When Exim created an email address using the login of the caller as the
5152 local part (e.g. when creating a From: or Sender: header line), it was not
5153 quoting the local part when it contained special characters such as @.
5155 7. Installed new OpenBSD configuration files.
5157 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5158 try to make them clearer.
5160 9. Callout options, other than the timeout value, were being ignored when
5161 verifying sender addresses in header lines. For example, when using
5163 verify = header_sender/callout=no_cache
5165 the cache was (incorrectly) being used.
5167 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5168 only the Cygwin environment.
5170 11. When return_path_on_delivery was set as a log selector, if different remote
5171 addresses in the same message used different return paths and parallel
5172 remote delivery occurred, the wrong values would sometimes be logged.
5173 (Whenever a remote delivery process finished, the return path value from
5174 the most recently started remote delivery process was logged.)
5176 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5177 lines when AUTH and/or TLS are in use. This is the "received protocol"
5178 field. Exim used to use "asmtp" for authenticated SMTP, without any
5179 indication (in the protocol name) for TLS use. Now it follows the RFC and
5180 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5181 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5182 names appear in log lines as well as in Received: header lines.
5184 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5185 extensions, and comparison tests, and to fix some bugs.
5187 14. Changes to the "personal" filter test:
5189 (1) The test was buggy in that it was just doing the equivalent of
5190 "contains" tests on header lines. For example, if a user's address was
5191 anne@some.where, the "personal" test would incorrectly be true for
5193 To: susanne@some.where
5195 This test is now done by extracting each address from the header in turn,
5196 and checking the entire address. Other tests that are part of "personal"
5197 are now done using regular expressions (for example, to check local parts
5198 of addresses in From: header lines).
5200 (2) The list of non-personal local parts in From: addresses has been
5201 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5202 taken from the Sieve specification recommendations.
5204 (3) If the message contains any header line starting with "List-" it is
5205 treated as non-personal.
5207 (4) The test for "circular" in the Subject: header line has been removed
5208 because it now seems ill-conceived.
5210 15. Minor typos in src/EDITME comments corrected.
5212 16. Installed latest exipick from John Jetmore.
5214 17. If headers_add on a router specified a text string that was too long for
5215 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5216 of string_sprintf() is now avoided.
5218 18. $message_body_size was not set (it was always zero) when running the DATA
5219 ACL and the local_scan() function.
5221 19. For the "mail" command in an Exim filter, no default was being set for
5222 the once_repeat time, causing a random time value to be used if "once" was
5223 specified. (If the value happened to be <= 0, no repeat happened.) The
5224 default is now 0s, meaning "never repeat". The "vacation" command was OK
5225 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5226 (I found it when inspecting the code).
5228 20. There is now an overall timeout for performing a callout verification. It
5229 defaults to 4 times the callout timeout, which applies to individual SMTP
5230 commands during the callout. The overall timeout applies when there is more
5231 than one host that can be tried. The timeout is checked before trying the
5232 next host. This prevents very long delays if there are a large number of
5233 hosts and all are timing out (e.g. when the network connections are timing
5234 out). The value of the overall timeout can be changed by specifying an
5235 additional sub-option for "callout", called "maxwait". For example:
5237 verify = sender/callout=5s,maxwait=20s
5239 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5240 to the end before writing, but this should make it even safer).
5242 22. Exim was forgetting that it had advertised PIPELINING for the second and
5243 subsequent messages on an SMTP connection. It was also not resetting its
5244 memory on STARTTLS and an internal HELO.
5246 23. When Exim logs an SMTP synchronization error within a session, it now
5247 records whether PIPELINING has been advertised or not.
5249 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5250 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5251 rather than long int.
5253 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5255 26. Added the never_mail option to autoreply.
5261 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5262 crash if the getsockname() call failed; this can happen if a connection is
5263 closed very soon after it is established. The problem was simply in the
5264 order in which certain operations were done, causing Exim to try to write
5265 to the SMTP stream before it had set up the file descriptor. The bug has
5266 been fixed by making things happen in the correct order.
5272 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5273 before the connection was closed, thus losing the rejection response.
5275 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5276 some early Solaris releases, but causes trouble in current releases where
5277 socklen_t is defined.
5279 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5282 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5285 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5286 that releasing the top of it at the end releases what was used for sub-
5287 expansions (unless the block got too big). However, discard this block if
5288 the first thing is a variable or header, so that we can use its block when
5289 it is dynamic (useful for very large $message_headers, for example).
5291 6. Lookups now cache *every* query, not just the most recent. A new, separate
5292 store pool is used for this. It can be recovered when all lookup caches are
5293 flushed. Lookups now release memory at the end of their result strings.
5294 This has involved some general refactoring of the lookup sources.
5296 7. Some code has been added to the store_xxx() functions to reduce the amount
5297 of flapping under certain conditions.
5299 8. log_incoming_interface used to affect only the <= reception log lines. Now
5300 it causes the local interface and port to be added to several more SMTP log
5301 lines, for example "SMTP connection from", and rejection lines.
5303 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5305 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5307 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5308 could be overwritten at the end of the current message (or the start of a
5309 new message if it was set in a HELO ACL). The value is now preserved for
5310 the duration of the SMTP connection.
5312 12. If a transport had a headers_rewrite setting, and a matching header line
5313 contained an unqualified address, that address was qualified, even if it
5314 did not match any rewriting rules. The underlying bug was that the values
5315 of the flags that permit the existence of unqualified sender and recipient
5316 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5317 non-local messages, and by -bnq for local messages) were not being
5318 preserved with the message after it was received.
5320 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5321 "next input=" as part of the text comprising the host identity instead of
5322 the correct text. The code was using the same buffer for two different
5323 strings. However, depending on which order the printing function evaluated
5324 its arguments, the bug did not always show up. Under Linux, for example, my
5325 test suite worked just fine.
5327 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5328 doesn't work with some older versions of Perl. It has been changed to "my",
5329 which in any case is probably the better facility to use.
5331 15. A really picky compiler found some instances of statements for creating
5332 error messages that either had too many or two few arguments for the format
5335 16. The size of the buffer for calls to the DNS resolver has been increased
5336 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5337 for addresses that have a lot of PTR records. This alleviates a problem; it
5338 does not fully solve it.
5340 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5341 buffer now truncates the list and logs the incident, which is the same
5342 action as happens when Exim is looking up a host name and its aliases.
5343 Previously in this situation something unpredictable would happen;
5344 sometimes it was "internal error: store_reset failed".
5346 18. If a server dropped the connection unexpectedly when an Exim client was
5347 using GnuTLS and trying to read a response, the client delivery process
5348 crashed while trying to generate an error log message.
5350 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5351 string, for example:
5353 warn message = H1: something\nH2: something
5355 the text was added as a single header line from Exim's point of view
5356 though it ended up OK in the delivered message. However, searching for the
5357 second and subsequent header lines using $h_h2: did not work. This has been
5358 fixed. Similarly, if a system filter added multiple headers in this way,
5359 the routers could not see them.
5361 20. Expanded the error message when iplsearch is called with an invalid key to
5362 suggest using net-iplsearch in a host list.
5364 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5365 ACLs are no longer actually imposed (and a message to that effect is
5368 22. If a "gecos" field in a passwd entry contained escaped characters, in
5369 particular, if it contained a \" sequence, Exim got it wrong when building
5370 a From: or a Sender: header from that name. A second bug also caused
5371 incorrect handling when an unquoted " was present following a character
5372 that needed quoting.
5374 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5375 was not being matched caselessly.
5377 24. Arranged for all hyphens in the exim.8 source to be escaped with
5380 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5381 properly. Recipient callout cache records were still being keyed to include
5382 the sender, even when use_sender was set false. This led to far more
5383 callouts that were necessary. The sender is no longer included in the key
5384 when use_sender is false.
5386 26. Added "control = submission" modifier to ACLs.
5388 27. Added the ${base62d: operator to decode base 62 numbers.
5390 28. dnsdb lookups can now access SRV records.
5392 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5393 the configuration file.
5395 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5396 (-v) mode. This makes the output for a verbose queue run more intelligible.
5398 31. Added a use_postmaster feature to recipient callouts.
5400 32. Added the $body_zerocount variable, containing the number of binary zero
5401 bytes in the message body.
5403 33. The time of last modification of the "new" subdirectory is now used as the
5404 "mailbox time last read" when there is a quota error for a maildir
5407 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5409 35. Added +ignore_unknown as a special item in host lists.
5411 36. Code for decoding IPv6 addresses in host lists is now included, even if
5412 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5413 address was recognized as an IP address, but was then not correctly decoded
5414 into binary, causing unexpected and incorrect effects when compared with
5421 1. Very minor rewording of debugging text in manualroute to say "list of
5422 hosts" instead of "hostlist".
5424 2. If verify=header_syntax was set, and a header line with an unqualified
5425 address (no domain) and a large number of spaces between the end of the
5426 name and the colon was received, the reception process suffered a buffer
5427 overflow, and (when I tested it) crashed. This was caused by some obsolete
5428 code that should have been removed. The fix is to remove it!
5430 3. When running in the test harness, delay a bit after writing a bounce
5431 message to get a bit more predictability in the log output.
5433 4. Added a call to search_tidyup() just before forking a reception process. In
5434 theory, someone could use a lookup in the expansion of smtp_accept_max_
5435 per_host which, without the tidyup, could leave open a database connection.
5437 5. Added the variables $recipient_data and $sender_data which get set from a
5438 lookup success in an ACL "recipients" or "senders" condition, or a router
5439 "senders" option, similar to $domain_data and $local_part_data.
5441 6. Moved the writing of debug_print from before to after the "senders" test
5444 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5445 problems for message scanning, either using a data ACL, or using
5446 local_scan() because the Received: header was not generated till after they
5447 were called (in order to set the time as the time of reception completion).
5448 I have revised the way this works. The header is now generated after the
5449 body is received, but before the ACL or local_scan() are called. After they
5450 are run, the timestamp in the header is updated.
5456 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5457 before starting a queue runner without re-exec. This happened only when
5458 deliver_drop_privilege was set or when the Exim user was set to root. The
5459 effect of the bug was that timeouts during subsequent deliveries caused
5460 crashes instead of being properly handled. The handler is now left at its
5461 default (and expected) setting.
5463 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5464 message, again when deliver_drop_privilege is set or Exim is run as root.
5465 The bug described in (1) was not present in this case, but the tidying up
5466 of the other signals was missing. I have made the two cases consistent.
5468 3. The ignore_target_hosts setting on a manualroute router was being ignored
5469 for hosts that were looked up using the /MX notation.
5471 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5474 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5475 operated on the sender address. After changing the $sender_address to <>
5476 for the sender address verify, Exim was re-instated it as the original
5477 (before rewriting) address, but remembering that it had rewritten it, so it
5478 wasn't rewriting it again. This bug also had the effect of breaking the
5479 sender address verification caching when the sender address was rewritten.
5481 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5482 This has been changed so that if the ip literal address matches
5483 ignore_target_hosts, the router declines.
5485 7. Added expansion conditions match_domain, match_address, and match_local_
5486 part (NOT match_host).
5488 8. The placeholder for the Received: header didn't have a length field set.
5490 9. Added code to Exim itself and to exim_lock to test for a specific race
5491 condition that could lead to file corruption when using MBX delivery. The
5492 issue is with the lockfile that is created in /tmp. If this file is removed
5493 after a process has opened it but before that process has acquired a lock,
5494 there is the potential for a second process to recreate the file and also
5495 acquire a lock. This could lead to two Exim processes writing to the file
5496 at the same time. The added code performs the same test as UW imapd; it
5497 checks after acquiring the lock that its file descriptor still refers to
5498 the same named file.
5500 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5501 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5504 11. Added the smtp_active_hostname option. If used, this will typically be made
5505 to depend on the incoming interface address. Because $interface_address is
5506 not set up until the daemon has forked a reception process, error responses
5507 that can happen earlier (such as "too many connections") no longer contain
5510 12. If an expansion in a condition on a "warn" statement fails because a lookup
5511 defers, the "warn" statement is abandoned, and the next ACL statement is
5512 processed. Previously this caused the whole ACL to be aborted.
5514 13. Added the iplsearch lookup type.
5516 14. Added ident_timeout as a log selector.
5518 15. Added tls_certificate_verified as a log selector.
5520 16. Added a global option tls_require_ciphers (compare the smtp transport
5521 option of the same name). This controls incoming TLS connections.
5523 17. I finally figured out how to make tls_require_ciphers do a similar thing
5524 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5525 before starting the TLS session.
5527 18. Tabs are now shown as \t in -bP output.
5529 19. If the log selector return_path_on_delivery was set, Exim crashed when
5530 bouncing a message because it had too many Received: header lines.
5532 20. If two routers both had headers_remove settings, and the first one included
5533 a superfluous trailing colon, the final name in the first list and the
5534 first name in the second list were incorrectly joined into one item (with a
5535 colon in the middle).
5541 1. Added -C and -D options to the exinext utility, mainly to make it easier
5542 to include in the automated testing, but these could be helpful when
5543 multiple configurations are in use.
5545 2. The exinext utility was not formatting the output nicely when there was
5546 an alternate port involved in the retry record key, nor when there was a
5547 message id as well (for retries that were specific to a specific message
5548 and a specific host). It was also confused by IPv6 addresses, because of
5549 the additional colons they contain. I have fixed the IPv4 problem, and
5550 patched it up to do a reasonable job for IPv6.
5552 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5553 delivery, the log line now contains "pipelined" if PIPELINING was used.
5555 4. An SMTP transport process used to panic and die if the bind() call to set
5556 an explicit outgoing interface failed. This has been changed; it is now
5557 treated in the same way as a connect() failure.
5559 5. A reference to $sender_host_name in the part of a conditional expansion
5560 that was being skipped was still causing a DNS lookup. This no longer
5563 6. The def: expansion condition was not recognizing references to header lines
5564 that used bh_ and bheader_.
5566 7. Added the _cache feature to named lists.
5568 8. The code for checking quota_filecount in the appendfile transport was
5569 allowing one more file than it should have been.
5571 9. For compatibility with Sendmail, the command line option
5579 and sets the incoming protocol and host name (for trusted callers). The
5580 host name and its colon can be omitted when only the protocol is to be set.
5581 Note the Exim already has two private options, -pd and -ps, that refer to
5582 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5583 "s", but I don't think that's a major issue.
5585 10. A number of refactoring changes to the code, none of which should affect
5588 (a) The number of logging options was getting close to filling up the
5589 32-bit word that was used as a bit map. I have split them into two classes:
5590 those that are passed in the argument to log_write(), and those that are
5591 only ever tested independently outside of that function. These are now in
5592 separate 32-bit words, so there is plenty of room for expansion again.
5593 There is no change in the user interface or the logging behaviour.
5595 (b) When building, for example, log lines, the code previously used a
5596 macro that called string_cat() twice, in order to add two strings. This is
5597 not really sufficiently general. Furthermore, there was one instance where
5598 it was actually wrong because one of the argument was used twice, and in
5599 one call a function was used. (As it happened, calling the function twice
5600 did not affect the overall behaviour.) The macro has been replaced by a
5601 function that can join an arbitrary number of extra strings onto a growing
5604 (c) The code for expansion conditions now uses a table and a binary chop
5605 instead of a serial search (which was left over from when there were very
5606 few conditions). Also, it now recognizes conditions like "pam" even when
5607 the relevant support is not compiled in: a suitably worded error message is
5608 given if an attempt is made to use such a condition.
5610 11. Added ${time_interval:xxxxx}.
5612 12. A bug was causing one of the ddress fields not to be passed back correctly
5613 from remote delivery subprocesses. The field in question was not being
5614 subsequently used, so this caused to problems in practice.
5616 13. Added new log selectors queue_time and deliver_time.
5618 14. Might have fixed a bug in maildirsizefile handling that threw up
5619 "unexpected character" debug warnings, and recalculated the data
5620 unnecessarily. In any case, I expanded the warning message to give more
5623 15. Added the message "Restricted characters in address" to the statements in
5624 the default ACL that block characters like @ and % in local parts.
5626 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5627 Three changes have been made:
5629 (a) There was a serious bug; a negative response to MAIL caused the whole
5630 recipient domain to be cached as invalid, thereby blocking all messages
5631 to all local parts at the same domain, from all senders. This bug has
5632 been fixed. The domain is no longer cached after a negative response to
5633 MAIL if the sender used is not empty.
5635 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5638 (c) A new callout option, "use_sender" has been added for people who want
5639 the modified behaviour.
5645 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5648 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5649 indeed breaks things for older releases.
5651 3. Added additional logging to the case where there is a problem reading data
5652 from a filter that is running in a subprocess using a pipe, in order to
5653 try to track down a specific problem.
5655 4. Testing facility fudge: when running in the test harness and attempting
5656 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5657 getting "No route to host". Convert this to a timeout.
5659 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5662 6. Some OS don't have socklen_t but use size_t instead. This affects the
5663 fifth argument of getsockopt() amongst other things. This is now
5664 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5665 can be set for individual OS. I have set it for SunOS5, OSF1, and
5666 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5667 some earlier ones do not.
5669 7. Change 4.30/15 was not doing the test caselessly.
5671 8. The standard form for an IPv6 address literal was being rejected by address
5672 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5673 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5674 this, as well as the form without the "IPv6" on the front (but only when
5675 address literals are enabled, of course).
5677 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5679 10. Exim crashed if a message with an empty sender address specified by -f
5680 encountered a router with an errors_to setting. This could be provoked only
5681 by a command such as
5685 where an empty string was supplied; "<>" did not hit this bug.
5687 11. Installed PCRE release 4.5.
5689 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5690 remained set. It is now erased.
5692 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5693 times from message ids (which are base 36 rather than the normal 62).
5695 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5696 were being counted as actual protocol errors, and logged if the log
5697 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5698 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5699 and DATA following a set of rejected RCPTs do not count as protocol errors.
5700 In other words, Exim assumes they were pipelined, though this may not
5701 actually be the case. Of course, in all cases the client gets an
5702 appropriate error code.
5704 15. If a lookup fails in an ACL condition, a message about the failure may
5705 be available; it is used if testing the ACL cannot continue, because most
5706 such messages specify what the cause of the deferral is. However, some
5707 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5708 that caused an old message to be retained and used if a later statement
5709 caused a defer, replacing the real cause of the deferral.
5711 16. If an IP address had so many PTR records that the DNS lookup buffer
5712 was not large enough to hold them, Exim could crash while trying to process
5713 the truncated data. It now detects and logs this case.
5715 17. Further to 4.21/58, another change has been made: if (and only if) the
5716 first line of a message (the first header line) ends with CRLF, a bare LF
5717 in a subsequent header line has a space inserted after it, so as not to
5718 terminate the header.
5720 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5721 unnecessarily, used atoi() instead of strtol(), and didn't check the
5722 termination when getting file sizes from file names by regex.
5724 19. Completely re-implemented the support for maildirsize files, in the light
5725 of a number of problems with the previous contributed implementation
5726 (4.30/29). In particular:
5728 . If the quota is zero, the maildirsize file is maintained, but no quota is
5731 . If the maildir directory does not exist, it is created before any attempt
5732 to write a maildirsize file.
5734 . The quota value in the file is just a cache; if the quota is changed in
5735 the transport, the new value overrides.
5737 . A regular expression is available for excluding directories from the
5740 20. The autoreply transport checks the characters in options that define the
5741 message's headers; it allows continued headers, but it was checking with
5742 isspace() after an embedded newline instead of explicitly looking for a
5745 21. If all the "regular" hosts to which an address was routed had passed their
5746 expiry times, and had not reached their retry times, the address was
5747 bounced, even if fallback hosts were defined. Now Exim should go on to try
5750 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5751 equivalent code in the SMTP transport. Some hosts send humungous responses
5752 to HELO/EHLO, more than 1024 it seems.
5754 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5755 gives compiler warnings in some environments. I've now done it "properly",
5758 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5759 (because of problems with the built-in one) was declared to return uschar *
5760 instead of char *, causing compiler failure.
5762 25. Fixed a file descriptor leak when processing alias/forward files.
5764 26. Fixed a minor format string issue in dbfn.c.
5766 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5768 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5769 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5772 29. When a sender address is verified, it is cached, to save repeating the test
5773 when there is more than one recipient in a message. However, when the
5774 verification involves a callout, it is possible for different callout
5775 options to be set for different recipients. It is too complicated to keep
5776 track of this in the cache, so now Exim always runs a verification when a
5777 callout is required, relying on the callout cache for the optimization.
5778 The overhead is duplication of the address routing, but this should not be
5781 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5782 to be verified with callout=postmaster, and the main callout worked but the
5783 postmaster check failed, the verification correctly failed. However, if a
5784 subsequent RCPT command asked for sender verification *without* the
5785 postmaster check, incorrect caching caused this verification also to fail,
5788 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5789 it was not caching the DNS options (qualify_single, search_parents) that
5790 were used when the lookup failed. A subsequent lookup with different
5791 options therefore always gave the same answer, though there were cases
5792 where it should not have. (Example: a "domains = !$mx_any" option on a
5793 dnslookup router: the "domains" option is always processed without any
5794 widening, but the router might have qualify_single set.) Now Exim uses the
5795 cached value only when the same options are set.
5797 32. Added John Jetmore's "exipick" utility to the distribution.
5799 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5800 than a timeout (e.g. a certificate is required, and is not provided), an
5801 Exim server now closes the connection immediately. Previously it waited for
5802 the client to close - but if the client is SSL, it seems that they each
5803 wait for each other, leading to a delay before one of them times out.
5805 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5806 maintained 0.8.x compatibility because I don't think many are using it, and
5807 it is clearly obsolete.
5809 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5812 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5813 client certificate was expired. A simple patch fixes this, though I don't
5814 understand the full logic of why the verify callback is called multiple
5817 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5818 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5819 which causes problems with some clients (such as the Certicom SSL Plus
5820 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5821 disables the coutermeasure allowing Eudora to connect."
5823 38. Exim was not checking that a write() to a log file succeeded. This could
5824 lead to Bad Things if a log got too big, in particular if it hit a file
5825 size limit. Exim now panics and dies if it cannot write to a log file, just
5826 as it does if it cannot open a log file.
5828 39. Modified OS/Makefile-Linux so that it now contains
5830 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5832 The two -D definitions ensure that Exim is compiled with large file
5833 support, which makes it possible to handle log files that are bigger than
5836 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5837 instance) a domain was checked against a named list that involved a lookup,
5838 causing $domain_data to be set, then another domain was checked against the
5839 same list, then the first domain was re-checked, the value of $domain_data
5840 after the final check could be wrong. In particular, if the second check
5841 failed, it could be set empty. This bug probably also applied to
5844 41. The strip_trailing_dot option was not being applied to the address given
5845 with the -f command-line option.
5847 42. The code for reading a message's header from the spool was incrementing
5848 $received_count, but never initializing it. This meant that the value was
5849 incorrect (doubled) while delivering a message in the same process in which
5850 it was received. In the most common configuration of Exim, this never
5851 happens - a fresh exec is done - but it can happen when
5852 deliver_drop_privilege is set.
5854 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5855 it now includes up to 150 characters of the unexpected data in the log
5858 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5859 and building data strings. The size of both of these buffers was 10 000
5860 bytes - far larger than anybody would *ever* want, thought I. Needless to
5861 say, somebody hit the limit. I have increased the maximum line length to
5862 20 000 and the maximum data length of concatenated lines to 100 000. I have
5863 also fixed two bugs, because there was no checking on these buffers. Tsk,
5864 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5865 buffer is too small.
5867 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5868 lsearch lookups. Now it does.
5870 46. When parsing a route_list item in a manualroute router, a fixed-length
5871 buffer was used for the list of hosts. I made this 1024 bytes long,
5872 thinking that nobody would ever have a list of hosts that long. Wrong.
5873 Somebody had a whole pile of complicated expansion conditions, and the
5874 string was silently truncated, leading to an expansion error. It turns out
5875 that it is easier to change to an unlimited length (owing to other changes
5876 that have happened since this code was originally written) than to build
5877 structure for giving a limitation error. The length of the item that
5878 expands into the list of hosts is now unlimited.
5880 47. The lsearch lookup could not handle data where the length of text line was
5881 more than 4095 characters. Such lines were truncated, leading to shortened
5882 data being returned. It should now handle lines of any length.
5884 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5885 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5888 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5889 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5890 no longer happens in comment lines. A list of replacements is now placed
5891 at the head of all of the source files, except those whose only change is
5892 to replace PERL_COMMAND in the very first #! line.
5894 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5895 messages on the queue, with a bottom-up merge sort, using code contributed
5896 by Michael Haardt. This should make operations like -bp somewhat faster on
5897 large queues. It won't affect queue runners, except when queue_run_in_order
5900 51. Installed eximstats 1.31 in the distribution.
5902 52. Added support for SRV lookups to the dnslookup router.
5904 53. If an ACL referred to $message_body or $message_body_end, the value was not
5905 reset for any messages that followed in the same SMTP session.
5907 54. The store-handling optimization for building very long strings was not
5908 differentiating between the different store pools. I don't think this
5909 actually made any difference in practice, but I've tidied it.
5911 55. While running the routers to verify a sender address, $sender_address
5912 was still set to the sender address. This is wrong, because when routing to
5913 send a bounce to the sender, it would be empty. Therefore, I have changed
5914 it so that, while verifying a sender address, $sender_address is set to <>.
5915 (There is no change to what happens when verifying a recipient address.)
5917 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5918 target A or AAAA records (if not already returned) without resetting the
5919 qualify_single or search_parents options of the DNS resolver. These are
5920 inappropriate in this case because the targets of MX and SRV records must
5921 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5922 target that, when qualified, matched something in the local domain. These
5923 two options are now turned off when doing these lookups.
5925 57. It seems that at least some releases of Reiserfs (which does not have the
5926 concept of a fixed number of inodes) returns zero and not -1 for the
5927 number of available inodes. This interacted badly with check_spool_inodes,
5928 which assumed that -1 was the "no such thing" setting. What I have done is
5929 to check that the total number of inodes is greater than zero before doing
5930 the test of how many are available.
5932 58. When a "warn" ACL statement has a log_message modifier, the message is
5933 remembered, and not repeated. This is to avoid a lot of repetition when a
5934 message has many recipients that cause the same warning to be written.
5935 However, Exim was preserving the list of already written lines for an
5936 entire SMTP session, which doesn't seem right. The memory is now reset if a
5937 new message is started.
5939 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5940 debugging output unless log_rewrite was also set.
5942 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5943 of (int)(handle) when we know that handle contains (void *)(-1).
5945 61. The Exim daemon panic-logs an error return when it closes the incoming
5946 connection. However "connection reset by peer" seems to be common, and
5947 isn't really an error worthy of noting specially, so that particular error
5950 62. When Exim is trying to find all the local interfaces, it used to panic and
5951 die if the ioctl to get the interface flags failed. However, it seems that
5952 on at least one OS (Solaris 9) it is possible to have an interface that is
5953 included in the list of interfaces, but for which you get a failure error
5954 for this call. This happens when the interface is not "plumbed" into a
5955 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5956 failure of the "get flags" call assumes that the interface is down.
5958 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5959 makes life easier for people who are doing arithmetic on fields extracted
5960 from dates, where you often get leading zeros that should not be
5961 interpreted as octal.
5963 64. Added qualify_domain to the redirect router, to override the global
5966 65. If a pathologically long header line contained very many addresses (the
5967 report of this problem mentioned 10 000) and each of them was rewritten,
5968 Exim could use up a very large amount of memory. (It kept on making new
5969 copies of the header line as it rewrote, and never released the old ones.)
5970 At the expense of a bit more processing, the header rewriting function has
5971 been changed so that it no longer eats memory in this way.
5973 66. The generation of the Received: header has been moved from the time that a
5974 message starts to be received, to the time that it finishes. The timestamp
5975 in the Received: header should now be very close to that of the <= log
5976 line. There are two side-effects of this change:
5978 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5979 logged header lines no longer include the local Received: line, because
5980 it has not yet been created. The same applies to a copy of the message
5981 that is returned to a non-SMTP sender when a message is rejected.
5983 (b) When a filter file is tested using -bf, no additional Received: header
5984 is added to the test message. After some thought, I decided that this
5987 This change does not affect the value of $received_for. It is still set
5988 after address rewriting, but before local_scan() is called.
5990 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5992 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5993 gave an unhelpful panic error message, and a defer error. I have managed to
5994 change this behaviour so that it now rejects any supplied certificate,
5995 which seems right, as the list of acceptable certificates is empty.
5997 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5998 gave an unhelpful defer error. I have not managed to make this reject any
5999 supplied certificates, but the error message it gives is "no certificate
6000 supplied", which is not helpful.
6002 70. exigrep's output now also includes lines that are not associated with any
6003 message, but which match the given pattern. Implemented by a patch from
6004 Martin Sluka, which also tidied up the Perl a bit.
6006 71. Recipient callout verification, like sender verification, was using <> in
6007 the MAIL FROM command. This isn't really the right thing, since the actual
6008 sender may affect whether the remote host accepts the recipient or not. I
6009 have changed it to use the actual sender in the callout; this means that
6010 the cache record is now keyed on a recipient/sender pair, not just the
6011 recipient address. There doesn't seem to be a real danger of callout loops,
6012 since a callout by the remote host to check the sender would use <>.
6013 [SEE ABOVE: changed after hitting problems.]
6015 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6016 temporary errors. However, in the case of such a code being given after
6017 the end of a data transmission (i.e. after ".") Exim was failing to write
6018 a retry record for the message. (Yes, there was some broken host that was
6019 actually sending 8xx at this point.)
6021 73. An unknown lookup type in a host list could cause Exim to panic-die when
6022 the list was checked. (An example that provoked this was putting <; in the
6023 middle of a list instead of at the start.) If this happened during a DATA
6024 ACL check, a -D file could be left lying around. This kind of configuration
6025 error no longer causes Exim to die; instead it causes a defer error. The
6026 incident is still logged to the main and panic logs.
6028 74. Buglet left over from Exim 3 conversion. The message "too many messages
6029 in one connection" was written to the rejectlog but not the mainlog, except
6030 when address rewriting (yes!) was being logged.
6032 75. Added write_rejectlog option.
6034 76. When a system filter was run not as root (that is, when system_filter_user
6035 was set), the values of the $n variables were not being returned to the
6036 main process; thus, they were not subsequently available in the $sn
6039 77. Added +return_path_on_delivery log selector.
6041 78. A connection timeout was being treated differently from recipients deferred
6042 when testing hosts_max_try with a message that was older than the host's
6043 retry timeout. (The host should not be counted, thus allowing all hosts to
6044 be tried at least once before bouncing.) This may have been the cause of an
6045 occasionally reported bug whereby a message would remain on the queue
6046 longer than the retry timeout, but would be bounced if a delivery was
6047 forced. I say "may" because I never totally pinned down the problem;
6048 setting up timeout/retry tests is difficult. See also the next item.
6050 79. The ultimate address timeout was not being applied to errors that involved
6051 a combination of host plus message (for example, a timeout on a MAIL
6052 command). When an address resolved to a number of possible hosts, and they
6053 were not all tried for each delivery (e.g. because of hosts_max_try), a
6054 message could remain on the queue longer than the retry timeout.
6056 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6059 81. Fixed an obscure SMTP outgoing bug which required at least the following
6060 conditions: (a) there was another message waiting for the same server;
6061 (b) the server returned 5xx to all RCPT commands in the first message so
6062 that the message was not completed; (c) the server dropped the connection
6063 or gave a negative response to the RSET that Exim sends to abort the
6064 transaction. The observed case was a dropped connection after DATA that had
6065 been sent in pipelining mode. That is, the server had advertised PIPELINING
6066 but was not implementing it correctly. The effect of the bug was incorrect
6067 behaviour, such as trying another host, and this could lead to a crash.
6073 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6074 and daemon.c were passed as pointers to ints; they should have been
6075 pointers to socklen_t variables (which are typically unsigned ints).
6077 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6080 3. Fixed a really odd bug that affected only the testing scheme; patching a
6081 certain fixed string in the binary changed the value of another string that
6082 happened to be identical to the end of the original first string.
6084 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6085 name", it returns that address as the IP address. On some operating
6086 systems (e.g. Solaris), it also passes back the IP address string as the
6087 "host name". However, on others (e.g. Linux), it passes back an empty
6088 string. Exim wasn't checking for this, and was changing the host name to an
6089 empty string, assuming it had been canonicalized.
6091 5. Although rare, it is permitted to have more than one PTR record for a given
6092 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6093 all the names associated with an address, because they do in Solaris.
6094 However, it seems that they do not in Linux for data that comes from the
6095 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6096 I found this out when I moved to a new Linux workstation and tried to run
6097 the Exim test suite.
6099 To get round this problem I have changed the code so that it now does its
6100 own call to the DNS to look up PTR records when searching for a host name.
6101 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6102 addresses that are only in /etc/hosts are still found.
6104 This behaviour is, however, controlled by an option called host_lookup_
6105 order, which defaults to "bydns:byaddr". If people want to use the other
6106 order, or indeed, just use one or the other means of lookup, they can
6107 specify it in this variable.
6109 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6110 some operating systems, this comes back from gethostbyaddr() as an empty
6111 string, and this is what Exim used to test for. However, it seems that in
6112 other systems, "." is yielded. Exim now tests for this case too.
6114 7. The values of check_spool_space and check_log_space are now held internally
6115 as a number of kilobytes instead of an absolute number of bytes. If a
6116 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6117 kilobyte. This means that much larger values can be stored.
6119 8. Exim monitor: an attempt to get the action menu when not actually pointing
6120 at a message produces an empty menu entitled "No message selected". This
6121 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6122 no entries in it ("Shell widget menu has zero width and/or height"). So I
6123 have added a single, blank menu entry in this case.
6125 9. Added ${quote_local_part.
6127 10. MIME decoding is now applied to the contents of Subject: header lines when
6130 11. Now that a reference to $sender_host_address automatically causes a reverse
6131 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6132 host lookup before query-style lookups in lists that might use this
6133 variable. This has therefore been abolished, and the "net-" prefix is no
6134 longer necessary for query-style lookups.
6136 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6137 have been a typo for LFLAGS, so it has been changed.
6139 13. The install script calls Exim with "-C /dev/null" in order to find the
6140 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6141 to be output. However, since Exim outputs its version number before the
6142 error, it didn't break the script. It just looked ugly. I fixed this by
6143 always allowing "-C /dev/null" if the caller is root.
6145 14. Ignore overlarge ACL variable number when reading spool file - insurance
6146 against a later release with more variables having written the file.
6148 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6149 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6150 this, as well as the form without the "IPv6" on the front.
6152 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6153 OS/Makefile-Darwin file.
6155 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6156 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6158 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6159 variable. However, some OS use ints for inodes. I've added cast to long int
6160 to get rid of the compiler warning.
6162 19. I had forgotten to lock out "/../" in configuration file names when
6163 ALT_CONFIG_PREFIX was set.
6165 20. Routers used for verification do not need to specify transports. However,
6166 if such a router generated a host list, and callout was configured, Exim
6167 crashed, because it could not find a port number from the (non-existent)
6168 transport. It now assumes port 25 in this circumstance.
6170 21. Added the -t option to exigrep.
6172 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6173 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6174 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6175 them is set, LOOKUP_LSEARCH is forced.
6177 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6179 24. Added sender and host information to the "rejected by local_scan()" log
6180 line; previously there was no indication of these.
6182 25. Added .include_if_exists.
6184 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6185 Linux. It turns out that not all file systems support this. Apparently some
6186 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6187 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6188 not supported on the file descriptor, is now ignored when Exim is trying to
6189 sync a directory. This applies only to Linux.
6191 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6193 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6194 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6197 29. Added support for maildirsize files from supplied patch (modified a bit).
6199 30. The use of :fail: followed by an empty string could lead Exim to respond to
6200 sender verification failures with (e.g.):
6202 550 Verification failed for <xxx>
6203 550 Sender verify failed
6205 where the first response line was missing the '-' that indicates it is not
6206 the final line of the response.
6208 31. The loop for finding the name of the user that called Exim had a hardwired
6209 limit of 10; it now uses the value of finduser_retries, which is used for
6210 all other user lookups.
6212 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6215 33. Exim was neglecting to zero errno before one call of strtol() when
6216 expanding a string and expecting an integer value. On some systems this
6217 resulted in spurious "integer overflow" errors. Also, it was casting the
6218 result into an int without checking.
6220 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6221 did not work. The code looks as if it has *never* worked, though it appears
6222 to have been documented since at least release 1.62. I have made it work.
6224 35. The "timeout_DNS" error in retry rules, also documented since at least
6225 1.62, also never worked. As it isn't clear exactly what this means, and
6226 clearly it isn't a major issue, I have abolished the feature by treating it
6227 as "timeout", and writing a warning to the main and panic logs.
6229 36. The display of retry rules for -brt wasn't always showing the error code
6232 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6233 timeout_connect_A, timeout_connect_MX.
6235 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6236 to the empty sender.
6238 39. The daemon was not analysing the content of -oX till after it had closed
6239 stderr and disconnected from the controlling terminal. This meant that any
6240 syntax errors were only noted on the panic log, and the return code from
6241 the command was 0. By re-arranging the code a little, I've made the
6242 decoding happen first, so such errors now appear on stderr, and the return
6243 code is 1. However, the actual setting up of the sockets still happens in
6244 the disconnected process, so errors there are still only recorded on the
6247 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6248 connections (as happens on some IP stacks) was logged at start up time as
6249 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6250 it from "IPv6 and IPv4", which means that two separate sockets are being
6253 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6254 says whether AF_INET or AF_INET6 was passed as an argument.
6256 42. Exiwhat output was messed up when time zones were included in log
6259 43. Exiwhat now gives more information about the daemon's listening ports,
6260 and whether -tls-on-connect was used.
6262 44. The "port" option of the smtp transport is now expanded.
6264 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6265 silently ignored. Now an error message is written to the main and panic
6268 46. There's a new ACL modifier called "logwrite" which writes to a log file
6269 as soon as it is encountered.
6271 47. Added $local_user_uid and $local_user_gid at routing time.
6273 48. Exim crashed when trying to verify a sender address that was being
6276 49. Exim was recognizing only a space character after ".include". It now also
6277 recognizes a tab character.
6279 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6280 extracting the relevant information from the specification. The man page no
6281 longer contains scrambled data for the -d option, and I've added a section
6282 at the front about calling Exim under different names.
6284 51. Added "extra_headers" argument to the "mail" command in filter files.
6286 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6289 53. Installed eximstats 1.29.
6291 54. Added transport_filter_timeout as a generic transport option.
6293 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6294 Cc: header lines. This was required by RFC 822, but it not required by RFC
6297 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6298 incoming messages that did not have them. Now it does so only if the
6299 message originates locally, that is, if there is no associated remote host
6300 address. When Resent- header lines are present, this applies to the Resent-
6301 lines rather than the non-Resent- lines.
6303 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6304 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6306 58. Messages for configuration errors now include the name of the main
6307 configuration file - useful now that there may be more than one file in a
6308 list (.included file names were always shown).
6310 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6311 for those rare installations that do not start the daemon as root or run it
6312 setuid root. I've cut out the call to initgroups() if the daemon is not
6315 60. The Exim user and group can now be bound into the binary as text strings
6316 that are looked up at the start of Exim's processing.
6318 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6320 62. Added $mailstore_basename variable.
6322 63. Installed patch to sieve.c from Michael Haardt.
6324 64. When Exim failed to open the panic log after failing to open the main log,
6325 the original message it was trying to log was written to stderr and debug
6326 output, but if they were not available (the usual case in production), it
6327 was lost. Now it is written to syslog before the two lines that record the
6328 failures to open the logs.
6330 65. Users' Exim filters run in subprocesses under the user's uid. It is
6331 possible for a "deliver" command or an alias in a "personal" command to
6332 provoke an address rewrite. If logging of address rewriting is configured,
6333 this fails because the process is not running as root or exim. There may be
6334 a better way of dealing with this, but for the moment (because 4.30 needs
6335 to be released), I have disabled address rewrite logging when running a
6336 filter in a non-root, non-exim process.
6342 1. The buildconfig auxiliary program wasn't quoting the value set for
6343 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6344 not defined. This bug was masked in 4.22 by the effect that was fixed in
6347 2. Some messages that were rejected after a message id was allocated were
6348 shown as "incomplete" by exigrep. It no longer does this for messages that
6349 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6351 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6352 have allow_domain_literals set, the ID did not get logged in the <= line.
6353 Domain literals are now always recognized in Message-ID: header lines.
6355 4. The first argument for a ${extract expansion item is the key name or field
6356 number. Leading and trailing spaces in this item were not being ignored,
6357 causing some misleading effects.
6359 5. When deliver_drop_privilege was set, single queue runner processes started
6360 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6361 same command in the process it spins off) were not dropping privilege.
6363 6. When the daemon running as "exim" started a queue runner, it always
6364 re-executed Exim in the spun-off process. This is a waste of effort when
6365 deliver_drop_privilege is set. The new process now just calls the
6366 queue-runner function directly.
6372 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6375 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6376 ignored. Though the use of -oP was forcing the writing of a pid file, it
6377 was always written to the default place.
6379 3. If the message "no IP address found for host xxxx" is generated during
6380 incoming verification, it is now followed by identification of the incoming
6381 connection (so you can more easily find what provoked it).
6383 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6385 5. Added some features to "harden" Exim a bit more against certain attacks:
6387 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6388 be put in Local/Makefile. This is like the never_users runtime option,
6389 but it cannot be overridden. The default setting is "root".
6391 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6392 prefix string with which any file named in a -C command line option
6395 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6396 is retained for -C and -D only if the caller of Exim is root. Without
6397 it, the exim user may also use -C and -D and retain privilege.
6399 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6400 command line option is disabled.
6402 6. Macro names set by the -D option must start with an upper case letter, just
6403 like macro names defined in the configuration file.
6405 7. Added "dereference=" facility to LDAP.
6407 8. Two instances of the typo "uknown" in the source files are fixed.
6409 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6410 the Configure-Makefile script screwed up while processing it.
6412 10. Incorporated PCRE 4.4.
6414 11. The SMTP synchronization check was not operating right at the start of an
6415 SMTP session. For example, it could not catch a HELO sent before the client
6416 waited for the greeting. There is now a check for outstanding input at the
6417 point when the greeting is written. Because of the duplex, asynchronous
6418 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6419 way, but not yet received, when the check is performed.
6421 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6422 on TCP/IP sockets, because this apparently causes some broken clients to
6425 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6426 unchanged) from the Cygwin maintainer.
6428 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6429 was supported instead of testing for mbx. Effectively a typo.
6431 15. The spa authenticator server code was not checking that the input it
6432 received was valid base64.
6434 16. The debug output line for the "set" modifier in ACLs was not showing the
6435 name of the variable that was being set.
6437 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6439 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6440 lookup on its own. Something else was needed to trigger the lookup. For
6441 example, a match in host_lookup or the need for a host name in a host list.
6442 Now, if $sender_host_name is referenced and the host name has not yet been
6443 looked up, a lookup is performed. If the lookup fails, the variable remains
6444 empty, and $host_lookup_failed is set to "1".
6446 19. Added "eqi" as a case-independent comparison operator.
6448 20. The saslauthd authentication condition could segfault if neither service
6449 nor realm was specified.
6451 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6452 error message that was logged was misleading, and incoming SMTP
6453 connections were dropped. The message is now more accurate, and temporary
6454 errors are given to SMTP connections.
6456 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6457 (except RSET) with a 421 error, until QUIT is received. However, it was
6458 failing to send a response to QUIT.
6460 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6461 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6462 ACL is now run after the helo_try_verify_hosts code.
6464 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6465 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6466 case-independent, so other case variants are also recognized). Apparently
6467 some systems use these upper case variants.
6469 25. If more than two messages were waiting for the same host, and a transport
6470 filter was specified for the transport, Exim sent two messages over the
6471 same TCP/IP connection, and then failed with "socket operation on non-
6472 socket" when it tried to send the third.
6474 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6476 27. The extern definition of crypt16() in expand.c was not being excluded when
6477 the OS had its own crypt16() function.
6479 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6480 as a preferred synonym for return_size_limit, both as an option and as an
6483 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6485 30. Changed the default configuration ACL to relax the local part checking rule
6486 for addresses that are not in any local domains. For these addresses,
6487 slashes and pipe symbols are allowed within local parts, but the sequence
6488 /../ is explicitly forbidden.
6490 31. SPA server authentication was not clearing the challenge buffer before
6493 32. log_message in a "warn" ACL statement was writing to the reject log as
6494 well as to the main log, which contradicts the documentation and doesn't
6495 seem right (because no rejection is happening). So I have stopped it.
6497 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6498 However, I am unable to do any testing of this.
6500 34. Fixed an infelicity in the appendfile transport. When checking directories
6501 for a mailbox, to see if any needed to be created, it was accidentally
6502 using path names with one or more superfluous leading slashes; tracing
6503 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6505 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6506 contents are added to the log line that is written for every discarded
6507 recipient. (Previously a log_message setting was ignored.)
6509 36. The ${quote: operator now quotes the string if it is empty.
6511 37. The install script runs exim in order to find its version number. If for
6512 some reason other than non-existence or emptiness, which it checks, it
6513 could not run './exim', it was installing it with an empty version number,
6514 i.e. as "exim-". This error state is now caught, and the installation is
6517 38. An argument was missing from the function that creates an error message
6518 when Exim fails to connect to the socket for saslauthd authentication.
6519 This could cause Exim to crash, or give a corrupted message.
6521 39. Added isip, isip4, and isip6 to ${if conditions.
6523 40. The ACL variables $acl_xx are now saved with the message, and can be
6524 accessed later in routers, transports, and filters.
6526 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6527 strings in the file are not string-expanded.
6529 42. If a MAIL command specified a SIZE value that was too large to fit into an
6530 int variable, the check against message_size_limit failed. Such values are
6531 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6532 day this will have to be increased, but I don't think I want to be around
6533 when emails are that large.
6540 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6541 iconv() is not standard in FreeBSD.
6543 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6544 IPv6 enabled. The observed symptom was a segmentation fault on return from
6545 the function os_common_find_running_interfaces() in src/os.c.
6547 3. In the check_special_case() function in daemon.c I had used "errno" as an
6548 argument name, which causes warnings on some systems. This was basically a
6549 typo, since it was named "eno" in the comments!
6551 4. The code that waits for the clock to tick (at a resolution of some fraction
6552 of a second) so as to ensure message-id uniqueness was always waiting for
6553 at least one whole tick, when it could have waited for less. [This is
6554 almost certainly not relevant at current processor speeds, where it is
6555 unlikely to ever wait at all. But we try to future-proof.]
6557 5. The function that sleeps for a time interval that includes fractions of a
6558 second contained a race. It did not block SIGALRM between setting the
6559 timer, and suspending (a couple of lines later). If the interval was short
6560 and the sigsuspend() was delayed until after it had expired, the suspension
6561 never ended. On busy systems this could lead to processes getting stuck for
6564 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6565 process, before it forks any delivery processes. The open lookup caching
6566 mechanism meant that the open file or database connection was passed into
6567 the delivery process. The problem was that delivery processes always tidy
6568 up cached lookup data. This could cause a problem for the next delivery
6569 process started by the queue runner, because the external queue runner
6570 process does not know about the closure. So the next delivery process
6571 still has data in the lookup cache. In the case of a file lookup, there was
6572 no problem because closing a file descriptor in a subprocess doesn't affect
6573 the parent. However, if the lookup was caching a connection to a database,
6574 the connection was closed, and the second delivery process was likely to
6575 see errors such as "PGSQL: query failed: server closed the connection
6576 unexpectedly". The problem has been fixed by closing all cached lookups
6577 in a queue runner before running a delivery process.
6579 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6580 seem to have the "const" qualifier which it has on other OS. I've
6583 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6584 *of the same type* (client or server) with the same public name that an
6585 error should be diagnosed.
6587 9. When Exim looked up a host name for an IP address, but failed to find the
6588 original IP address when looking up the host name (a safety check), it
6589 output the message "<ip address> does not match any IP for NULL", which was
6590 confusing, to say the least. The bug was that the host name should have
6591 appeared instead of "NULL".
6593 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6594 user that is built into the binary, and the -C or -D options is used, root
6595 privilege is dropped before the configuration file is read. In addition,
6596 logging is switched to stderr instead of the normal log files. If the
6597 configuration then re-defines the Exim user, the unprivileged environment
6598 is probably not what is expected, so Exim logs a panic warning message (but
6601 However, if deliver_drop_privilege is set, the unprivileged state may well
6602 be exactly what is intended, so the warning has been cut out in that case,
6603 and Exim is allowed to try to write to its normal log files.
6609 1. smtp_return_error_details was not giving details for temporary sender
6610 or receiver verification errors.
6612 2. Diagnose a configuration error if two authenticators have the same public
6615 3. Exim used not to create the message log file for a message until the first
6616 delivery attempt. This could be confusing when incoming messages were held
6617 for policy or load reasons. The message log file is now created at the time
6618 the message is received, and an initial "Received" line is written to it.
6620 4. The automatically generated man page for command line options had a minor
6621 bug that caused no ill effects; however, a more serious problem was that
6622 the procedure for building the man page automatically didn't always
6623 operate. Consequently, release 4.20 contains an out-of-date version. This
6624 shouldn't happen again.
6626 5. When building Exim with embedded Perl support, the script that builds the
6627 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6628 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6630 6. The freeze_tell option was not being used for messages that were frozen on
6631 arrival, either by an ACL or by local_scan().
6633 7. Added the smtp_incomplete_transaction log selector.
6635 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6636 was accepting AUTH without a new EHLO.
6638 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6639 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6642 10. Logging of TCP/IP connections (when configured) now happens in the main
6643 daemon process instead of the child process, so that the TCP/IP connection
6644 count is more accurate (but it can never be perfect).
6646 11. The use of "drop" in a nested ACL was not being handled correctly in the
6647 outer ACL. Now, if condition failure induced by the nested "drop" causes
6648 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6649 or "require"), the connection is dropped.
6651 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6652 that yield "discard" can now be used with an "accept" or a "discard" verb,
6653 but an error is generated for any others (because I can't see a useful way
6654 to define what should happen).
6656 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6657 are now processed in the same way as lines in the Exim configuration file.
6658 In particular, continuation lines are supported.
6660 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6662 15. Added -ti meaning -t -i.
6664 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6665 domains, and warn by logging if others are found.
6667 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6668 returned from GIFCONF when Exim is trying to find the list of interfaces on
6669 a host. The code in os.c has been modified to copy each ifreq to an aligned
6670 structure in all cases.
6672 Also, in some cases, the returned ifreq's were being copied to a 'struct
6673 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6674 means the last couple of bytes of an IPv6 address could be chopped if the
6675 ifreq contained only a normal sockaddr (14 bytes storage).
6677 18. Named domain lists were not supported in the hosts_treat_as_local option.
6678 An entry such as +xxxx was not recognized, and was treated as a literal
6681 19. Ensure that header lines added by a DATA ACL are included in the reject log
6682 if the ACL subsequently rejects the message.
6684 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6685 MD5 (which is deprecated).
6687 21. When testing a filter file using -bf, Exim was writing a message when it
6688 took the sender from a "From " line in the message, but it was not doing so
6689 when it took $return_path from a Return-Path: header line. It now does.
6691 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6692 with a valid header line field name (a series of printing characters
6693 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6695 23. Changed "disc" in the source to "disk" to conform to the documentation and
6696 the book and for uniformity.
6698 24. Ignore Sendmail's -Ooption=value command line item.
6700 25. When execve() failed while trying to run a command in a pipe transport,
6701 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6702 could be confused with a return value of 69 from the command itself. This
6703 has been changed to 127, the value the shell returns if it is asked to run
6704 a non-existent command. The wording for the related log line suggests a
6705 non-existent command as the problem.
6707 26. If received_header_text expands to an empty string, do not add a Received:
6708 header line to the message. (Well, it adds a token one on the spool, but
6709 marks it "old" so that it doesn't get used or transmitted.)
6711 27. Installed eximstats 1.28 (addition of -nt option).
6713 28. There was no check for failure on the call to getsockname() in the daemon
6714 code. This can fail if there is a shortage of resources on the system, with
6715 ENOMEM, for example. A temporary error is now given on failure.
6717 29. Contrary to the C standard, it seems that in some environments, the
6718 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6719 program. Exim now does this explicitly; it affects the formatting of
6720 timestamps using strftime().
6722 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6723 complaints. I've now initialized all the variables, to avoid this.
6725 32. Header lines added by a system filter were not being "seen" during
6726 transport-time rewrites.
6728 33. The info_callback() function passed to OpenSSL is set up with type void
6729 (*)(SSL *, int, int), as described somewhere. However, when calling the
6730 function (actually a macro) that sets it up, the type void(*)() is
6731 expected. I've put in a cast to prevent warnings from picky compilers.
6733 34. If a DNS black list lookup found a CNAME record, but there were no A
6734 records associated with the domain it pointed at, Exim crashed.
6736 35. If a DNS black list lookup returned more than one A record, Exim ignored
6737 all but the first. It now scans all returned addresses if a particular IP
6738 value is being sought. In this situation, the contents of the
6739 $dnslist_value variable are a list of all the addresses, separated by a
6742 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6743 to accept a host name and all its aliases if the forward lookup for any of
6744 them yielded the IP address of the incoming connection. Now it accepts only
6745 those names whose forward lookup yields the correct IP address. Any other
6746 names are discarded. This closes a loophole whereby a rogue DNS
6747 administrator could create reverse DNS records to break through a
6748 wildcarded host restriction in an ACL.
6750 37. If a user filter or a system filter that ran in a subprocess used any of
6751 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6752 the wrong values were passed to the pipe command ($thisaddress had the
6753 value of $0, $0 had the value of $1, etc). This bug was introduced by
6754 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6756 38. Improved the line breaking for long SMTP error messages from ACLs.
6757 Previously, if there was no break point between 40 and 75 characters, Exim
6758 left the rest of the message alone. Two changes have been made: (a) I've
6759 reduced the minimum length to 35 characters; (b) if it can't find a break
6760 point between 35 and 75 characters, it looks ahead and uses the first one
6761 that it finds. This may give the occasional overlong line, but at least the
6762 remaining text gets split now.
6764 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6765 file descriptors might be low, and that setting 1000 would always raise it.
6766 It turns out that in some environments, the limit is already over 1000 and
6767 that lowering it causes trouble. So now Exim takes care not to decrease it.
6769 40. When delivering a message, the value of $return_path is set to $sender_
6770 address at the start of routing (routers may change the value). By an
6771 oversight, this default was not being set up when an address was tested by
6772 -bt or -bv, which affected the outcome if any router or filter referred to
6775 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6776 log only when "message" or "log_message" are set. However, if one of the
6777 conditions was an address verification, or a call to a nested ACL, the
6778 messages generated by the underlying test were being passed through. This
6779 no longer happens. The underlying message is available in $acl_verify_
6780 message for both "message" and "log_message" expansions, so it can be
6781 passed through if needed.
6783 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6784 new expansion $bh_ to give the encoded byte string without charset
6785 translation. Translation happens only if iconv() is available; HAVE_ICONV
6786 indicates this at build time. HEADERS_CHARSET gives the charset to
6787 translate to; headers_charset can change it in the configuration, and
6788 "headers charset" can change it in an individual filter file.
6790 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6791 that creates RFC 2047 encoded "words" labels them as that charset instead
6792 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6793 expansion operator; (ii) when Exim creates a From: line for a local
6794 message; (iii) when a header line is rewritten to include a "phrase" part.
6796 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6797 buggy, causing it to skip the first lines of messages whose message ID
6798 ended in 'D'. This would not have bitten before Exim release 4.14, because
6799 message IDs were unlikely to end in 'D' before then. The effect was to have
6800 incorrect size information for certain domains.
6802 45. #include "config.h" was missing at the start of the crypt16.c module. This
6803 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6806 46. If there was a timeout during a "random" callout check, Exim treated it as
6807 a failure of the random address, and carried on sending RSET and the real
6808 address. If the delay was just some slowness somewhere, the response to the
6809 original RCPT would be taken as a response to RSET and so on, causing
6810 mayhem of various kinds.
6812 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6813 when I implemented it. It didn't allow for the fact that some option values
6814 may legitimately be negative (e.g. size_addition), and it didn't even do
6815 the right test for positive values.
6817 48. Domain names in DNS records are case-independent. Exim always looks them up
6818 in lower case. Some resolvers return domain names in exactly the case they
6819 appear in the zone file, that is, they may contain uppercase letters. Not
6820 all resolvers do this - some return always lower case. Exim was treating a
6821 change of case by a resolver as a change of domain, similar to a widening
6822 of a domain abbreviation. This triggered its re-routing code and so it was
6823 trying to route what was effectively the same domain again. This normally
6824 caused routing to fail (because the router wouldn't handle the domain
6825 twice). Now Exim checks for this case specially, and just changes the
6826 casing of the domain that it ultimately uses when it transmits the message
6829 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6832 50. If a filter generated a file delivery with a non-absolute name (possible if
6833 no home directory exists for the router), the forbid_file option was not
6836 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6837 the existing equality matching.
6839 52. Exim was using ints instead of ino_t variables in some places where it was
6840 dealing with inode numbers.
6842 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6843 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6844 called TMPDIR, and if it finds it is different, it changes its value.
6846 54. The smtp_printf() function is now made available to local_scan() so
6847 additional output lines can be written before returning. There is also an
6848 smtp_fflush() function to enable the detection of a dropped connection.
6849 The variables smtp_input and smtp_batched_input are exported to
6852 55. Changed the default runtime configuration: the message "Unknown user"
6853 has been removed from the ACL, and instead placed on the localuser router,
6854 using the cannot_route_message feature. This means that any verification
6855 failures that generate their own messages won't get overridden. Similarly,
6856 the "Unrouteable address" message that was in the ACL for unverifiable
6857 relay addresses has also been removed.
6859 56. Added hosts_avoid_esmtp to the smtp transport.
6861 57. The exicyclog script was not checking for the esoteric option
6862 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6863 will work only if exicyclog is run under the appropriate euid.
6865 58. Following a discussion on the list, the rules by which Exim recognises line
6866 endings on incoming messages have been changed. The -dropcr and drop_cr
6867 options are now no-ops, retained only for backwards compatibility. The
6868 following line terminators are recognized: LF CRLF CR. However, special
6869 processing applies to CR:
6871 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6872 nor a local message in the state where . is a terminator.
6874 (ii) If a bare CR is encountered in a header line, an extra space is added
6875 after the line terminator so as not to end the header. The reasoning
6876 behind this is that bare CRs in header lines are most likely either
6877 to be mistakes, or people trying to play silly games.
6879 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6880 was being incorrectly given as 18 bytes larger than it should have been.
6881 This is a VOB (very old bug).
6883 60. This may never have affected anything current, but just in case it has:
6884 When the local host is found other than at the start of a list of hosts,
6885 the local host, those with the same MX, and any that follow, are discarded.
6886 When the list in question was part of a longer list of hosts, the following
6887 hosts (not currently being processed) were also being discarded. This no
6888 longer happens. I'm not sure if this situation could ever has previously
6891 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6894 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6895 header. This is something that is recommended in a new Internet Draft, and
6896 is something that is documented as being done by Sendmail. There are two
6897 possible values. For messages generated by the autoreply transport, Exim
6900 Auto-Submitted: auto-replied
6902 whereas for all other generated messages (e.g. bounces) it adds
6904 Auto-Submitted: auto-generated
6906 63. The "personal" condition in filters now includes a test for the
6907 Auto-Submitted: header. If it contains the string "auto-" the message it
6908 not considered personal.
6910 64. Added rcpt_include_affixes as a generic transport option.
6912 65. Added queue_only_override (default true).
6914 66. Added the syslog_duplication option.
6916 67. If what should have been the first header line of a message consisted of
6917 a space followed by a colon, Exim was mis-interpreting it as a header line.
6918 It isn't of course - it is syntactically invalid and should therefore be
6919 treated as the start of the message body. The misbehaviour could have
6920 caused a number of strange effects, including loss of data in subsequent
6921 header lines, and spool format errors.
6923 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6924 client host had authenticated. This control can now be exercised by an ACL
6925 for more flexibility.
6927 69. By default, callouts do not happen when testing with -bh. There is now a
6928 variant, -bhc, which does actually run the callout code, including
6929 consulting and updating the callout cache.
6931 70. Added support for saslauthd authentication, courtesy of Alexander
6934 71. If statvfs() failed on the spool or log directories while checking their
6935 size for availability, Exim confusingly gave the error "space shortage".
6936 Furthermore, in debugging mode it crashed with a floating point exception.
6937 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6938 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6939 serious problem, Exim now writes to the main and panic logs when this
6940 happens, with details of the failure. It then refuses to accept the
6941 incoming message, giving the message "spool directory problem" or "log
6942 directory problem" with a 421 code for SMTP messages.
6944 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6945 0, 1, and 2 exist, because some OS complain for execs without them (see
6946 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6947 descriptors. However, the code omitted to check that the open succeeded,
6948 causing mysterious errors if for some reason the permissions on /dev/null
6949 got screwed. Now Exim writes a message to the main and panic logs, and
6950 bombs out if it can't open /dev/null.
6952 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6953 interact so that it is all more flexible. It is supposed to remain
6954 backwards compatible. Also added extra_local_interfaces.
6956 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6957 to bomb out with an assertion failure - to the client this appears as a
6958 connection drop. This problem occurs in the part of the code that was taken
6959 from the Samba project. Fortunately, the assertion is in a very simple
6960 function, so I have fixed this by reproducing the function inline in the
6961 one place where it is called, and arranging for authentication to fail
6962 instead of killing the process with assert().
6964 75. The SPA client code was not working when the server requested OEM rather
6965 than Unicode encoding.
6967 76. Added code to make require_files with a specific uid setting more usable in
6968 the case where statting the file as root fails - usually a non-root-mounted
6969 NFS file system. When this happens and the failure is EACCES, Exim now
6970 forks a subprocess and does the per-uid checking as the relevant uid.
6972 77. Added process_log_path.
6974 78. If log_file_path was not explicitly set, a setting of check_log_space or
6975 check_log_inodes was ignored.
6977 79. If a space check for the spool or log partitions fails, the incident is now
6978 logged. Of course, in the latter case the data may get lost...
6980 80. Added the %p formatting code to string_format() so that it can be used to
6981 print addresses in debug_print(). Adjusted all the address printing in the
6982 debugging in store.c to use %p rather than %d.
6984 81. There was a concern that a line of code in smtp_in.c could overflow a
6985 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6986 initially expressed, the concern was not well-founded, because trailing
6987 spaces are removed early. However, if the trailing spaces were followed by
6988 a NULL, they did not get removed, so the overflow was possible. Two fixes
6991 (a) I re-wrote the offending code in a cleaner fashion.
6992 (b) If an incoming SMTP command contains a NULL character, it is rejected
6995 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6996 runs initgroups(), so that if the Exim user is in any additional groups,
6997 they will be used during message reception.
7003 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob