1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Replace the store_release() internal interface with store_newblock(),
12 which internalises the check required to safely use the old one, plus
13 the allocate and data copy operations duplicated in both (!) of the
16 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
17 modifier. This matches the restriction on the commandline.
19 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
20 Previously only the last row was returned.
22 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
23 we assumed that tags in the header were well-formed, and parsed the
24 element content after inspecting only the first char of the tag.
25 Assumptions at that stage could crash the receive process on malformed
28 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
29 While running the DKIM ACL we operate on the Permanent memory pool so that
30 variables created with "set" persist to the DATA ACL. Also (at any time)
31 DNS lookups that fail create cache records using the Permanent pool. But
32 expansions release any allocations made on the current pool - so a dnsdb
33 lookup expansion done in the DKIM ACL releases the memory used for the
34 DNS negative-cache, and bad things result. Solution is to switch to the
35 Main pool for expansions.
36 While we're in that code, add checks on the DNS cache during store_reset,
37 active in the testsuite.
38 Problem spotted, and debugging aided, by Wolfgang Breyha.
40 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
41 When none of the hosts presented to a transport match an already-open
42 connection, close it and proceed with the list. Previously we would
43 queue the message. Spotted by Lena with Yahoo, probably involving
50 JH/01 Rework error string handling in TLS interface so that the caller in
51 more cases is responsible for logging. This permits library-sourced
52 string to be attached to addresses during delivery, and collapses
53 pairs of long lines into single ones.
55 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
56 during configuration. Wildcards are allowed and expanded.
58 JH/02 Rework error string handling in DKIM to pass more info back to callers.
59 This permits better logging.
61 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
62 do not close it down and have the child transport start it up again on
63 the passed-on TCP connection. Instead, proxy the child (and any
64 subsequent ones) for TLS via a unix-domain socket channel. Logging is
65 affected: the continued delivery log lines do not have any DNSSEC, TLS
66 Certificate or OCSP information. TLS cipher information is still logged.
68 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
69 identical IP addresses on different listening ports. Will also affect
72 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
73 add noisy ifdef guards to special-case this sillyness.
74 Patch from Bernd Kuhls.
76 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
77 than 255 are no longer allowed.
79 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
80 Disable the session-cache too, which might reduce our load. Since we
81 currrectly use a new context for every connection, both as server and
82 client, there is no benefit for these.
83 GnuTLS appears to not support tickets server-side by default (we don't
84 call gnutls_session_ticket_enable_server()) but client side is enabled
85 by default on recent versions (3.1.3 +) unless the PFS priority string
88 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
89 <https://reproducible-builds.org/specs/source-date-epoch/>.
91 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
92 the check for any unsuccessful recipients did not notice the limit, and
93 erroneously found still-pending ones.
95 JH/08 Pipeline CHUNKING command and data together, on kernels that support
96 MSG_MORE. Only in-clear (not on TLS connections).
98 JH/09 Avoid using a temporary file during transport using dkim. Unless a
99 transport-filter is involved we can buffer the headers in memory for
100 creating the signature, and read the spool data file once for the
101 signature and again for transmission.
103 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
104 4.77 as the kernel support then wasn't solid, having issues in 64bit
105 mode. Now, it's been long enough. Add support for FreeBSD also.
107 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
108 case where the routing stage had gathered several addresses to send to
109 a host before calling the transport for the first, we previously failed
110 to close down TLS in the old transport process before passing the TCP
111 connection to the new process. The new one sent a STARTTLS command
112 which naturally failed, giving a failed delivery and bloating the retry
113 database. Investigation and fix prototype from Wolfgang Breyha.
115 JH/12 Fix check on SMTP command input synchronisation. Previously there were
116 false-negatives in the check that the sender had not preempted a response
117 or prompt from Exim (running as a server), due to that code's lack of
118 awareness of the SMTP input buffering.
120 PP/04 Add commandline_checks_require_admin option.
121 Exim drops privileges sanely, various checks such as -be aren't a
122 security problem, as long as you trust local users with access to their
123 own account. When invoked by services which pass untrusted data to
124 Exim, this might be an issue. Set this option in main configuration
125 AND make fixes to the calling application, such as using `--` to stop
128 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
129 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
130 responses to those, into a single TLS record each way (this usually means
131 a single packet). As a side issue, smtp_enforce_sync now works on TLS
134 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
135 affects you only if you're dancing at the edge of the param size limits.
136 If you are, and this message makes sense to you, then: raise the
137 configured limit or use OpenSSL 1.1. Nothing we can do for older
140 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
141 cmdline element to get the documented default one. Previously it was
144 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
145 are unsuable for non-message-specific reasons - eg. connection timeout,
146 banner-time rejection.
148 JH/16 Fix logging of delivery remote port, when specified by router, under
151 PP/06 Repair manualroute's ability to take options in any order, even if one
152 is the name of a transport.
155 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
157 JH/17 Change the list-building routines interface to use the expanding-string
158 triplet model, for better allocation and copying behaviour.
160 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
161 Previously it was constructed the first time a possibly-matching string
162 was met in the configuration file input during startup; now it is done
165 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
166 compatible one, to avoid the (poorly documented) possibility of a config
167 file in the working directory redirecting the DB files, possibly correpting
168 some existing file. CVE-2017-10140 assigned for BDB.
170 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
171 cache-hot. Previously, although the result was properly cached, the
172 initial verify call returned a defer.
174 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
175 the main verify for receipient in uncached-mode.
177 JH/22 Retire historical build files to an "unsupported" subdir. These are
178 defined as "ones for which we have no current evidence of testing".
180 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
181 if present. Previously it was ignored.
183 JH/24 Start using specified-initialisers in C structure init coding. This is
184 a C99 feature (it's 2017, so now considered safe).
186 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
187 if was a fixed-sized field and bitmask ops via macros; it is now more
190 PP/07 GitHub PR 56: Apply MariaDB build fix.
191 Patch provided by Jaroslav Škarvada.
193 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
194 during Coverity cleanups [4.87 JH/47]
195 Diagnosis and fix provided by Michael Fischer v. Mollard.
197 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
198 the right size to place the terminating semicolon on its own folded
199 line, the header hash was calculated to an incorrect value thanks to
200 the (relaxed) space the fold became.
202 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
203 and confused the parent.
205 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
206 which could crash as a result. This could lead to undeliverable messages.
208 JH/28 Logging: "next input sent too soon" now shows where input was truncated
211 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
212 matters on fast-turnover and PID-randomising systems, which were getting
213 out-of-order delivery.
215 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
216 a possibly-overlapping copy. The symptom was that "Remote host closed
217 connection in response to HELO" was logged instead of the actual 4xx
220 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
221 Previously only that bufferd was discarded, resulting in SYMTP command
224 JH/32 DKIM: when a message has multiple signatures matching an identity given
225 in dkim_verify_signers, run the dkim acl once for each. Previously only
226 one run was done. Bug 2189.
228 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
229 "panic the current process" to "deliberately defer". The panic log is
230 still written with the problem list name; the mail and reject logs now
231 get a temp-reject line for the message that was being handled, saying
232 something like "domains check lookup or other defer". The SMTP 451
233 message is still "Temporary local problem".
235 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
236 A crafted sequence of BDAT commands could result in in-use memory beeing
237 freed. CVE-2017-16943.
239 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
240 from SMTP input. Previously it was always done; now only done for DATA
241 and not BDAT commands. CVE-2017-16944.
243 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
244 to the message (such as an overlong header line). Previously this was
245 not done and we did not exit BDAT mode. Followon from the previous item
246 though a different problem.
252 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
253 than -2003 did; needs libidn2 in addition to libidn.
255 JH/02 The path option on a pipe transport is now expanded before use.
257 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
258 Patch provided by "Björn", documentation fix added too.
260 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
261 missing a wire-to-host endian conversion.
263 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
264 close after a BDAT command line could be taken as a following command,
265 giving a synch failure. Fix by only checking for synch immediately
266 before acknowledging the chunk.
268 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
269 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
270 macro. Patches provided by Josh Soref.
272 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
273 Previously we did not; the RFC seems ambiguous and VRFY is not listed
274 by IANA as a service extension. However, John Klensin suggests that we
277 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
278 the dkim code may be unix-mode line endings rather than smtp wire-format
279 CRLF, so prepend a CR to any bare LF.
281 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
282 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
284 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
285 the first were themselves being wrongly included in the feed into dkim
286 processing; with most chunk sizes in use this resulted in an incorrect
287 body hash calculated value.
289 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
290 DKIM signature block, for verification. Although advised against by
291 standards it is specifically not ruled illegal.
293 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
295 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
296 missing a body hash (the bh= tag).
298 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
299 It seems that HAProxy sends the Proxy Protocol information in clear and
300 only then does a TLS startup, so do the same.
302 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
303 TCP connections (such as for Spamd) unless the daemon successfully set
304 Fast Open mode on its listening sockets. This fixes breakage seen on
305 too-old kernels or those not configured for Fast Open, at the cost of
306 requiring both directions being enabled for TFO, and TFO never being used
307 by non-daemon-related Exim processes.
309 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
310 endings, at least on the first header line. Try to canonify any that get
311 past that check, despite the cost.
313 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
314 now limited to an arbitrary five deep, while parsing addresses with the
315 strip_excess_angle_brackets option enabled.
317 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
318 instead leave the unprompted TLS handshake in socket buffer for the
319 TLS library to consume.
321 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
323 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
325 JH/16 Drop variables when they go out of scope. Memory management drops a whole
326 region in one operation, for speed, and this leaves assigned pointers
327 dangling. Add checks run only under the testsuite which checks all
328 variables at a store-reset and panics on a dangling pointer; add code
329 explicitly nulling out all the variables discovered. Fixes one known
330 bug: a transport crash, where a dangling pointer for $sending_ip_address
331 originally assigned in a verify callout, is re-used.
333 PP/06 Drop '.' from @INC in various Perl scripts.
335 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
337 PP/08 Reduce a number of compilation warnings under clang; building with
338 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
339 should be warning-free.
341 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
343 HS/01 Fix portability problems introduced by PP/08 for platforms where
344 realloc(NULL) is not equivalent to malloc() [SunOS et al].
346 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
347 chunk. This allows us to accept broken chunked messages. We need a more
348 general solution here.
350 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
351 already-broken messages in the queue.
353 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
355 JH/19 Fix reference counting bug in routing-generated-address tracking.
361 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
362 supports it and a size is available (ie. the sending peer gave us one).
364 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
365 years of being deprecated). The replacements are the ACLs
366 acl_smtp_mime and acl_not_smtp_mime.
368 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
369 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
370 or even an in-clear connection were permitted. Now, if the host lookup
371 was dnssec and dane was requested then the host is only used if the
372 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
373 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
374 if one fails this test.
375 This means that a poorly-configured remote DNS will make it incommunicado;
376 but it protects against a DNS-interception attack on it.
378 JH/04 Bug 1810: make continued-use of an open smtp transport connection
379 non-noisy when a race steals the message being considered.
381 JH/05 If main configuration option tls_certificate is unset, generate a
382 self-signed certificate for inbound TLS connections.
384 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
385 in rewrites and routers.
387 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
388 and logged a warning sing 4.83; now they are a configuration file error.
390 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
391 (lacking @domain). Apply the same qualification processing as RCPT.
393 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
395 JH/10 Support ${sha256:} applied to a string (as well as the previous
398 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
399 a cutthrough deliver is pending, as we always want to make a connection.
400 This also avoids re-routing the message when later placing the cutthrough
401 connection after a verify cache hit.
402 Do not update it with the verify result either.
404 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
405 when routing results in more than one destination address.
407 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
408 signing (which inhibits the cutthrough capability). Previously only
409 the presence of an option was tested; now an expansion evaluating as
410 empty is permissible (obviously it should depend only on data available
411 when the cutthrough connection is made).
413 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
414 the relevant preceding SMTP command did not note the pipelining mode.
416 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
417 Previously they were not counted.
419 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
420 as one having no matching records. Previously we deferred the message
421 that needed the lookup.
423 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
424 distinguished as "(=".
426 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
427 for missing MX records. Previously it only worked for missing A records.
429 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
431 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
432 after the data-go-ahead and data-ack. Patch from Jason Betts.
434 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
435 even for a "none" policy. Patch from Tony Meyer.
437 JH/22 Fix continued use of a connection for further deliveries. If a port was
438 specified by a router, it must also match for the delivery to be
441 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
442 When one of the recipients of a message was unsuitable for the connection
443 (has no matching addresses), we lost track of needing to mark it
444 deferred. As a result mail would be lost.
446 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
448 JH/25 Decoding ACL controls is now done using a binary search; the source code
449 takes up less space and should be simpler to maintain. Merge the ACL
450 condition decode tables also, with similar effect.
452 JH/26 Fix problem with one_time used on a redirect router which returned the
453 parent address unchanged. A retry would see the parent address marked as
454 delivered, so not attempt the (identical) child. As a result mail would
457 JH/27 Fix a possible security hole, wherein a process operating with the Exim
458 UID can gain a root shell. Credit to http://www.halfdog.net/ for
459 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
462 JH/28 Enable {spool,log} filesystem space and inode checks as default.
463 Main config options check_{log,spool}_{inodes,space} are now
464 100 inodes, 10MB unless set otherwise in the configuration.
466 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
467 Previously it only applied to the main-section connection policy
470 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
472 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
473 by me. Added RFC7919 DH primes as an alternative.
475 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
476 are not in the system include path.
478 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
479 GnuTLS, when a session startup failed (eg because the client disconnected)
480 Exim did stdio operations after fclose. This was exposed by a recent
481 change which nulled out the file handle after the fclose.
483 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
484 signed directly by the cert-signing cert, rather than an intermediate
485 OCSP-signing cert. This is the model used by LetsEncrypt.
487 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
489 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
490 an incoming connection.
492 HS/02 Bug 1802: Do not half-close the connection after sending a request
495 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
496 fallback to "prime256v1".
498 JH/34 SECURITY: Use proper copy of DATA command in error message.
499 Could leak key material. Remotely exploitable. CVE-2016-9963.
505 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
506 and 3.4.4 - once the server is enabled to respond to an OCSP request
507 it does even when not requested, resulting in a stapling non-aware
508 client dropping the TLS connection.
510 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
511 support variable-length bit vectors. No functional change.
513 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
514 The I= interface field on outgoing lines is now after the H= remote
515 host field, same as incoming lines. There is a separate
516 outgoing_interface log selector which allows you to disable the
519 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
520 If not running log_selector +smtp_connection the mainlog would be held
521 open indefinitely after a "too many connections" event, including to a
522 deleted file after a log rotate. Leave the per net connection logging
523 leaving it open for efficiency as that will be quickly detected by the
524 check on the next write.
526 HS/01 Bug 1671: Fix post transport crash.
527 Processing the wait-<transport> messages could crash the delivery
528 process if the message IDs didn't exist for some reason. When
529 using 'split_spool_directory=yes' the construction of the spool
530 file name failed already, exposing the same netto behaviour.
532 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
533 mime_regex ACL conditions.
535 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
536 to DSN fail messages (bounces): remote IP, remote greeting, remote response
537 to HELO, local diagnostic string.
539 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
540 log line to debug. Even when configured with a tls authenticator many
541 client connections are expected to not authenticate in this way, so
542 an authenticate fail is not an error.
544 HS/02 Add the Exim version string to the process info. This way exiwhat
545 gives some more detail about the running daemon.
547 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
548 matter for fast-change records such as DNSBLs.
550 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
551 retry record, even if constant. There may be multiple transports with
552 different interface settings and the retry behaviour needs to be kept
555 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
557 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
559 JH/10 Bug 840: fix log_defer_output option of pipe transport
561 JH/11 Bug 830: use same host for all RCPTS of a message, even under
562 hosts_randomize. This matters a lot when combined with mua_wrapper.
564 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
565 ${quote_pgsql:<string>} operator.
567 JH/13 Bug 1708: avoid misaligned access in cached lookup.
569 JH/14 Change header file name for freeradius-client. Relevant if compiling
570 with Radius support; from the Gentoo tree and checked under Fedora.
572 JH/15 Bug 1712: Introduce $prdr_requested flag variable
574 JH/16 Bug 1714: Permit an empty string as expansion result for transport
575 option transport_filter, meaning no filtering.
577 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
579 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
580 defaults to "*" (all hosts). The variable is now available when not built
581 with TLS, default unset, mainly to enable keeping the testsuite sane.
582 If a server certificate is not supplied (via tls_certificate) an error is
583 logged, and clients will find TLS connections fail on startup. Presumably
584 they will retry in-clear.
585 Packagers of Exim are strongly encouraged to create a server certificate
586 at installation time.
588 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
589 with the $config_file variable.
591 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
592 in transport context, after the attempt, and per-recipient. The latter type
593 is per host attempted. The event data is the error message, and the errno
594 information encodes the lookup type (A vs. MX) used for the (first) host,
595 and the trailing two digits of the smtp 4xx response.
597 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
598 to write to mainlog (or rejectlog, paniclog) in the window between file
599 creation and permissions/ownership being changed. Particularly affects
600 installations where exicyclog is run as root, rather than exim user;
601 result is that the running daemon panics and dies.
603 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
605 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
606 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
607 "pri" and "weight". Note that the previous implicit priority given by the
608 list order is no longer honoured.
610 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
613 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
614 by defining SUPPORT_SOCKS.
616 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
617 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
618 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
619 variables are renamed to proxy_{local,external}_{address,port}.
621 JH/27 Move Internationalisation support from Experimental to mainline, enabled
622 for a build by defining SUPPORT_I18N
624 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
625 of the query string, and make ${quote_redis:} do that quoting.
627 JH/29 Move Events support from Experimental to mainline, enabled by default
628 and removable for a build by defining DISABLE_EVENT.
630 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
632 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
633 cached by the daemon.
635 JH/32 Move Redis support from Experimental to mainline, enabled for a build
636 by defining LOOKUP_REDIS. The libhiredis library is required.
638 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
639 keys are given for lookup.
641 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
642 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
643 only supported when built with TLS support. The PolarSSL SHA routines
644 are still used when the TLS library is too old for convenient support.
646 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
647 openssl_options), for security. OpenSSL forces this from version 1.1.0
648 server-side so match that on older versions.
650 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
651 allocation for $value could be released as the expansion processing
652 concluded, but leaving the global pointer active for it.
654 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
655 and to use the domains and local_parts ACL conditions.
657 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
658 incorrectly not doubled on cutthrough transmission, hence seen as a
659 body-termination at the receiving system - resulting in truncated mails.
660 Commonly the sender saw a TCP-level error, and retransmitted the message
661 via the normal store-and-forward channel. This could result in duplicates
662 received - but deduplicating mailstores were liable to retain only the
663 initial truncated version.
665 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
667 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
669 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
670 we're in there, support oversigning also; bug 1309.
672 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
674 HS/04 Add support for keep_environment and add_environment options.
676 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
677 either intentional arithmetic overflow during PRNG, or testing config-
680 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
681 delivery resulted in actual delivery. Cancel cutthrough before DATA
684 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
685 rejects a recipient: pass the reject to the originator.
687 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
688 Many were false-positives and ignorable, but it's worth fixing the
691 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
692 for the new environment-manipulation done at startup. Move the routines
693 from being local to tls.c to being global via the os.c file.
695 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
696 an extract embedded as result-arg for a map, the first arg for extract
697 is unavailable so we cannot tell if this is a numbered or keyed
698 extraction. Accept either.
704 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
707 JH/02 The smtp transport option "multi_domain" is now expanded.
709 JH/03 The smtp transport now requests PRDR by default, if the server offers
712 JH/04 Certificate name checking on server certificates, when exim is a client,
713 is now done by default. The transport option tls_verify_cert_hostnames
714 can be used to disable this per-host. The build option
715 EXPERIMENTAL_CERTNAMES is withdrawn.
717 JH/05 The value of the tls_verify_certificates smtp transport and main options
718 default to the word "system" to access the system default CA bundle.
719 For GnuTLS, only version 3.0.20 or later.
721 JH/06 Verification of the server certificate for a TLS connection is now tried
722 (but not required) by default. The verification status is now logged by
723 default, for both outbound TLS and client-certificate supplying inbound
726 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
729 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
730 Status Notification (bounce) messages are now MIME format per RFC 3464.
731 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
732 under the control of the dsn_advertise_hosts option, and routers may
733 have a dsn_lasthop option.
735 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
736 default, modifiable by a malware= option. The list separator for
737 the options can now be changed in the usual way. Bug 68.
739 JH/10 The smtp_receive_timeout main option is now expanded before use.
741 JH/11 The incoming_interface log option now also enables logging of the
742 local interface on delivery outgoing connections.
744 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
745 if the interface and destination host and port all match.
747 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
750 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
751 Patch from Andrew Lewis.
753 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
754 now supports optional time-restrictions, weighting, and priority
755 modifiers per server. Patch originally by <rommer@active.by>.
757 JH/16 The spamd_address main option now supports a mixed list of local
758 and remote servers. Remote servers can be IPv6 addresses, and
759 specify a port-range.
761 JH/17 Bug 68: The spamd_address main option now supports an optional
762 timeout value per server.
764 JH/18 Bug 1581: Router and transport options headers_add/remove can
765 now have the list separator specified.
767 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
770 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
773 JH/21 Support for the A6 type of dns record is withdrawn.
775 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
776 rather than the verbs used.
778 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
779 from 255 to 1024 chars.
781 JH/24 Verification callouts now attempt to use TLS by default.
783 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
784 are generic router options now. The defaults didn't change.
786 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
787 Original patch from Alexander Shikoff, worked over by JH.
789 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
790 files if ZCAT_COMMAND is not executable.
792 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
794 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
796 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
797 Normally benign, it bites when the pair was led to by a CNAME;
798 modern usage is to not canonicalize the domain to a CNAME target
799 (and we were inconsistent anyway for A-only vs AAAA+A).
801 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
803 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
804 when evaluating $sender_host_dnssec.
806 JH/31 Check the HELO verification lookup for DNSSEC, adding new
807 $sender_helo_dnssec variable.
809 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
811 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
813 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
815 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
816 documented as working, but never had. Support all but $spam_report.
818 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
819 added for tls authenticator.
821 HS/03 Add perl_taintmode main config option
827 TL/01 When running the test suite, the README says that variables such as
828 no_msglog_check are global and can be placed anywhere in a specific
829 test's script, however it was observed that placement needed to be near
830 the beginning for it to behave that way. Changed the runtest perl
831 script to read through the entire script once to detect and set these
832 variables, reset to the beginning of the script, and then run through
833 the script parsing/test process like normal.
835 TL/02 The BSD's have an arc4random API. One of the functions to induce
836 adding randomness was arc4random_stir(), but it has been removed in
837 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
838 function when detected.
840 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
841 cause callback expansion.
843 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
844 syntax errors in an expansion can be treated as a string instead of
845 logging or causing an error, due to the internal use of bool_lax
846 instead of bool when processing it.
848 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
849 server certificates when making smtp deliveries.
851 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
853 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
855 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
857 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
858 Merged patch from Sebastian Wiedenroth.
860 JH/05 Fix results-pipe from transport process. Several recipients, combined
861 with certificate use, exposed issues where response data items split
862 over buffer boundaries were not parsed properly. This eventually
863 resulted in duplicates being sent. This issue only became common enough
864 to notice due to the introduction of connection certificate information,
865 the item size being so much larger. Found and fixed by Wolfgang Breyha.
867 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
868 size buffer was used, resulting in syntax errors when an expansion
871 JH/07 Add support for directories of certificates when compiled with a GnuTLS
872 version 3.3.6 or later.
874 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
875 is EXPERIMENTAL_EVENT, the main-configuration and transport options
876 both become "event_action", the variables become $event_name, $event_data
877 and $event_defer_errno. There is a new variable $verify_mode, usable in
878 routers, transports and related events. The tls:cert event is now also
879 raised for inbound connections, if the main configuration event_action
882 TL/06 In test suite, disable OCSP for old versions of openssl which contained
883 early OCSP support, but no stapling (appears to be less than 1.0.0).
885 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
886 server certificate names available under the smtp transport option
887 "tls_verify_cert_hostname" now do not permit multi-component wildcard
890 JH/10 Time-related extraction expansions from certificates now use the main
891 option "timezone" setting for output formatting, and are consistent
892 between OpenSSL and GnuTLS compilations. Bug 1541.
894 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
895 encoded parameter in the incoming message. Bug 1558.
897 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
898 include certificate info, eximon was claiming there were spoolfile
901 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
903 JH/14 Log delivery-related information more consistently, using the sequence
904 "H=<name> [<ip>]" wherever possible.
906 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
907 are problematic for Debian distribution, omit them from the release
910 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
912 JH/16 Fix string representation of time values on 64bit time_t architectures.
915 JH/17 Fix a null-indirection in certextract expansions when a nondefault
916 output list separator was used.
921 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
922 checkers that were complaining about end of non-void function with no
925 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
926 This was a regression introduced in 4.83 by another bugfix.
928 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
930 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
931 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
937 TF/01 Correctly close the server side of TLS when forking for delivery.
939 When a message was received over SMTP with TLS, Exim failed to clear up
940 the incoming connection properly after forking off the child process to
941 deliver the message. In some situations the subsequent outgoing
942 delivery connection happened to have the same fd number as the incoming
943 connection previously had. Exim would try to use TLS and fail, logging
944 a "Bad file descriptor" error.
946 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
947 utilities have not been installed.
949 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
950 temporary space as the ACL may create new global variables.
952 TL/01 LDAP support uses per connection or global context settings, depending
953 upon the detected version of the libraries at build time.
955 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
956 to extract and use the src ip:port in logging and expansions as if it
957 were a direct connection from the outside internet. PPv2 support was
958 updated based on HAProxy spec change in May 2014.
960 JH/02 Add ${listextract {number}{list}{success}{fail}}.
962 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
963 Properly escape header and check for NULL return.
965 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
968 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
970 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
971 characters in header names, implemented as a verify condition.
972 Contributed by Michael Fischer v. Mollard.
974 TL/05 Rename SPF condition results err_perm and err_temp to standardized
975 results permerror and temperror. Previous values are deprecated but
976 still accepted. In a future release, err_perm and err_temp will be
977 completely removed, which will be a backward incompatibility if the
978 ACL tests for either of these two old results. Patch contributed by
979 user bes-internal on the mailing list.
981 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
983 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
984 selectors, in both main and reject logs.
986 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
989 JH/07 Add malware type "sock" for talking to simple daemon.
991 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
993 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
994 routers/transports under cutthrough routing.
996 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
997 numbers. Touch up "bool" conditional to keep the same definition.
999 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1001 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1003 JH/12 Expand items in router/transport headers_add or headers_remove lists
1004 individually rather than the list as a whole. Bug 1452.
1006 Required for reasonable handling of multiple headers_ options when
1007 they may be empty; requires that headers_remove items with embedded
1008 colons must have them doubled (or the list-separator changed).
1010 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1011 view the policy declared in the DMARC record. Currently, $dmarc_status
1012 is a combined value of both the record presence and the result of the
1015 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1017 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1018 dnslookup router and the smtp transport (applying to the forward
1021 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1022 of ldap servers used for a specific lookup. Patch provided by Heiko
1025 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1026 New variable $lookup_dnssec_authenticated for observability.
1028 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1029 Patch submitted by Lars Timman.
1031 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1033 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1034 Requires trusted mode and valid format message id, aborts otherwise.
1035 Patch contributed by Heiko Schlichting.
1037 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1038 certextract with support for various fields. Bug 1358.
1040 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1041 is requested by default, modifiable by smtp transport option
1044 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1045 operate on certificate variables to give certificate fingerprints
1046 Also new ${sha256:cert_variable}.
1048 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1050 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1051 Christian Aistleitner.
1053 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1055 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1056 file. Patch from Wolfgang Breyha.
1058 JH/25 Expand the coverage of the delivery $host and $host_address to
1059 client authenticators run in verify callout. Bug 1476.
1061 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1062 align with daemon_smtp_ports. Bug 72.
1064 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1065 support and error reporting did not work properly.
1067 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1068 and is readable. Patch from Andrew Colin Kissa.
1070 TL/14 Enhance documentation of ${run expansion and how it parses the
1071 commandline after expansion, particularly in the case when an
1072 unquoted variable expansion results in an empty value.
1074 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1076 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1077 by renaming away from it. Addresses GH issue 15, reported by
1080 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1082 TL/15 SECURITY: prevent double expansion in math comparison functions
1083 (can expand unsanitized data). Not remotely exploitable.
1090 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1092 PP/02 Make -n do something, by making it not do something.
1093 When combined with -bP, the name of an option is not output.
1095 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1098 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1099 $sender_host_name and config options to manage this, and basic check
1102 PP/05 DSCP support for outbound connections and control modifier for inbound.
1104 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1105 (Only plugin which currently uses this is kerberos4, which nobody should
1106 be using, but we should make it available and other future plugins might
1107 conceivably use it, even though it would break NAT; stuff *should* be
1108 using channel bindings instead).
1110 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1111 name; added for Sendmail compatibility; requires admin caller.
1112 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1113 just ignore it); requires trusted caller.
1114 Also parse but ignore: -Ac -Am -X<logfile>
1117 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1119 TL/02 Add +smtp_confirmation as a default logging option.
1121 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1122 Patch by Magnus Holmgren from 2007-02-20.
1124 TL/04 Bugzilla 1281 - Spec typo.
1125 Bugzilla 1283 - Spec typo.
1126 Bugzilla 1290 - Spec grammar fixes.
1128 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1130 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1132 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1133 mailing list by by Dmitry Isaikin.
1135 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1137 JH/02 Support "G" suffix to numbers in ${if comparisons.
1139 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1141 NM/01 Bugzilla 1197 - Spec typo
1142 Bugzilla 1196 - Spec examples corrections
1144 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1146 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1147 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1150 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1151 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1153 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1154 "acl {{name}{arg}...}", and optional args on acl condition
1157 JH/05 Permit multiple router/transport headers_add/remove lines.
1159 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1161 JH/07 Avoid using a waiting database for a single-message-only transport.
1162 Performance patch from Paul Fisher. Bugzilla 1262.
1164 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1167 JH/09 Add $headers_added variable, with content from use of ACL modifier
1168 add_header (but not yet added to the message). Bugzilla 199.
1170 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1171 Pulled from Bugzilla 817 by Wolfgang Breyha.
1173 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1175 (nb: this is the same fix as in Exim 4.80.1)
1177 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1180 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1183 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1185 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1186 advertise SMTP AUTH mechanism to us, instead of a generic
1187 protocol violation error. Also, make Exim more robust to bad
1188 data from the Dovecot auth socket.
1190 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1192 When a queue runner is handling a message, Exim first routes the
1193 recipient addresses, during which it prunes them based on the retry
1194 hints database. After that it attempts to deliver the message to
1195 any remaining recipients. It then updates the hints database using
1198 So if a recipient address works intermittently, it can get repeatedly
1199 deferred at routing time. The retry hints record remains fresh so the
1200 address never reaches the final cutoff time.
1202 This is a fairly common occurrence when a user is bumping up against
1203 their storage quota. Exim had some logic in its local delivery code
1204 to deal with this. However it did not apply to per-recipient defers
1205 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1207 This change adds a proper retry rule check during routing so that the
1208 final cutoff time is checked against the message's age. We only do
1209 this check if there is an address retry record and there is not a
1210 domain retry record; this implies that previous attempts to handle
1211 the address had the retry_use_local_parts option turned on. We use
1212 this as an approximation for the destination being like a local
1213 delivery, as in LMTP.
1215 I suspect this new check makes the old local delivery cutoff check
1216 redundant, but I have not verified this so I left the code in place.
1218 TF/02 Correct gecos expansion when From: is a prefix of the username.
1220 Test 0254 submits a message to Exim with the header
1224 When I ran the test suite under the user fanf2, Exim expanded
1225 the header to contain my full name, whereas it should have added
1226 a Resent-Sender: header. It erroneously treats any prefix of the
1227 username as equal to the username.
1229 This change corrects that bug.
1231 GF/01 DCC debug and logging tidyup
1232 Error conditions log to paniclog rather than rejectlog.
1233 Debug lines prefixed by "DCC: " to remove any ambiguity.
1235 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1237 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1238 Bug spotted by Jeremy Harris; was flawed since initial commit.
1239 Would have resulted in OCSP responses post-SNI triggering an Exim
1240 NULL dereference and crash.
1242 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1244 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1245 Bug detection, analysis and fix by Samuel Thibault.
1246 Bugzilla 1331, Debian bug #698092.
1248 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1250 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1251 Server implementation by Todd Lyons, client by JH.
1252 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1253 config variable "prdr_enable" controls whether the server
1254 advertises the facility. If the client requests PRDR a new
1255 acl_data_smtp_prdr ACL is called once for each recipient, after
1256 the body content is received and before the acl_smtp_data ACL.
1257 The client is controlled by both of: a hosts_try_prdr option
1258 on the smtp transport, and the server advertisement.
1259 Default client logging of deliveries and rejections involving
1260 PRDR are flagged with the string "PRDR".
1262 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1263 fclose(). Diagnosis by Todd Lyons.
1265 PP/17 Update configure.default to handle IPv6 localhost better.
1266 Patch by Alain Williams (plus minor tweaks).
1269 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1270 This is now consistent with GnuTLS, and is now documented: the
1271 previous undocumented portable approach to treating the option as
1272 unset was to force an expansion failure. That still works, and
1273 an empty string is now equivalent.
1275 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1276 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1277 not performing validation itself.
1279 PP/20 Added force_command boolean option to pipe transport.
1280 Patch from Nick Koston, of cPanel Inc.
1282 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1285 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1287 PP/21 Fix eximon continuous updating with timestamped log-files.
1288 Broken in a format-string cleanup in 4.80, missed when I repaired the
1289 other false fix of the same issue.
1290 Report and fix from Heiko Schlichting.
1293 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1294 Report from Prashanth Katuri.
1296 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1297 It's SecureTransport, so affects any MacOS clients which use the
1298 system-integrated TLS libraries, including email clients.
1300 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1301 using a MIME ACL for non-SMTP local injection.
1302 Report and assistance in diagnosis by Warren Baker.
1304 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1306 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1308 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1309 last id that failed so it may be referenced in subsequent ACL's.
1311 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1314 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1315 ldap library initialization, allowing self-signed CA's to be
1316 used. Also properly sets require_cert option later in code by
1317 using NULL (global ldap config) instead of ldap handle (per
1318 session). Bug diagnosis and testing by alxgomz.
1320 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1321 the src/util/ subdirectory.
1323 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1324 renamed to Transport Post Delivery Action by Jeremy Harris, as
1327 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1328 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1329 redis_servers = needs to be configured which will be used by the redis
1330 lookup. Patch from Warren Baker, of The Packet Hub.
1332 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1334 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1335 hostname or reverse DNS when processing a host list. Used suggestions
1336 from multiple comments on this bug.
1338 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1340 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1341 Missed a few lines, added it to make the runtest require no keyboard
1344 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1345 contains upper case chars. Make router use caseful_local_part.
1347 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1348 support when GnuTLS has been built with p11-kit.
1354 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1356 This, or similar/improved, will also be change PP/11 of 4.82.
1362 PP/01 Handle short writes when writing local log-files.
1363 In practice, only affects FreeBSD (8 onwards).
1364 Bugzilla 1053, with thanks to Dmitry Isaikin.
1366 NM/01 Bugzilla 949 - Documentation tweak
1368 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1371 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1373 PP/02 Implemented gsasl authenticator.
1375 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1377 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1378 `pkg-config foo` for cflags/libs.
1380 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1381 with rest of GSASL and with heimdal_gssapi.
1383 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1384 `pkg-config foo` for cflags/libs for the TLS implementation.
1386 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1387 properties get this fed in as external SSF. A number of robustness
1388 and debugging improvements to the cyrus_sasl authenticator.
1390 PP/08 cyrus_sasl server now expands the server_realm option.
1392 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1393 Patch by Jeremy Harris.
1395 PP/10 Added dbmjz lookup type.
1397 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1399 PP/12 MAIL args handles TAB as well as SP, for better interop with
1400 non-compliant senders.
1401 Analysis and variant patch by Todd Lyons.
1403 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1404 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1405 Patch from Dirk Mueller <dmueller@suse.com>
1407 PP/13 tls_peerdn now print-escaped for spool files.
1408 Observed some $tls_peerdn in wild which contained \n, which resulted
1409 in spool file corruption.
1411 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1412 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1413 or write after TLS renegotiation, which otherwise led to messages
1416 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1417 as a tracking header (ie: a signed header comes before the signature).
1418 Patch from Wolfgang Breyha.
1420 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1421 comma-sep list; embedded commas doubled.
1423 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1425 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1427 Report and patch from Dmitry Banschikov.
1429 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1430 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1431 failures. We appear to now support TLS1.1+ with Exim.
1433 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1434 lets Exim select keys and certificates based upon TLS SNI from client.
1435 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1436 before an outbound SMTP session. New log_selector, +tls_sni.
1438 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1439 NULL dereference. Report and patch from Alun Jones.
1441 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1442 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1443 Not seeing resolver debug output on NetBSD, but suspect this is a
1444 resolver implementation change.
1446 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1447 Left warnings. Added "eximon gdb" invocation mode.
1449 PP/21 Defaulting "accept_8bitmime" to true, not false.
1451 PP/22 Added -bw for inetd wait mode support.
1453 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1454 locate the relevant includes and libraries. Made this the default.
1456 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1457 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1459 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1460 This may cause build issues on older platforms.
1462 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1463 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1464 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1465 Added SNI support via GnuTLS too.
1466 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1468 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1470 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1471 Applied second patch from Janne, implementing suggestion to default
1472 multiple-strings-in-record handling to match SPF spec.
1474 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1476 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1477 read-only, out of scope).
1478 Patch from Wolfgang Breyha, report from Stuart Northfield.
1480 PP/29 Fix three issues highlighted by clang analyser static analysis.
1481 Only crash-plausible issue would require the Cambridge-specific
1482 iplookup router and a misconfiguration.
1483 Report from Marcin Mirosław.
1485 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1487 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1488 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1489 As part of this, removing so much warning spew let me fix some minor
1490 real issues in debug logging.
1492 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1493 assignment on my part. Fixed.
1495 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1496 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1497 Janne Snabb (who went above and beyond: thank you).
1499 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1500 string otherwise requires a connection and a bunch more work and it's
1501 relatively easy to get wrong. Should also expose TLS library linkage
1504 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1505 64-bit ${eval} (JH/03).
1507 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1508 GNU libc to support some of the 64-bit stuff, should not lead to
1509 conflicts. Defined before os.h is pulled in, so if a given platform
1510 needs to override this, it can.
1512 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1513 protection layer was required, which is not implemented.
1514 Bugzilla 1254, patch from Wolfgang Breyha.
1516 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1517 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1518 tls_dhparam take prime identifiers. Also unbreak combination of
1519 OpenSSL+DH_params+TLSSNI.
1521 PP/39 Disable SSLv2 by default in OpenSSL support.
1527 PP/01 Solaris build fix for Oracle's LDAP libraries.
1528 Bugzilla 1109, patch from Stephen Usher.
1530 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1532 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1535 TF/02 Fix a couple more cases where we did not log the error message
1536 when unlink() failed. See also change 4.74-TF/03.
1538 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1539 lock up or crash if it happened to be inside a call to libc when it
1540 got a SIGUSR1 from exiwhat.
1542 The SIGUSR1 handler appends the current process status to the process
1543 log which is later printed by exiwhat. It used to use the general
1544 purpose logging code to do this, but several functions it calls are
1545 not safe for signals.
1547 The new output code in the SIGUSR1 handler is specific to the process
1548 log, and simple enough that it's easy to inspect for signal safety.
1549 Removing some special cases also simplifies the general logging code.
1550 Removing the spurious timestamps from the process log simplifies
1553 TF/04 Improved ratelimit ACL condition.
1555 The /noupdate option has been deprecated in favour of /readonly which
1556 has clearer semantics. The /leaky, /strict, and /readonly update modes
1557 are mutually exclusive. The update mode is no longer included in the
1558 database key; it just determines when the database is updated. (This
1559 means that when you upgrade Exim will forget old rate measurements.)
1561 Exim now checks that the per_* options are used with an update mode that
1562 makes sense for the current ACL. For example, when Exim is processing a
1563 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1564 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1565 must specify per_mail/readonly. If you omit the update mode it defaults to
1566 /leaky where that makes sense (as before) or /readonly where required.
1568 The /noupdate option is now undocumented but still supported for
1569 backwards compatibility. It is equivalent to /readonly except that in
1570 ACLs where /readonly is required you may specify /leaky/noupdate or
1571 /strict/noupdate which are treated the same as /readonly.
1573 A useful new feature is the /count= option. This is a generalization
1574 of the per_byte option, so that you can measure the throughput of other
1575 aggregate values. For example, the per_byte option is now equivalent
1576 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1578 The per_rcpt option has been generalized using the /count= mechanism
1579 (though it's more complicated than the per_byte equivalence). When it is
1580 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1581 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1582 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1583 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1584 Note that using per_rcpt with a non-readonly update mode in more than
1585 one ACL will cause the recipients to be double-counted. (The per_mail
1586 and per_byte options don't have this problem.)
1588 The handling of very low rates has changed slightly. If the computed rate
1589 is less than the event's count (usually one) then this event is the first
1590 after a long gap. In this case the rate is set to the same as this event's
1591 count, so that the first message of a spam run is counted properly.
1593 The major new feature is a mechanism for counting the rate of unique
1594 events. The new per_addr option counts the number of different
1595 recipients that someone has sent messages to in the last time period. It
1596 behaves like per_rcpt if all the recipient addresses are different, but
1597 duplicate recipient addresses do not increase the measured rate. Like
1598 the /count= option this is a general mechanism, so the per_addr option
1599 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1600 example, measure the rate that a client uses different sender addresses
1601 with the options per_mail/unique=$sender_address. There are further
1602 details in the main documentation.
1604 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1606 TF/06 Removed a few PCRE remnants.
1608 TF/07 Automatically extract Exim's version number from tags in the git
1609 repository when doing development or release builds.
1611 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1612 Bugzilla 879. Patch from Paul Fisher.
1614 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1615 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1618 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1620 PP/05 Variable $av_failed, true if the AV scanner deferred.
1621 Bugzilla 1078. Patch from John Horne.
1623 PP/06 Stop make process more reliably on build failure.
1624 Bugzilla 1087. Patch from Heiko Schlittermann.
1626 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1627 Bugzilla 1089. Patch from Heiko Schlittermann.
1629 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1630 Bugzilla 1131. Patch from Holger Weiß.
1632 PP/09 Handle IPv6 addresses with SPF.
1633 Bugzilla 860. Patch from Wolfgang Breyha.
1635 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1637 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1640 PP/11 match_* no longer expand right-hand-side by default.
1641 New compile-time build option, EXPAND_LISTMATCH_RHS.
1642 New expansion conditions, "inlist", "inlisti".
1644 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1646 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1648 PP/14 fix log_write() format string regression from TF/03.
1649 Bugzilla 1152. Patch from Dmitry Isaikin.
1655 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1657 PP/02 Harmonised TLS library version reporting; only show if debugging.
1658 Layout now matches that introduced for other libraries in 4.74 PP/03.
1660 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1662 PP/04 New "dns_use_edns0" global option.
1664 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1667 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1668 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1670 TK/01 Updated PolarSSL code to 0.14.2.
1671 Bugzilla 1097. Patch from Andreas Metzler.
1673 PP/07 Catch divide-by-zero in ${eval:...}.
1674 Fixes bugzilla 1102.
1676 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1679 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1680 format-string attack -- SECURITY: remote arbitrary code execution.
1682 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1683 time unintentionally subject to list matching rules, letting the header
1684 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1685 arbitrary string expansion). This allowed for information disclosure.
1687 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1688 INT_MIN/-1 -- value coerced to INT_MAX.
1694 NM/01 Workaround for PCRE version dependency in version reporting
1697 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1698 This fixes portability to compilers other than gcc, notably
1699 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1701 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1702 makefiles for portability to HP-UX and POSIX correctness.
1704 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1705 Also via indented variable definition in the Makefile.
1706 (Debugging by Oliver Heesakkers).
1708 PP/02 Restore caching of spamd results with expanded spamd_address.
1709 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1711 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1712 Improves build reliability. Fix from: Frank Elsner
1714 NM/02 Fix wide character breakage in the rfc2047 coding
1715 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1717 NM/03 Allow underscore in dnslist lookups
1718 Fixes bug 1026. Patch from Graeme Fowler
1720 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1721 Code patches from Adam Ciarcinski of NetBSD.
1723 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1726 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1727 is logged, to avoid truncation. Patch from John Horne.
1729 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1730 Patch from Jakob Hirsch.
1732 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1733 SQL string expansion failure details.
1734 Patch from Andrey Oktyabrski.
1736 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1737 Patch from Simon Arlott.
1739 PP/09 New lookups functionality failed to compile on old gcc which rejects
1740 extern declarations in function scope.
1741 Patch from Oliver Fleischmann
1743 PP/10 Use sig_atomic_t for flags set from signal handlers.
1744 Check getgroups() return and improve debugging.
1745 Fixed developed for diagnosis in bug 927 (which turned out to be
1748 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1749 Patch from Mark Zealey.
1751 PP/12 Bugzilla 1056: Improved spamd server selection.
1752 Patch from Mark Zealey.
1754 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1755 Based on patch from Heiko Schlittermann.
1757 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1758 Patch from Uwe Doering, sign-off by Michael Haardt.
1760 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1761 variable declaration deep within a block. Bug and patch from
1764 PP/15 lookups-Makefile IRIX compatibility coercion.
1766 PP/16 Make DISABLE_DKIM build knob functional.
1768 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1769 Patch by Simon Arlott
1771 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1772 variable argument macros. Our copy now differs from upstream.
1778 TF/01 Failure to get a lock on a hints database can have serious
1779 consequences so log it to the panic log.
1781 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1782 controlled using the smtp_confirmation log selector.
1784 TF/03 Include the error message when we fail to unlink a spool file.
1786 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1787 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1788 for maintaining out-of-tree patches for some time.
1790 PP/01 Bugzilla 139: Documentation and portability issues.
1791 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1792 Handle per-OS dynamic-module compilation flags.
1794 PP/02 Let /dev/null have normal permissions.
1795 The 4.73 fixes were a little too stringent and complained about the
1796 permissions on /dev/null. Exempt it from some checks.
1797 Reported by Andreas M. Kirchwitz.
1799 PP/03 Report version information for many libraries, including
1800 Exim version information for dynamically loaded libraries. Created
1801 version.h, now support a version extension string for distributors
1802 who patch heavily. Dynamic module ABI change.
1804 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1805 privilege escalation vulnerability whereby the Exim run-time user
1806 can cause root to append content of the attacker's choosing to
1809 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1812 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1813 If dropping privileges for untrusted macros, we disabled normal logging
1814 on the basis that it would fail; for the Exim run-time user, this is not
1815 the case, and it resulted in successful deliveries going unlogged.
1816 Fixed. Reported by Andreas Metzler.
1822 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1823 only prepend for the Resent-* case. Fixes regression introduced in
1824 Exim 4.70 by NM/22 for Bugzilla 607.
1826 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1827 increasing numbers of administrators be bitten by this.
1829 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1831 PP/03 Bugzilla 994: added openssl_options main configuration option.
1833 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1835 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1837 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1839 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1840 without a peer certificate, leading to a segfault because of an
1841 assumption that peers always have certificates. Be a little more
1842 paranoid. Problem reported by Martin Tscholak.
1844 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1845 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1846 NB: ClamAV planning to remove STREAM in "middle of 2010".
1847 CL also introduces -bmalware, various -d+acl logging additions and
1848 more caution in buffer sizes.
1850 PP/09 Implemented reverse_ip expansion operator.
1852 PP/10 Bugzilla 937: provide a "debug" ACL control.
1854 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1856 PP/12 Bugzilla 973: Implement --version.
1858 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1860 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1862 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1864 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1865 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1866 ignore trailing whitespace.
1868 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1870 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1873 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1874 Notification from Dr Andrew Aitchison.
1876 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1877 ExtendedDetectionInfo response format.
1878 Notification from John Horne.
1880 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1883 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1884 XSL and documented dependency on system catalogs, with examples of how
1887 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1890 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1891 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1892 configuration file which is writeable by the Exim user or group.
1894 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1895 of configuration files to cover files specified with the -C option if
1896 they are going to be used with root privileges, not just the default
1899 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1900 option (effectively making it always true).
1902 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1903 files to be used while preserving root privileges.
1905 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1906 that rogue child processes cannot use them.
1908 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1909 run-time user, instead of root.
1911 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1912 Exim run-time user without dropping privileges.
1914 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1915 result string, instead of calling string_vformat() twice with the same
1918 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1919 for other users. Others should always drop root privileges if they use
1920 -C on the command line, even for a whitelisted configure file.
1922 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1924 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1930 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1931 $data_path, and $header_path variables; fixed documentation bugs and
1934 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1935 exipick to access non-standard spools, including the "frozen" queue
1938 NM/01 Bugzilla 965: Support mysql stored procedures.
1939 Patch from Alain Williams
1941 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1943 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1944 Patch from Andreas Metzler
1946 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1947 Patch from Kirill Miazine
1949 NM/05 Bugzilla 671: Added umask to procmail example.
1951 JJ/03 installed exipick 20100323.0, fixing doc bug
1953 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1954 directory. Notification and patch from Dan Rosenberg.
1956 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1958 TK/02 Improve log output when DKIM signing operation fails.
1960 MH/01 Treat the transport option dkim_domain as a colon separated
1961 list, not as a single string, and sign the message with each element,
1962 omitting multiple occurences of the same signer.
1964 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1965 Bugzilla 985, 986. Patch by Simon Arlott
1967 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1968 Patch by Simon Arlott
1970 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1971 MBX locking. Notification from Dan Rosenberg.
1977 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1979 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1981 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1983 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1985 NM/04 Bugzilla 915: Flag broken perl installation during build.
1991 TK/01 Added patch by Johannes Berg that expands the main option
1992 "spamd_address" if it starts with a dollar sign.
1994 TK/02 Write list of recipients to X-Envelope-Sender header when building
1995 the mbox-format spool file for content scanning (suggested by Jakob
1998 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1999 (http://www.dcc-servers.net/) support via dccifd. Activated by
2000 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2002 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2003 by Mark Daniel Reidel <mr@df.eu>.
2005 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2006 When building exim an external PCRE library is now needed -
2007 PCRE is a system library on the majority of modern systems.
2008 See entry on PCRE_LIBS in EDITME file.
2010 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2011 conversation. Added nologin parameter to request.
2012 Patch contributed by Kirill Miazine.
2014 TF/01 Do not log submission mode rewrites if they do not change the address.
2016 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2018 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2019 log files in place. Contributed by Roberto Lima.
2021 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2023 TF/03 Bugzilla 615: When checking the local_parts router precondition
2024 after a local_part_suffix or local_part_prefix option, Exim now
2025 does not use the address's named list lookup cache, since this
2026 contains cached lookups for the whole local part.
2028 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2029 Robert Millan. Documentation is in experimental-spec.txt.
2031 TF/04 Bugzilla 668: Fix parallel build (make -j).
2033 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2035 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2036 Patch provided by Jan Srzednicki.
2038 TF/05 Leading white space used to be stripped from $spam_report which
2039 wrecked the formatting. Now it is preserved.
2041 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2042 that they are available at delivery time.
2044 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2046 TF/08 TLS error reporting now respects the incoming_interface and
2047 incoming_port log selectors.
2049 TF/09 Produce a more useful error message if an SMTP transport's hosts
2050 setting expands to an empty string.
2052 NM/06 Bugzilla 744: EXPN did not work under TLS.
2053 Patch provided by Phil Pennock.
2055 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2056 Patch provided by Richard Godbee.
2058 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2059 acl_smtp_notquit, added index entry.
2061 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2062 Patch provided by Eugene Bujak.
2064 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2065 accept(). Patch provided by Maxim Dounin.
2067 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2068 Patch provided by Phil Pennock.
2070 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2072 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2073 Patch provided by Brad "anomie" Jorsch.
2075 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2076 Patch provided by Dean Brooks.
2078 TK/05 Add native DKIM support (does not depend on external libraries).
2080 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2081 Patch provided by Graeme Fowler.
2083 NM/16 Bugzilla 851: Documentation example syntax fix.
2085 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2087 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2090 NM/19 Bugzilla 745: TLS version reporting.
2091 Patch provided by Phil Pennock.
2093 NM/20 Bugzilla 167: bool: condition support.
2094 Patch provided by Phil Pennock.
2096 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2097 clients. Patch provided by Phil Pennock.
2099 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2100 Patch provided by Brad "anomie" Jorsch.
2102 NM/23 Bugzilla 687: Fix misparses in eximstats.
2103 Patch provided by Heiko Schlittermann.
2105 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2106 Patch provided by Heiko Schlittermann.
2108 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2109 plus update to original patch.
2111 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2113 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2114 Patch provided by David Brownlee.
2116 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2118 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2120 NM/30 Bugzilla 888: TLS documentation bugfixes.
2122 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2124 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2125 Unlike the original bugzilla I have changed all shell scripts in src tree.
2127 NM/33 Bugzilla 898: Transport filter timeout fix.
2128 Patch by Todd Rinaldo.
2130 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2131 Patch by Serge Demonchaux.
2133 NM/35 Bugzilla 39: Base64 decode bug fixes.
2134 Patch by Jakob Hirsch.
2136 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2138 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2140 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2142 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2148 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2149 ALT-N's libdkim that I have put here:
2150 http://duncanthrax.net/exim-experimental/
2152 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2153 were called 'true' and it seems that C99 defines that as a reserved
2154 keyword to be used with 'bool' variable types. That means you could
2155 not include C99-style headers which use bools without triggering
2156 build errors in sieve.c.
2158 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2159 as mailq or other aliases. Changed the --help handling significantly
2160 to do whats expected. exim_usage() emits usage/help information.
2162 SC/01 Added the -bylocaldomain option to eximstats.
2164 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2166 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2168 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2174 PH/01 Another patch from the Sieve maintainer.
2176 PH/02 When an IPv6 address is converted to a string for single-key lookup
2177 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2178 dots are used instead of colons so that keys in lsearch files need not
2179 contain colons. This was done some time before quoting was made available
2180 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2181 (notated using the quote facility) so as to distinguish them from IPv4
2182 keys. This meant that lookups for IP addresses in host lists did not work
2183 for iplsearch lookups.
2185 This has been fixed by arranging for IPv6 addresses to be expressed with
2186 colons if the lookup type is iplsearch. This is not incompatible, because
2187 previously such lookups could never work.
2189 The situation is now rather anomalous, since one *can* have colons in
2190 ordinary lsearch keys. However, making the change in all cases is
2191 incompatible and would probably break a number of configurations.
2193 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2196 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2197 conversion specification without a maximum field width, thereby enabling
2198 a rogue spamd server to cause a buffer overflow. While nobody in their
2199 right mind would setup Exim to query an untrusted spamd server, an
2200 attacker that gains access to a server running spamd could potentially
2201 exploit this vulnerability to run arbitrary code as the Exim user.
2203 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2204 $primary_hostname instead of what libspf2 thinks the hosts name is.
2206 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2207 a directory entry by the name of the lookup key. Previously, if a
2208 symlink pointed to a non-existing file or a file in a directory that
2209 Exim lacked permissions to read, a lookup for a key matching that
2210 symlink would fail. Now it is enough that a matching directory entry
2211 exists, symlink or not. (Bugzilla 503.)
2213 PH/03 The body_linecount and body_zerocount variables are now exported in the
2216 PH/04 Added the $dnslist_matched variable.
2218 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2219 This means they are set thereafter only if the connection becomes
2222 PH/06 Added the client_condition to authenticators so that some can be skipped
2223 by clients under certain conditions.
2225 PH/07 The error message for a badly-placed control=no_multiline_responses left
2226 "_responses" off the end of the name.
2228 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2230 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2231 (without spaces) instead of just copying the configuration text.
2233 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2235 PH/11 Added $max_received_linelength.
2237 PH/12 Added +ignore_defer and +include_defer to host lists.
2239 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2240 way in which PCRE > 7.0 is built.
2242 PH/14 Implemented queue_only_load_latch.
2244 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2245 MAIL command. The effect was to mangle the value on 64-bit systems.
2247 PH/16 Another patch from the Sieve maintainer.
2249 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2251 PH/18 If a system quota error occurred while trying to create the file for
2252 a maildir delivery, the message "Mailbox is full" was not appended to the
2253 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2254 only to a quota excession during the actual writing of the file.
2256 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2257 characters?) which causes problems in log lines. The DN values are now
2258 passed through string_printing() before being added to log lines.
2260 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2261 and InterBase are left for another time.)
2263 PH/21 Added message_body_newlines option.
2265 PH/22 Guard against possible overflow in moan_check_errorcopy().
2267 PH/23 POSIX allows open() to be a macro; guard against that.
2269 PH/24 If the recipient of an error message contained an @ in the local part
2270 (suitably quoted, of course), incorrect values were put in $domain and
2271 $local_part during the evaluation of errors_copy.
2277 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2278 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2281 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2282 issue a MAIL command.
2284 PH/02 In an ACL statement such as
2286 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2288 if a client was not listed at all, or was listed with a value other than
2289 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2290 the condition was not true (as it should be), so access was not denied.
2291 The bug was that the ! inversion was incorrectly passed on to the second
2292 item. This has been fixed.
2294 PH/03 Added additional dnslists conditions == and =& which are different from
2295 = and & when the dns lookup returns more than one IP address.
2297 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2298 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2300 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2301 FSYNC, which compiles an option called disable_fsync that allows for
2302 bypassing fsync(). The documentation is heavily laced with warnings.
2304 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2306 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2307 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2308 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2309 including adding "make clean"; (3) Added -fPIC when compiling the test
2310 dynamically loaded module, to get rid of a warning.
2312 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2313 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2314 = 0s. The bug is otherwise harmless.
2316 PH/07 There was a bug in the dovecot authenticator such that the value of
2317 $auth1 could be overwritten, and so not correctly preserved, after a
2318 successful authentication. This usually meant that the value preserved by
2319 the server_setid option was incorrect.
2321 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2323 PH/09 Installed PCRE release 7.0.
2325 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2326 run for batched SMTP input. It is now run at the start of every message
2327 in the batch. While fixing this I discovered that the process information
2328 (output by running exiwhat) was not always getting set for -bs and -bS
2329 input. This is fixed, and it now also says "batched" for BSMTP.
2331 PH/11 Added control=no_pipelining.
2333 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2334 patch, slightly modified), and move the expansion of helo_data till after
2335 the connection is made in the smtp transport (so it can use these
2338 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2340 PH/14 Added log_selector = +pid.
2342 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2344 PH/16 Add ${if forany and ${if forall.
2346 PH/17 Added dsn_from option to vary the From: line in DSNs.
2348 PH/18 Flush SMTP output before performing a callout, unless control =
2349 no_callout_flush is set.
2351 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2352 was true (the default) a successful delivery failed to delete the retry
2353 item, thus causing premature timeout of the address. The bug is now
2356 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2358 PH/21 Long custom messages for fakedefer and fakereject are now split up
2359 into multiline responses in the same way that messages for "deny" and
2360 other ACL rejections are.
2362 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2363 with slight modification.
2365 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2366 draft, changing the syntax and factoring some duplicate code".
2368 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2369 for deliveries of the second and subsequent messages over the same SMTP
2372 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2373 ${reduce, with only minor "tidies".
2375 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2377 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2378 expansion side effects.
2380 PH/27 When a message times out after an over-quota error from an Exim-imposed
2381 quota, the bounce message says "mailbox is full". This message was not
2382 being given when it was a system quota that was exceeded. It now should
2385 MH/03 Made $recipients available in local_scan(). local_scan() already has
2386 better access to the recipient list through recipients_list[], but
2387 $recipients can be useful in postmaster-provided expansion strings.
2389 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2390 in the case of a MAIL command with additional options following the
2391 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2392 were accidentally chopped off.
2394 PH/29 SMTP synchronization checks are implemented when a command is read -
2395 there is a check that no more input is waiting when there shouldn't be
2396 any. However, for some commands, a delay in an ACL can mean that it is
2397 some time before the response is written. In this time, more input might
2398 arrive, invalidly. So now there are extra checks after an ACL has run for
2399 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2400 pipelining has not been advertised.
2402 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2404 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2405 returned in the SMTP response when smtp_return_error_details was false.
2406 This has been fixed.
2408 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2409 instead of the C I/O that was originally supplied, because problems were
2410 reported on Solaris.
2412 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2413 Exim which did not show up earlier: it was assuming that a call to
2414 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2415 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2416 was a macro that became an assignment, so it seemed to work. This has
2417 changed to a proper function call with a void return, hence the compile
2418 error. Exim's code has been fixed.
2420 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2423 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2425 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2427 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2428 because it turns out to be too big or there is a timeout) there is no
2429 "Completed" line in the log. When some messages of this type were
2430 selected by exigrep, they were listed as "not completed". Others were
2431 picked up by some special patterns. I have improved the selection
2432 criteria to be more general.
2434 PH/38 The host_find_failed option in the manualroute router can now be set
2435 to "ignore", to completely ignore a host whose IP address cannot be
2436 found. If all hosts are ignored, the behaviour is controlled by the new
2437 host_all_ignored option.
2439 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2440 homing or because of multiple MX records with /mx) generated more than
2441 one IP address, and the following item turned out to be the local host,
2442 all the secondary addresses of the first item were incorrectly removed
2443 from the list, along with the local host and any following hosts (which
2444 is what is supposed to happen).
2446 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2447 whoever called Exim into the -H file. In the case of the daemon it was
2448 behaving confusingly. When first started, it used values for whoever
2449 started the daemon, but after a SIGHUP it used the Exim user (because it
2450 calls itself on a restart). I have changed the code so that it now always
2453 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2454 message are rejected with the same error (e.g. no authentication or bad
2455 sender address), and a DATA command is nevertheless sent (as can happen
2456 with PIPELINING or a stupid MUA), the error message that was given to the
2457 RCPT commands is included in the rejection of the DATA command. This is
2458 intended to be helpful for MUAs that show only the final error to their
2461 PH/42 Another patch from the Sieve maintainer.
2463 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2464 Eximstats - Fixed some broken HTML links and added missing column headers
2466 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2467 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2469 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2471 PH/43 Yet another patch from the Sieve maintainer.
2473 PH/44 I found a way to check for a TCP/IP connection going away before sending
2474 the response to the final '.' that terminates a message, but only in the
2475 case where the client has not sent further data following the '.'
2476 (unfortunately, this is allowed). However, in many cases there won't be
2477 any further data because there won't be any more messages to send. A call
2478 to select() can be used: if it shows that the input is "ready", there is
2479 either input waiting, or the socket has been closed. An attempt to read
2480 the next input character can distinguish the two cases. Previously, Exim
2481 would have sent an OK response which the client would never have see.
2482 This could lead to message repetition. This fix should cure that, at
2483 least in a lot of common cases.
2485 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2486 advertised in response to EHLO.
2492 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2493 fixed by 4.65/MH/01 (is this a record?) are fixed:
2495 (i) An empty string was always treated as zero by the numeric comparison
2496 operators. This behaviour has been restored.
2498 (ii) It is documented that the numeric comparison operators always treat
2499 their arguments as decimal numbers. This was broken in that numbers
2500 starting with 0 were being interpreted as octal.
2502 While fixing these problems I realized that there was another issue that
2503 hadn't been noticed. Values of message_size_limit (both the global option
2504 and the transport option) were treated as octal if they started with 0.
2505 The documentation was vague. These values are now always treated as
2506 decimal, and I will make that clear in the documentation.
2512 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2513 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2516 MH/01 Don't check that the operands of numeric comparison operators are
2517 integers when their expansion is in "skipping" mode (fixes bug
2518 introduced by 4.64-PH/07).
2520 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2521 child addresses, Exim now panics and dies. Previously, because the count
2522 is held in a short int, deliveries were likely to be lost. As such a
2523 large number of recipients for a single message is ridiculous
2524 (performance will be very, very poor), I have chosen to impose a limit
2525 rather than extend the field.
2531 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2532 leftover -K file (the existence of which was triggered by #402).
2533 While we were at it, introduced process PID as part of the -K
2534 filename. This should rule out race conditions when creating
2537 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2538 processing considerably. Previous code took too long for large mails,
2539 triggering a timeout which in turn triggers #401.
2541 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2542 in the DK code in transports.c. sendfile() is not really portable,
2543 hence the _LINUX specificness.
2545 TF/01 In the add_headers option to the mail command in an Exim filter,
2546 there was a bug that Exim would claim a syntax error in any
2547 header after the first one which had an odd number of characters
2550 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2551 callout verification, Exim cached a "reject" for the entire domain. This
2552 is correct for most verifications, but it is not correct for a recipient
2553 verification with use_sender or use_postmaster set, because in that case
2554 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2555 case of MAIL FROM:<> rejection from other early rejections (e.g.
2556 rejection of HELO). When verifying a recipient using a non-null MAIL
2557 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2558 Whatever the result of the callout, the value of the domain cache is
2559 left unchanged (for any other kind of callout, getting as far as trying
2560 RCPT means that the domain itself is ok).
2562 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2565 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2566 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2567 session, the daemon ignores ECONNECT errors and logs others; it now
2568 ignores EPIPE as well.
2570 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2571 (quoted-printable decoding).
2573 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2574 later the small subsequent patch to fix an introduced bug.
2576 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2578 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2580 PH/08 An error is now given if message_size_limit is specified negative.
2582 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2583 to be given (somewhat) arbitrary names.
2585 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2588 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2589 miscellaneous code fixes
2591 PH/10 Added the log_reject_target ACL modifier to specify where to log
2594 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2595 hostname. This is wrong, because it relates to the incoming message (and
2596 probably the interface on which it is arriving) and not to the outgoing
2597 callout (which could be using a different interface). This has been
2598 changed to use the value of the helo_data option from the smtp transport
2599 instead - this is what is used when a message is actually being sent. If
2600 there is no remote transport (possible with a router that sets up host
2601 addresses), $smtp_active_hostname is used.
2603 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2604 tweaks were necessary in order to get it to work (see also 21 below):
2605 (a) The code assumed that strncpy() returns a negative number on buffer
2606 overflow, which isn't the case. Replaced with Exim's string_format()
2608 (b) There were several signed/unsigned issues. I just did the minimum
2609 hacking in of casts. There is scope for a larger refactoring.
2610 (c) The code used strcasecmp() which is not a standard C function.
2611 Replaced with Exim's strcmpic() function.
2612 (d) The code set only $1; it now sets $auth1 as well.
2613 (e) A simple test gave the error "authentication client didn't specify
2614 service in request". It would seem that Dovecot has changed its
2615 interface. Fortunately there's a specification; I followed it and
2616 changed what the client sends and it appears to be working now.
2618 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2621 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2622 address A is aliased to B and C, where B exists and C does not. Without
2623 -v the output is "A verified" because verification stops after a
2624 successful redirection if more than one address is generated. However,
2625 with -v the child addresses are also verified. Exim was outputting "A
2626 failed to verify" and then showing the successful verification for C,
2627 with its parentage. It now outputs "B failed to verify", showing B's
2628 parentage before showing the successful verification of C.
2630 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2631 look up a TXT record in a specific list after matching in a combined
2634 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2635 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2636 they consult the DNS. I had assumed they would set it the way they
2637 wanted; and indeed my experiments on Linux seem to show that in some
2638 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2639 To be on the safe side, however, I have now made the interface to
2640 host_find_byname() similar to host_find_bydns(), with an argument
2641 containing the DNS resolver options. The host_find_byname() function now
2642 sets these options at its start, just as host_find_bydns() does. The smtp
2643 transport options dns_qualify_single and dns_search_parents are passed to
2644 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2645 of host_find_byname() use the default settings of RES_DEFNAMES
2646 (qualify_single) but not RES_DNSRCH (search_parents).
2648 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2649 spool_read_header() do less string testing, by means of a preliminary
2650 switch on the second character of optional "-foo" lines. (This is
2651 overdue, caused by the large number of possibilities that now exist.
2652 Originally there were few.) While I was there, I also converted the
2653 str(n)cmp tests so they don't re-test the leading "-" and the first
2654 character, in the hope this might squeeze out yet more improvement.
2656 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2657 flag allowing group syntax was set by the header_syntax check but not
2658 turned off, possible causing trouble later; (2) The flag was not being
2659 set at all for the header_verify test, causing "group"-style headers to
2660 be rejected. I have now set it in this case, and also caused header_
2661 verify to ignore an empty address taken from a group. While doing this, I
2662 came across some other cases where the code for allowing group syntax
2663 while scanning a header line wasn't quite right (mostly, not resetting
2664 the flag correctly in the right place). These bugs could have caused
2665 trouble for malformed header lines. I hope it is now all correct.
2667 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2668 with the "reply" argument non-NULL. The code, however (which originally
2669 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2670 but it didn't always do it. This confused somebody who was copying the
2671 code for some other use. I have removed all the tests.
2673 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2674 feature that was used to support insecure browsers during the U.S. crypto
2675 embargo. It requires special client support, and Exim is probably the
2676 only MTA that supported it -- and would never use it because real RSA is
2677 always available. This code has been removed, because it had the bad
2678 effect of slowing Exim down by computing (never used) parameters for the
2679 RSA_EXPORT functionality.
2681 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2682 authenticator to fail if there's a tab character in the incoming data
2683 (there should never be unless someone is messing about, as it's supposed
2684 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2685 if the connection is using TLS or if the remote IP is the same as the
2686 local IP, and the "valid-client-cert option" if a client certificate has
2689 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2690 authenticators. This can be used for authorization after authentication
2691 succeeds. (In the case of plaintext, it servers for both authentication
2694 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2695 if any retry times were supplied.
2697 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2698 connection, where there is no client IP address to check. In this
2699 situation, the verify now always succeeds.
2701 PH/25 Applied John Jetmore's -Mset patch.
2703 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2705 PH/27 In a string expansion for a processed (not raw) header when multiple
2706 headers of the same name were present, leading whitespace was being
2707 removed from all of them, but trailing whitespace was being removed only
2708 from the last one. Now trailing whitespace is removed from each header
2709 before concatenation. Completely empty headers in a concatenation (as
2710 before) are ignored.
2712 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2713 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2715 PH/29 [Removed. This was a change that I later backed out, and forgot to
2716 correct the ChangeLog entry (that I had efficiently created) before
2717 committing the later change.]
2719 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2720 address errors (4xx response to RCPT) over the same connection as other
2721 messages routed to the same hosts. Such deliveries are always "forced",
2722 so retry times are not inspected. This resulted in far too many retries
2723 for the affected addresses. The effect occurred only when there were more
2724 hosts than the hosts_max_try setting in the smtp transport when it had
2725 the 4xx errors. Those hosts that it had tried were not added to the list
2726 of hosts for which the message was waiting, so if all were tried, there
2727 was no problem. Two fixes have been applied:
2729 (i) If there are any address or message errors in an SMTP delivery, none
2730 of the hosts (tried or untried) are now added to the list of hosts
2731 for which the message is waiting, so the message should not be a
2732 candidate for sending over the same connection that was used for a
2733 successful delivery of some other message. This seems entirely
2734 reasonable: after all the message is NOT "waiting for some host".
2735 This is so "obvious" that I'm not sure why it wasn't done
2736 previously. Hope I haven't missed anything, but it can't do any
2737 harm, as the worst effect is to miss an optimization.
2739 (ii) If, despite (i), such a delivery is accidentally attempted, the
2740 routing retry time is respected, so at least it doesn't keep
2741 hammering the server.
2743 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2744 in ${readsocket because some servers need this prod.
2746 PH/32 Added some extra debug output when updating a wait-xxx database.
2748 PH/33 The hint "could be header name not terminated by colon", which has been
2749 given for certain expansion errors for a long time, was not being given
2750 for the ${if def:h_colon_omitted{... case.
2752 PH/34 The spec says: "With one important exception, whenever a domain list is
2753 being scanned, $domain contains the subject domain." There was at least
2754 one case where this was not true.
2756 PH/35 The error "getsockname() failed: connection reset by peer" was being
2757 written to the panic log as well as the main log, but it isn't really
2758 panic-worthy as it just means the connection died rather early on. I have
2759 removed the panic log writing for the ECONNRESET error when getsockname()
2762 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2763 runs only) independently of the message's sender address. This meant
2764 that, if the 4xx error was in fact related to the sender, a different
2765 message to the same recipient with a different sender could confuse
2766 things. In particular, this can happen when sending to a greylisting
2767 server, but other circumstances could also provoke similar problems.
2768 I have changed the default so that the retry time for these errors is now
2769 based a combination of the sender and recipient addresses. This change
2770 can be overridden by setting address_retry_include_sender=false in the
2773 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2774 remote server are returned as part of bounce messages. This was not
2775 happening for LMTP over a pipe (the lmtp transport), but now it is the
2776 same for both kinds of LMTP.
2778 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2779 in header lines that were in fact CNAMEs. This is no longer the case.
2781 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2782 and queue runs started by the daemon processed all messages. This has
2783 been fixed so that -R and -S can now usefully be given with -q<time>.
2785 PH/40 Import PCRE release 6.7 (fixes some bugs).
2787 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2789 PH/42 Give an error if -q is specified more than once.
2791 PH/43 Renamed the variables $interface_address and $interface_port as
2792 $received_ip_address and $received_port, to make it clear that these
2793 values apply to message reception, and not to the outgoing interface when
2794 a message is delivered. (The old names remain recognized, of course.)
2796 PH/44 There was no timeout on the connect() call when using a Unix domain
2797 socket in the ${readsocket expansion. There now is.
2799 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2800 be meaningful with "accept".
2802 SC/01 Eximstats V1.43
2803 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2805 SC/02 Eximstats V1.44
2806 Use a glob alias rather than an array ref in the generated
2807 parser. This improves both readability and performance.
2809 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2810 Collect SpamAssassin and rejection statistics.
2811 Don't display local sender or destination tables unless
2812 there is data to show.
2813 Added average volumes into the top table text output.
2815 SC/04 Eximstats V1.46
2816 Collect data on the number of addresses (recipients)
2817 as well as the number of messages.
2819 SC/05 Eximstats V1.47
2820 Added 'Message too big' to the list of mail rejection
2821 reasons (thanks to Marco Gaiarin).
2823 SC/06 Eximstats V1.48
2824 Mainlog lines which have GMT offsets and are too short to
2825 have a flag are now skipped.
2827 SC/07 Eximstats V1.49 (Alain Williams)
2828 Added the -emptyok flag.
2830 SC/08 Eximstats V1.50
2831 Fixes for obtaining the IP address from reject messages.
2833 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2834 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2835 whitespace changes from 4.64-PH/27
2837 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2840 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2841 are found, allow negative numbers in numeric criteria)
2843 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2845 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2846 to match changes made in 4.64-PH/43
2848 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2850 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2851 "domain" is actually an IP address, and force a failure. This locks out
2852 those revolvers/nameservers that support "A-for-A" lookups, in
2853 contravention of the specifications.
2855 PH/48 When a host name was looked up from an IP address, and the subsequent
2856 forward lookup of the name timed out, the host name was left in
2857 $sender_host_name, contrary to the specification.
2859 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2860 restricted to single-key lookups, Exim was not diagnosing an error if
2861 * or *@ was used with a query-style lookup.
2863 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2865 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2866 long ago, but noone interested enough thought of it. Let's just say that
2867 the "1.1" means that there are some new functions that weren't there at
2868 some point in the past.
2870 PH/51 Error processing for expansion failure of helo_data from an smtp
2871 transport during callout processing was broken.
2873 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2874 tested/used via the -bh/-bhc/-bs options.
2876 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2877 bug, fixed in subsequent PCRE releases).
2879 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2880 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2882 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2888 SC/01 Use a glob alias rather than an array ref in eximstats generated
2889 parser. This improves both readability and performance.
2891 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2892 Don't display local sender or destination tables in eximstats unless
2893 there is data to show.
2894 Added average volumes into the eximstats top table text output.
2896 SC/03 Collect data on the number of addresses (recipients) as well
2897 as the number of messages in eximstats.
2899 TF/01 Correct an error in the documentation for the redirect router. Exim
2900 does (usually) call initgroups() when daemonizing.
2902 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2903 with consistent privilege compared to when running as a daemon.
2905 TF/03 Note in the spec that $authenticated_id is not set for local
2906 submissions from trusted users.
2908 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2909 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2911 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2912 by adding some example configuration directives to the default
2913 configuration file. A little bit of work is required to uncomment the
2914 directives and define how usernames and passwords are checked, but
2915 there is now a framework to start from.
2917 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2918 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2919 without this. I don't know how relevant this is to other LDAP libraries.
2921 PH/02 Add the verb name to the "unknown ACL verb" error.
2923 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2925 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2927 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2928 directory not expanded when it should be if an expanded home directory
2929 was set for the address (which is overridden by the transport).
2931 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2934 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2935 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2936 because it is too late at that time, and has no effect.
2938 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2939 security issue with \' (bugzilla #107). I could not use the
2940 PQescapeStringConn() function, because it needs a PGconn value as one of
2943 PH/08 When testing addresses using -bt, indicate those final addresses that
2944 are duplicates that would not cause an additional delivery. At least one
2945 person was confused, thinking that -bt output corresponded to deliveries.
2946 (Suppressing duplicates isn't a good idea as you lose the information
2947 about possibly different redirections that led to the duplicates.)
2949 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2950 systems where poll() doesn't work, in particular OS X.
2952 PH/10 Added more information to debugging output for retry time not reached.
2954 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2955 operations in malware.c.
2957 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2960 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2961 syslog_duplication set false, log lines that would normally be written
2962 both the the main log and to the reject log were not written to syslog at
2965 PH/14 In the default configuration, change the use of "message" in ACL warn
2966 statements to "add_header".
2968 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2969 not followed by a command (e.g. "seen endif").
2971 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2972 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2975 PH/17 Added extra conditions to the default value of delay_warning_condition
2979 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2980 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2981 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2984 The Auto-Submitted: and various List- headers are standardised, whereas I
2985 don't think Precedence: ever was.
2987 PH/18 Refactored debugging code in route_finduser() to show more information,
2988 in particular, the error code if getpwnam() issues one.
2990 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2991 This is apparently needed in addition to the PH/07 change above to avoid
2992 any possible encoding problems.
2994 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2995 but not after initializing Perl.
2997 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2998 output them only if debugging. By default they are written stderr,
2999 apparently, which is not desirable.
3001 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3004 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3007 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3009 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3010 authenticated or an ident call has been made. Suppress the default
3011 values for $authenticated_id and $authenticated_sender (but permit -oMai
3012 and -oMas) when testing with -bh.
3014 PH/24 Re-jigged the order of the tests in the default configuration so that the
3015 tests for valid domains and recipients precede the DNS black list and CSA
3016 tests, on the grounds that those ones are more expensive.
3018 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3019 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3020 This bug exists in every version of Exim that I still have, right back to
3023 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3024 However, an attempt to turn on case-sensitivity in a regex key by
3025 including (?-i) didn't work because the subject string was already
3026 lowercased, and the effects were non-intuitive. It turns out that a
3027 one-line patch can be used to allow (?-i) to work as expected.
3033 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3034 other effects) broke the use of negated acl sub-conditions.
3036 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3039 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3040 "Deny" causes Exim to reject the incoming connection with a 554 error.
3041 Unfortunately, if there is a major crisis, such as a disk failure,
3042 tcp-wrappers gives "deny", whereas what one would like would be some
3043 kind of temporary error. A kludge has been added to help with this.
3044 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3045 554 error is used if errno is still zero or contains ENOENT (which occurs
3046 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3049 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3051 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3052 errors. Otherwise a message that provokes a temporary error (when other
3053 messages do not) can cause a whole host to time out.
3055 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3056 addresses were routed directly to files or pipes from a redirect router.
3057 File deliveries just didn't batch; pipe deliveries might have suffered
3060 PH/06 A failure to get a lock for a hints database would erroneously always say
3061 "Failed to get write lock", even when it was really a read lock.
3063 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3064 of 0600. This has been changed to use the value of the lockfile_mode
3065 option (which defaults to 0600).
3067 PH/08 Applied small patch from the Sieve maintainer.
3069 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3070 folder from quota calculations, a direct delivery into this folder messed
3071 up the contents of the maildirsize file. This was because the regex was
3072 used only to exclude .Trash (or whatever) when the size of the mailbox
3073 was calculated. There was no check that a delivery was happening into an
3074 excluded directory. This bug has been fixed by ignoring all quota
3075 processing for deliveries into excluded directories.
3077 PH/10 Added the maildirfolder_create_regex option to appendfile.
3083 PH/01 The code for finding all the local interface addresses on a FreeBSD
3084 system running IPv6 was broken. This may well have applied to all BSD
3085 systems, as well as to others that have similar system calls. The broken
3086 code found IPv4 interfaces correctly, but gave incorrect values for the
3087 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3088 that it would not match correctly against @[] and not recognize the IPv6
3091 PH/02 The ipliteral router was not recognizing addresses of the form user@
3092 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3094 PH/03 Added disable_ipv6.
3096 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3097 decoded form, because it is most often used to construct To: headers
3098 lines in autoreplies, and the decoded form may well be syntactically
3099 invalid. However, $reply_address has leading white space removed, and all
3100 newlines turned into spaces so that the autoreply transport does not
3103 PH/05 If group was specified without a user on a router, and no group or user
3104 was specified on a transport, the group from the router was ignored.
3106 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3107 for visible compile-time settings that can be used to change these
3108 numbers, for those that want even more. Backwards compatibility with old
3109 spool files has been maintained. However, going back to a previous Exim
3110 release will lost any variables that are in spool files.
3112 PH/07 Two small changes when running in the test harness: increase delay when
3113 passing a TCP/IP connection to a new process, in case the original
3114 process has to generate a bounce, and remove special handling of
3115 127.0.0.2 (sic), which is no longer necessary.
3117 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3118 be the same on different OS.
3120 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3123 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3124 whether --show-vars was specified or not
3126 JJ/02 exipick: Added support for new ACL variable spool format introduced
3129 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3130 syntactically invalid From: or Reply-to: line, and a filter used this to
3131 generate an autoreply, and therefore failed to obtain an address for the
3132 autoreply, Exim could try to deliver to a non-existent relative file
3133 name, causing unrelated and misleading errors. What now happens is that
3134 it logs this as a hard delivery error, but does not attempt to create a
3137 PH/11 The exinext utility has a -C option for testing purposes, but although
3138 the given file was scanned by exinext itself; it wasn't being passed on
3139 when Exim was called.
3141 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3142 an end-of-file indication when reading a command response.
3144 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3145 compiled. In many other places in Exim, IPv6 addresses are always
3146 recognized, so I have changed this. It also means that IPv4 domain
3147 literals of the form [IPV4:n.n.n.n] are now always recognized.
3149 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3150 used if the router is not running as root, for example, when verifying at
3151 ACL time, or when using -bh. The debugging output from this situation was
3152 non-existent - all you got was a failure to exec. I have made two
3155 (a) Failures to set uid/gid, the current directory, or a process leader
3156 in a subprocess such as that created by queryprogram now generate
3157 suitable debugging output when -d is set.
3159 (b) The queryprogram router detects when it is not running as root,
3160 outputs suitable debugging information if -d is set, and then runs
3161 the subprocess without attempting to change uid/gid.
3163 PH/15 Minor change to Makefile for building test_host (undocumented testing
3166 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3167 additional section of a DNS packet that returns MX or SRV records.
3168 Instead, it always explicitly searches for A/AAAA records. This avoids
3169 major problems that occur when a DNS server includes only records of one
3170 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3171 fixed another bug: if SRV records were looked up and the corresponding
3172 address records were *not* found in the additional section, the port
3173 values from the SRV records were lost.
3175 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3176 using the correct key (the original address) when searching the retry
3177 rules in order to find which one to use for generating the retry hint.
3179 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3180 adding the default one. Similarly, if it contains a Reply-To: header, the
3181 errors_reply_to option, if set, is not used.
3183 PH/19 When calculating a retry time, Exim used to measure the "time since
3184 failure" by looking at the "first failed" field in the retry record. Now
3185 it does not use this if it is later than than the arrival time of the
3186 message. Instead it uses the arrival time. This makes for better
3187 behaviour in cases where some deliveries succeed, thus re-setting the
3188 "first failed" field. An example is a quota failure for a huge message
3189 when small messages continue to be delivered. Without this change, the
3190 "time since failure" will always be short, possible causing more frequent
3191 delivery attempts for the huge message than are intended.
3192 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3194 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3195 $1, $2, $3) because the numerical variables can be reset during some
3196 expansion items (e.g. "match"), thereby losing the authentication data.
3198 PH/21 Make -bV show the size of off_t variables so that the test suite can
3199 decide whether to run tests for quotas > 2G.
3201 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3202 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3203 filecount value is greater than 2G or if a quota value is greater than 2G
3204 on a system where the size of off_t is not greater than 4, a panic error
3207 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3208 never match. The debug and -bh output now contains an explicit error
3209 message indicating a malformed IPv4 address or mask.
3211 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3212 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3213 PH/23 above applies.
3215 PH/25 Do not write to syslog when running in the test harness. The only
3216 occasion when this arises is a failure to open the main or panic logs
3217 (for which there is an explicit test).
3219 PH/26 Added the /no_tell option to "control=freeze".
3221 PH/27 If a host name lookup failed very early in a connection, for example, if
3222 the IP address matched host_lookup and the reverse lookup yielded a name
3223 that did not have a forward lookup, an error message of the form "no IP
3224 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3225 could be logged. Now it outputs the IP address instead of "NULL".
3227 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3228 allows the sender and the authenticated sender to be set when
3229 submitting a message from within Exim. Since child_open_exim() is
3230 documented for local_scan(), the new function should be too.
3232 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3233 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3234 results in an empty string is now treated as unset.
3236 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3238 PH/31 Added sender_verify_fail logging option.
3240 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3241 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3242 tidied the source and removed it altogether.
3244 PH/33 When a queue run was abandoned because the load average was too high, a
3245 log line was always written; now it is written only if the queue_run log
3246 selector is set. In addition, the log line for abandonment now contains
3247 information about the queue run such as the pid. This is always present
3248 in "start" and "stop" lines but was omitted from the "abandon" line.
3250 PH/34 Omit spaces between a header name and the colon in the error message that
3251 is given when verify = headers_syntax fails (if there are lots of them,
3252 the message gets confusing).
3254 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3255 names, as there are now some PTR records that contain slashes. This check
3256 is only to protect against broken name servers that fall over on strange
3257 characters, so the fact that it applies to all lookups doesn't matter.
3259 PH/36 Now that the new test suite is complete, we can remove some of the
3260 special code in Exim that was needed for the old test suite. For example,
3261 sorting DNS records because real resolvers return them in an arbitrary
3262 order. The new test suite's fake resolver always returns records in the
3265 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3266 bounces) except when queue_only is set, to avoid logging races between
3267 the different processes.
3269 PH/38 Panic-die if .include specifies a non-absolute path.
3271 PH/39 A tweak to the "H" retry rule from its user.
3273 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3274 a label. They prevented compilation on older perls.
3276 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3277 a warning to be raised on newish perls.
3279 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3280 on queue. Changes to match documented behaviour of showing count of
3281 messages matching specified criteria.
3283 PH/40 Changed the default ident timeout from 30s to 5s.
3285 PH/41 Added support for the use of login_cap features, on those BSD systems
3286 that have them, for controlling the resources used by pipe deliveries.
3288 PH/42 The content-scanning code uses fopen() to create files in which to put
3289 message data. Previously it was not paying any attention to the mode of
3290 the files. Exim runs with umask(0) because the rest of the code creates
3291 files with open(), and sets the required mode explicitly. Thus, these
3292 files were ending up world-writeable. This was not a big issue, because,
3293 being within the spool directory, they were not world-accessible. I have
3294 created a function called modefopen, which takes an additional mode
3295 argument. It sets umask(777), creates the file, chmods it to the required
3296 mode, then resets the umask. All the relevant calls to fopen() in the
3297 content scanning code have been changed to use this function.
3299 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3300 to 24 hours. This avoids potential overflow problems when processing G
3301 and H retry rules. I suspect nobody ever tinkers with this value.
3303 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3305 PH/45 When the plaintext authenticator is running as a client, the server's
3306 challenges are checked to ensure they are valid base64 strings. By
3307 default, the authentication attempt is cancelled if an invalid string is
3308 received. Setting client_ignore_invalid_base64 true ignores these errors.
3309 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3310 they are received. Thus, the responses can be made to depend on the
3311 challenges. If an invalid string is ignored, an empty string is placed in
3314 PH/46 Messages that are created by the autoreply transport now contains a
3315 References: header, in accordance with RFCs 2822 and 3834.
3317 PH/47 Added authenticated_sender_force to the smtp transport.
3319 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3321 PH/49 Installed latest patch from the Sieve maintainer.
3323 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3324 was also set, the appendfile transport was unnecessarily scanning a
3325 directory of message files (e.g. for maildir delivery) to find the count
3326 of files (along with the size), even though it did not need this
3327 information. It now does the scan only if it needs to find either the
3328 size of the count of files.
3330 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3332 PH/52 Two bugs concerned with error handling when the smtp transport is
3335 (i) Exim was not creating retry information for temporary errors given
3336 for individual recipients after the DATA command when the smtp transport
3337 was used in LMTP mode. This meant that they could be retried too
3338 frequently, and not timed out correctly.
3340 (ii) Exim was setting the flag that allows error details to be returned
3341 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3342 recipients that were returned after the DATA command.
3344 PH/53 This is related to PH/52, but is more general: for any failing address,
3345 when detailed error information was permitted to be returned to the
3346 sender, but the error was temporary, then after the final timeout, only
3347 "retry timeout exceeded" was returned. Now it returns the full error as
3348 well as "retry timeout exceeded".
3350 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3351 do this, and (what is worse) MTAs that accept it.
3353 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3354 will now be deprecated.
3356 PH/56 New os.c-cygwin from the Cygwin maintainer.
3358 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3359 formats (previously only available in exim formats via -bpr, -bpru,
3360 and -bpra. Now also available in native and exiqgrep formats)
3362 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3363 with very large, slow to parse queues
3365 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3367 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3369 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3370 responses to authentication challenges, though it was showing the
3371 challenges; (ii) I've removed the CR characters from the debug output for
3374 PH/58 Allow for the insertion of a newline as well as a space when a string
3375 is turned into more than one encoded-word during RFC 2047 encoding. The
3376 Sieve code now uses this.
3378 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3379 data_4xx, lost_connection, tls_required.
3381 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3382 message was being sent as an SMTP response.
3384 PH/61 Add -l and -k options to exicyclog.
3386 PH/62 When verifying, if an address was redirected to one new address, so that
3387 verification continued, and the new address failed or deferred after
3388 having set something in $address_data, the value of $address_data was not
3389 passed back to the ACL. This was different to the case when no
3390 redirection occurred. The value is now passed back in both cases.
3392 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3393 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3394 use login_cap.h, so on its own it isn't the distinguishing feature. The
3395 new name refers directly to the setclassresources() function.
3397 PH/65 Added configuration files for NetBSD3.
3399 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3401 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3402 is preferred over IPv4.
3404 PH/68 The bounce_return_message and bounce_return_body options were not being
3405 honoured for bounces generated during the reception of non-SMTP messages.
3406 In particular, this applied to messages rejected by the ACL. This bug has
3407 been fixed. However, if bounce_return_message is true and bounce_return_
3408 body is false, the headers that are returned for a non-SMTP message
3409 include only those that have been read before the error was detected.
3410 (In the case of an ACL rejection, they have all been read.)
3412 PH/69 The HTML version of the specification is now built in a directory called
3413 spec_html instead of spec.html, because the latter looks like a path with
3414 a MIME-type, and this confuses some software.
3416 PH/70 Catch two compiler warnings in sieve.c.
3418 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3419 function verify_get_ident() calls ip_connect() to connect a socket, but
3420 if the "connect()" function timed out, ip_connect() used to close the
3421 socket. However, verify_get_ident() also closes the socket later, and in
3422 between Exim writes to the log, which may get opened at this point. When
3423 the socket was closed in ip_connect(), the log could get the same file
3424 descriptor number as the socket. This naturally causes chaos. The fix is
3425 not to close the socket in ip_connect(); the socket should be closed by
3426 the function that creates it. There was only one place in the code where
3427 this was missing, in the iplookup router, which I don't think anybody now
3428 uses, but I've fixed it anyway.
3430 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3431 well as to direct DNS lookups. Otherwise the handling of names in host
3432 lists is inconsistent and therefore confusing.
3438 PH/01 Two changes to the default runtime configuration:
3440 (1) Move the checks for relay_from_hosts and authenticated clients from
3441 after to before the (commented out) DNS black list checks.
3443 (2) Add control=submission to the relay_from_hosts and authenticated
3444 clients checks, on the grounds that messages accepted by these
3445 statements are most likely to be submissions.
3447 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3449 (1) Generate an error if the third argument for the ${prvs expansion is
3452 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3455 (3) Reset the variables that are obtained from the first argument of
3456 ${prvscheck and used in the second argument before leaving the code,
3457 because their memory is reclaimed, so using them afterwards may do
3460 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3461 one (it's much easier than Tom thought :-).
3463 (5) Because of (4), we can now allow for the use of $prvscheck_result
3464 inside the third argument.
3466 PH/03 For some reason, the default setting of PATH when running a command from
3467 a pipe transport was just "/usr/bin". I have changed it to
3470 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3471 anything to be listed in the output from -bV.
3473 PH/05 When a filter generated an autoreply, the entire To: header line was
3474 quoted in the delivery log line, like this:
3476 => >A.N.Other <ano@some.domain> <original@ddress> ...
3478 This has been changed so that it extracts the operative address. There
3479 may be more than one such address. If so, they are comma-separated, like
3482 => >ano@some.domain,ona@other.domain <original@ddress> ...
3484 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3485 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3486 not being looked up in the rDNS to get a host name, Exim was showing the
3487 IP address twice in Received: lines, even though the IP addresses were
3488 identical. For example:
3490 Received: from [1.2.3.4] (helo=[1.2.3.4])
3492 However, if the real host name was known, it was omitting the HELO data
3493 if it matched the actual IP address. This has been tidied up so that it
3494 doesn't show the same IP address twice.
3496 PH/07 When both +timestamp and +memory debugging was on, the value given by
3497 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3498 called by the debug printing, thereby overwriting the timestamp buffer.
3499 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3501 PH/08 When the original message was included in an autoreply transport, it
3502 always said "this is a copy of the message, including all the headers",
3503 even if body_only or headers_only was set. It now gives an appropriate
3506 PH/09 Applied a patch from the Sieve maintainer which:
3508 o fixes some comments
3509 o adds the (disabled) notify extension core
3510 o adds some debug output for the result of if/elsif tests
3511 o points to the current vacation draft in the documentation
3512 and documents the missing references header update
3516 o fixes a bug in processing the envelope test (when testing
3517 multiple envelope elements, the last element determined the
3520 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3521 Electronic Mail") by including:
3523 Auto-submitted: auto-generated
3525 in the messages that it generates (bounce messages and others, such as
3526 warnings). In the case of bounce messages for non-SMTP messages, there was
3527 also a typo: it was using "Auto_submitted" (underscore instead of
3528 hyphen). Since every message generated by Exim is necessarily in response
3529 to another message, thes have all been changed to:
3531 Auto-Submitted: auto-replied
3533 in accordance with these statements in the RFC:
3535 The auto-replied keyword:
3537 - SHOULD be used on messages sent in direct response to another
3538 message by an automatic process,
3540 - MUST NOT be used on manually-generated messages,
3542 - MAY be used on Delivery Status Notifications (DSNs) and Message
3543 Disposition Notifications (MDNs),
3545 - MUST NOT be used on messages generated by automatic or periodic
3546 processes, except for messages which are automatic responses to
3549 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3550 to the default Received: header definition.
3552 PH/12 Added log selector acl_warn_skipped (default on).
3554 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3555 variables because (a) they are in the wrong storage pool and (b) even if
3556 they were copied, it wouldn't work properly because of the caching.
3558 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3559 checking when decoding. Apparently there are clients that generate
3560 overlong encoded strings. Why am I not surprised?
3562 PH/15 If the first argument of "${if match_address" was not empty, but did not
3563 contain an "@" character, Exim crashed. Now it writes a panic log message
3564 and treats the condition as false.
3566 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3568 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3569 extension "envelope-auth". The code is finished and in agreement with
3570 other implementations, but there is no documentation so far and in fact,
3571 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3572 not changing the active code.
3574 Print executed "if" and "elsif" statements when debugging is used. This
3575 helps a great deal to understand what a filter does.
3577 Document more things not specified clearly in RFC3028. I had all this
3578 sorted out, when out of a sudden new issues came to my mind. Oops."
3580 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3583 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3584 canonical form (as documented). However, after a host name lookup from
3585 the IP address, check_host() was doing a simple string comparison with
3586 addresses acquired from the DNS when checking that the found name did
3587 have the original IP as one of its addresses. Since any found IPv6
3588 addresses are likely to be in abbreviated form, the comparison could
3589 fail. Luckily, there already exists a function for doing the comparison
3590 by converting both addresses to binary, so now that is used instead of
3591 the text comparison.
3593 PH/20 There was another similar case to PH/19, when a complete host name was
3594 given in a host list; looking up its IP address could give an abbreviated
3595 form, whereas the current host's name might or might not be abbreviated.
3596 The same fix has been applied.
3602 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3603 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3606 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3607 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3609 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3611 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3612 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3613 Also, for "Auto-Submitted", treat anything other than "no" as
3614 non-personal, in accordance with RFC 3834. (Previously it treated
3615 anything starting "auto-" as non-personal.)
3617 TF/01 The control=submission/name=... option had a problem with syntax
3618 errors if the name included a slash character. The /name= option
3619 now slurps the rest of the string, so it can include any characters
3620 but it must come last in the list of options (after /sender_retain
3623 PH/05 Some modifications to the interface to the fake nameserver for the new
3631 TK/01 Added the "success_on_redirect" address verification option. See
3632 NewStuff for rationale and an example.
3634 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3636 PH/02 Patch to exigrep to allow it to work on syslog lines.
3638 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3639 fread() to skip over the body file's header line, because in Cygwin the
3640 header line is locked and is inaccessible.
3642 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3643 co-exist for some time) to make it clear that it is the Exim ID that is
3644 referenced, not the Message-ID: header line.
3646 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3647 string_format() function, because snprintf() does not exist on all
3650 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3651 ${stat: expansion item.
3653 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3654 protocol synchronization error", to keep the pedants happy.
3656 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3657 well as for IRIX systems, when gcc is being used. See the host.c source
3660 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3662 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3665 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3666 transport and to the smtp transport in LMTP mode.
3668 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3670 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3671 run a filter in a subprocess. This could lead to confusion in subsequent
3672 lookups in the parent process. There should also be a search_tidyup() at
3673 the end of the subprocess.
3675 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3676 only if the host matched helo_try_verify_hosts, which caused the
3677 verification to occur when the EHLO/HELO command was issued. The ACL just
3678 tested the remembered result. Now, if a previous verification attempt has
3679 not happened, "verify = helo" does it there and then.
3681 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3683 TK/03 Fix log output including CR from clamd.
3685 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3686 exist provoked a memory error which could cause a segfault.
3688 PH/15 Installed PCRE 6.2
3690 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3692 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3693 of the problem. Specifically, suggested +O2 rather than +O1 for the
3696 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3698 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3699 record, the port was not correctly logged when the outgoing_port log
3700 selector was set (it logged the transort's default port).
3702 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3703 fallback_hosts, and "hosts" in the smtp transport.
3705 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3706 host errors such as "Connection refused".
3708 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3709 authentication with radiusclient 0.4.9:
3711 - Error returned from rc_read_config was caught wrongly
3712 - Username/password not passed on to radius server due to wrong length.
3714 The presumption is that some radiusclient API changes for 4.51/PH/17
3715 were not taken care of correctly. The code is still untested by me (my
3716 Linux distribution still has 0.3.2 of radiusclient), but it was
3717 contributed by a Radius user.
3719 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3720 expanding the "port" option of the smtp transport.
3722 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3723 while reading a MIME header. Thanks to Tom Hughes for a patch.
3725 PH/24 Include config.h inside local_scan.h so that configuration settings are
3728 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3729 that in an ACL for RCPT (for example), you can examine exactly what was
3732 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3733 commands, but it was not correctly comparing the address with the actual
3734 client host address. Thus, it would show the EHLO address in Received:
3735 header lines when this was not necessary.
3737 PH/27 Added the % operator to ${eval:}.
3739 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3740 it should be ignoring failures (because with -C, for example, it has lost
3741 privilege). It wasn't ignoring creation failures other than "already
3744 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3747 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3748 because an input line was too long, either on its own, or by virtue of
3749 too many continuations, the temporary file was not being removed, and the
3750 return code was incorrect.
3752 PH/31 Missing "BOOL" in function definition in filtertest.c.
3754 PH/32 Applied Sieve patches from the maintainer.
3756 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3758 PH/33 Added "verify = not_blind".
3760 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3761 Local/Makefile (with some defaults set). These are used in built scripts
3762 such as exicyclog, but they have never been used in the exim_install
3763 script (though there are many overriding facilities there). I have
3764 arranged that the exim_install script now takes note of these two
3767 PH/35 Installed configuration files for Dragonfly.
3769 PH/36 When a locally submitted message by a trusted user did not contain a
3770 From: header, and the sender address was obtained from -f or from an SMTP
3771 MAIL command, and the trusted user did not use -F to supply a sender
3772 name, $originator_name was incorrectly used when constructing a From:
3773 header. Furthermore, $originator_name was used for submission mode
3774 messages from external hosts without From: headers in a similar way,
3775 which is clearly wrong.
3777 PH/37 Added control=suppress_local_fixups.
3779 PH/38 When log_selector = +received_sender was set, and the addition of the
3780 sender made the log line's construction buffer exactly full, or one byte
3781 less than full, an overflow happened when the terminating "\n" was
3784 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3785 when the result of a list match is failure because a DNS lookup failed.
3787 PH/40 RM_COMMAND is now used in the building process.
3789 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3790 the "build-* directories that it finds.
3792 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3793 address in a domain literal was a prefix of an interface address.
3795 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3796 when verifying a sender address, unless rewrite_headers is false.
3798 PH/44 Wrote a long comment about why errors_to addresses are verified as
3799 recipients, not senders.
3801 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3802 the ratelimit ACL was added.
3804 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3806 PH/46 Added extra information about PostgreSQL errors to the error string.
3808 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3809 avoiding the need to install special zones in a real server. This is
3810 backwards compatible; if it can't find the fake resolver, it drops back.
3811 Thus, both old and new test suites can be run.
3813 TF/02 Added util/ratelimit.pl
3815 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3816 clock is set back in time.
3818 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3819 Candler <B.Candler@pobox.com>.
3821 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3822 OK for addresses that are the result of redirections.
3824 PH/48 A number of further additions for the benefit of the new test suite,
3825 including a fake gethostbyname() that interfaces to the fake DNS resolver
3828 TF/06 The fix for widen_domains has also been applied to qualify_single and
3829 search_parents which are the other dnslookup options that can cause
3832 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3835 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3837 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3838 service" if it encountered an MX record with an empty target hostname.
3839 The message is now "an MX or SRV record indicated no SMTP service".
3841 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3842 if the DNS of the sending site is misconfigured. This is quite a
3843 common situation. This change restores the behaviour of treating a
3844 helo verification defer as a failure.
3846 PH/51 If self=fail was set on a router, the bounce message did not include the
3847 actual error message.
3853 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3855 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3856 command itself ended in error, the underlying message about the transport
3857 filter timeout was being overwritten with the pipe command error. Now the
3858 underlying error message should be appended to the second error message.
3860 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3862 PH/02 Reduce the amount of output that "make" produces by default. Full output
3863 can still be requested.
3865 PH/03 The warning log line about a condition test deferring for a "warn" verb
3866 was being output only once per connection, rather than after each
3867 occurrence (because it was using the same function as for successful
3868 "warn" verbs). This seems wrong, so I have changed it.
3870 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3871 it should not have, which might have caused a crash in the right
3872 circumstances, but probably never did.
3874 PH/04 Installed a modified version of Tony Finch's patch to make submission
3875 mode fix the return path as well as the Sender: header line, and to
3876 add a /name= option so that you can make the user's friendly name appear
3879 TF/03 Added the control = fakedefer ACL modifier.
3881 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3882 Mark Lowes for thorough testing.
3884 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3886 TK/03 Merged latest SRS patch from Miles Wilton.
3888 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3889 with the definition in sysexits.h (which is #included earlier).
3890 Fortunately, Exim does not actually use EX_OK. The code used to try to
3891 preserve the sysexits.h value, by assuming that macro definitions were
3892 scanned for macro replacements. I have been disabused of this notion,
3893 so now the code just undefines EX_OK before #including unistd.h.
3895 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3896 in the smtp transport. When a block could not be written in a single
3897 write() function, the timeout was being re-applied to each part-write.
3898 This seems wrong - if the receiver was accepting one byte at a time it
3899 would take for ever. The timeout is now adjusted when this happens. It
3900 doesn't have to be particularly precise.
3902 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3903 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3905 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3906 without a domain if the check to <postmaster@domain> fails.
3908 SC/01 Eximstats: added -xls and the ability to specify output files
3909 (patch written by Frank Heydlauf).
3911 SC/02 Eximstats: use FileHandles for outputting results.
3913 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3915 SC/04 Eximstats: fixed display of large numbers with -nvr option
3917 SC/05 Eximstats: fixed merging of reports with empty tables.
3919 SC/06 Eximstats: added the -include_original_destination flag
3921 SC/07 Eximstats: removed tabs and trailing whitespace.
3923 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3925 TK/06 MBOX spool code: Add real "From " MBOX separator line
3926 so the .eml file is really in mbox format (even though
3927 most programs do not really care). Patch from Alex Miller.
3929 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3930 The latter is generated from $received_to and is only set if the
3931 message has one envelope recipient. SA can use these headers,
3932 obviously out-of-the-box. Patch from Alex Miller.
3934 PH/08 The ${def test on a variable was returning false if the variable's
3935 value was "0", contrary to what the specification has always said!
3936 The result should be true unless the variable is empty.
3938 PH/09 The syntax error of a character other than { following "${if
3939 def:variable_name" (after optional whitespace) was not being diagnosed.
3940 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3941 accidental colon was present, for example, could give incorrect results.
3943 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3944 result is used (not including appendfile, where other changes are about
3947 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3948 This involved changing a lot of size variables from int to off_t. It
3949 should work with maildirs and everything.
3951 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3952 spamd dying while we are connected to it.
3954 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3957 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3958 function for BDB 4.3.
3960 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3962 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3963 the sources for PCRE have been re-arranged and more files are now
3966 PH/15 The code I had for printing potentially long long variables in PH/11
3967 above was not the best (it lost precision). The length of off_t variables
3968 is now inspected at build time, and an appropriate printing format (%ld
3969 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3970 to be "long long int" or "long int". This is needed for the internal
3971 formatting function string_vformat().
3973 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3974 the configuration file to be ":syslog", then the script "guesses" where
3975 the logs files are, rather than using the compiled in default. In our
3976 case the guess is not the same as the compiled default, so the script
3977 suddenly stopped working when I started to use syslog. The patch checks
3978 to see if log_file_path is "". If so, it attempts to read it from exim
3979 with no configuration file to get the compiled in version, before it
3980 falls back to the previous guessing code."
3982 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3983 implementing BATV in an Exim configuration. See NewStuff for the gory
3986 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3987 Makefile that are specific to HP-UX.
3989 PH/18 If the "use_postmaster" option was set for a recipient callout together
3990 with the "random" option, the postmaster address was used as the MAIL
3991 FROM address for the random test, but not for the subsequent recipient
3992 test. It is now used for both.
3994 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3995 patch removes a few documentation additions to RFC 3028, because the
3996 latest draft now contains them. It adds the new en;ascii-case comparator
3997 and a new error check for 8bit text in MIME parts. Comparator and
3998 require names are now matched exactly. I enabled the subaddress
3999 extension, but it is not well tested yet (read: it works for me)."
4001 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4002 rework some of the code of TK/09 above to avoid the hardwired use of
4003 "%lld" and "long long". Replaced the call to snprintf() with a call to
4006 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4007 records point to non-existent hosts", "retry timeout exceeded", and
4008 "retry time not reached for any host after a long failure period".
4010 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4011 experimental DomainKeys support:
4013 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4014 (2) On an error such as an illegally used "control", the wrong name for
4015 the control was given.
4017 These problems did NOT occur unless DomainKeys support was compiled.
4019 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4021 PH/24 Added ${if match_ip condition.
4023 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4024 databases so that it will be absolutely obvious if a crash occurs in the
4025 DB library. This is a regular occurrence (often caused by mis-matched
4028 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4029 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4030 picked up on a user's system that detects such things. There doesn't seem
4031 to be a gcc warning option for this - only an attribute that has to be
4032 put on the function's prototype. It seems that in Fedora Core 4 they have
4033 set this on a number of new functions. No doubt there will be more in due
4036 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4037 specify a transport. However, if an address that was verified by such a
4038 router was the subject of a callout, Exim crashed because it tried to
4039 read the rcpt_include_affixes from the non-existent transport. Now it
4040 just assumes that the setting of that option is false. This bug was
4041 introduced by 4.51/PH/31.
4043 PH/28 Changed -d+all to exclude +memory, because that information is very
4044 rarely of interest, but it makes the output a lot bigger. People tend to
4045 do -d+all out of habit.
4047 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4048 code in os-type was giving problems when libc.so lives in lib64, like on
4051 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4052 aren't the modern standard, and it seems that some systems' include files
4053 don't always have them. Exim was already checking for some of the newer
4054 ones like T_AAAA, and defining it itself. I've added checks for all the
4055 record types that Exim uses.
4057 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4058 not automatically generating a new one, as it is supposed to. This
4059 prevented TLS from working. If the file did exist, but contained invalid
4060 data, a new version was generated, as expected. It was only the case of a
4061 non-existent file that was broken.
4063 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4064 with a change in libdomainkeys > 0.64.
4066 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4067 from DNS. If the selector record carries the flag, it now has
4068 precedence over the domain-wide flag.
4070 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4072 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4073 the use of a port name that isn't defined in /etc/services) occurred, the
4074 message was deferred as in a normal delivery, and thus remained on the
4075 spool, instead of being failed because of the mua_wrapper setting. This
4076 is now fixed, and I tidied up some of the mua_wrapper messages at the
4079 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4080 the messages in a hash of arrays rather than using individual hashes.
4081 This is a bit cleaner and results in dramatic memory savings, albeit
4082 at a slight CPU cost.
4084 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4085 as requested by Marc Sherman.
4087 SC/10 Eximstats: added histograms for user specified patterns as requested
4090 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4092 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4093 fopen() in the content-scanning modules that did not already have it.
4099 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4100 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4102 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4104 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4106 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4107 to test Sieve filters that use "vacation".
4109 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4110 that changes the way the GnuTLS parameters are stored in the cache file.
4111 The new format can be generated externally. For backward compatibility,
4112 if the data in the cache doesn't make sense, Exim assumes it has read an
4113 old-format file, and it generates new data and writes a new file. This
4114 means that you can't go back to an older release without removing the
4117 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4118 work if there are any delivery delays because "one_time" forces the
4119 parent to be marked "delivered", so its unseen clone is never tried
4120 again. For this reason, Exim now forbids the simultaneous setting of
4123 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4124 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4125 entry if you want to know the details. The fix, however, neglected to
4126 consider the case when local delivery batching is involved. The test for
4127 "previously delivered" was not happening when checking to see if an
4128 address could be batched with a previous (undelivered) one; under
4129 certain circumstances this could lead to multiple deliveries to the same
4132 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4133 in its include files, and this causes problems building Exim.
4135 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4136 header_syntax) but Exim was just ignoring anything given after a slash.
4137 In particular, this caused confusion with an attempt to use "verify =
4138 reverse_host_lookup/defer_ok". An error is now given when options are
4139 supplied for verify items that do not have them. (Maybe reverse_host_
4140 lookup should have a defer_ok option, but that's a different point.)
4142 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4143 defined by RFC 821) to 2048, because there were problems with some AUTH
4144 commands, and RFC 1869 says the size should be increased for extended
4145 SMTP commands that take arguments.
4147 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4150 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4151 "unknown" error; now it says that the functionality isn't in the binary.
4153 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4154 an address' error message when a string expansion fails (syntax or
4155 whatever). Otherwise the password may appear in the log. Following change
4156 PH/42 below, there is no longer a chance of it appearing in a bounce
4159 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4161 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4162 of its addresses was ever tried. (Bugzilla bug #2.)
4164 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4165 the result incorrectly in the debug output. (It correctly added a newline
4166 to what was transported.)
4168 TF/01 Added $received_time.
4170 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4171 commented out examples of how to interface to a virus scanner and to
4172 SpamAssassin. Also added commented examples of av_scanner and
4173 spamd_address settings.
4175 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4176 and controls are allowed in which ACLs. There were a couple of minor
4177 errors. Some of the entries in the conditions table (which is a table of
4178 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4179 negation of where the condition IS allowed.
4181 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4183 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4184 header file does not have a version number, so I've had to invent a new
4185 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4186 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4187 radiusclient), but it was contributed by a Radius user.
4189 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4190 files or directories, for OpenSSL.
4192 PH/19 When an Exim process that is running as root has to create an Exim log
4193 file, it does so in a subprocess that runs as exim:exim so as to get the
4194 ownership right at creation (otherwise, other Exim processes might see
4195 the file with the wrong ownership). There was no test for failure of this
4196 fork() call, which would lead to the process getting stuck as it waited
4197 for a non-existent subprocess. Forks do occasionally fail when resources
4198 run out. I reviewed all the other calls to fork(); they all seem to check
4201 PH/20 When checking for unexpected SMTP input at connect time (before writing
4202 the banner), Exim was not dealing correctly with a non-positive return
4203 from the read() function. If the client had disconnected by this time,
4204 the result was a log entry for a synchronization error with an empty
4205 string after "input=" when read() returned zero. If read() returned -1
4206 (an event I could not check), uninitialized data bytes were printed.
4207 There were reports of junk text (parts of files, etc) appearing after
4210 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4212 PH/22 Added support for macro redefinition, and (re)definition in between
4213 driver and ACL definitions.
4215 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4216 forgetting to use the resulting value; it was using the unexpanded value.
4218 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4219 hadn't been configured. The fix is from Juergen Kreileder, who
4220 understands it better than I do:
4222 "Here's what I see happening with three configured cyrus_sasl
4223 authenticators configured (plain, login, cram-md5):
4225 On startup auth_cyrus_sasl_init() gets called for each of these.
4226 This means three calls to sasl_listmech() without a specified mech_list.
4227 => SASL tests which mechs of all available mechs actually work
4228 => three warnings about OTP not working
4229 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4231 With the patch, sasl_listmech() also gets called three times. But now
4232 SASL's mech_list option is set to the server_mech specified in the the
4233 authenticator. Or in other words, the answer from sasl_listmech()
4234 gets limited to just the mech you're testing for (which is different
4236 => the return list contains just 'plain' or 'login', 'cram-md5' or
4237 nothing depending on the value of ob->server_mech.
4239 I've just tested the patch: Authentication still works fine,
4240 unavailable mechs specified in the exim configuration are still
4241 caught, and the auth.log warnings about OTP are gone."
4243 PH/25 When debugging is enabled, the contents of the command line are added
4244 to the debugging output, even when log_selector=+arguments is not
4247 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4248 answer is "GNU", and only if the return is "GNU/something" is the answer
4251 PH/27 $acl_verify_message is now set immediately after the failure of a
4252 verification in an ACL, and so is available in subsequent modifiers. In
4253 particular, the message can be preserved by coding like this:
4255 warn !verify = sender
4256 set acl_m0 = $acl_verify_message
4258 Previously, $acl_verify_message was set only while expanding "message"
4259 and "log_message" when a very denied access.
4261 PH/28 Modified OS/os.c-Linux with
4263 -#ifndef OS_LOAD_AVERAGE
4264 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4266 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4267 nomenclature these days.)
4269 PH/29 Installed patch from the Sieve maintainer that adds the options
4270 sieve_useraddress and sieve_subaddress to the redirect router.
4272 PH/30 In these circumstances:
4273 . Two addresses routed to the same list of hosts;
4274 . First host does not offer TLS;
4275 . First host accepts first address;
4276 . First host gives temporary error to second address;
4277 . Second host offers TLS and a TLS session is established;
4278 . Second host accepts second address.
4279 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4280 and peerdn, if requested) that were in fact used only for the second
4283 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4284 attention to any local part prefix or suffix that was matched by the
4285 router that accepted the address. It now behaves in the same way as it
4286 does for delivery: the affixes are removed from the local part unless
4287 rcpt_include_affixes is set on the transport.
4289 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4290 timeout during the DATA phase of an incoming message.
4292 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4293 applied a patch sanctioned by the Sieve maintainer.
4295 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4296 the uid or gid is negative. A case of a negative gid caused this to be
4297 noticed. The fix allows for either to be negative.
4299 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4300 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4303 PH/36 The change PH/12 above was broken. Fixed it.
4305 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4306 the grounds that routing the same address twice would always produce the
4307 same answer. This might have been true once, but it is certainly no
4308 longer true now. Routing a child address may depend on the previous
4309 routing that produced that child. Some complicated redirection strategies
4310 went wrong when messages had multiple recipients, and made Exim's
4311 behaviour dependent on the order in which the addresses were given.
4313 I have moved the duplicate checking until after the routing is complete.
4314 Exim scans the addresses that are assigned to local and remote
4315 transports, and removes any duplicates. This means that more work will be
4316 done, as duplicates will always all be routed, but duplicates are
4317 presumably rare, so I don't expect this is of any significance.
4319 For deliveries to pipes, files, and autoreplies, the duplicate checking
4320 still happens during the routing process, since they are not going to be
4323 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4324 It corrects a timeout issue with spamd. This is Ian's comment: "The
4325 background is that sometimes spamd either never reads data from a
4326 connection it has accepted, or it never writes response data. The exiscan
4327 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4328 blindly assumes that writes won't block so it may never time out."
4330 PH/39 Allow G after quota size as well as K and M.
4332 PH/40 The value set for $authenticated_id in an authenticator may not contain
4333 binary zeroes or newlines because the value is written to log lines and
4334 to spool files. There was no check on this. Now the value is run through
4335 the string_printing() function so that such characters are converted to
4336 printable escape sequences.
4338 PH/41 $message_linecount is a new variable that contains the total number of
4339 lines in the message. Compare $body_linecount, which is the count for the
4342 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4343 bounce and delay warning messages, except in certain special cases, which
4346 (a) An SMTP error message from a remote host;
4347 (b) A message specified in a :fail: redirection;
4348 (c) A message specified in a "fail" command in a system filter;
4349 (d) A message specified in a FAIL return from the queryprogram router;
4350 (e) A message specified by the cannot_route_message router option.
4352 In these cases only, Exim does include the error details in bounce and
4353 warning messages. There are also a few cases where bland messages such
4354 as "unrouteable address" or "local delivery error" are given.
4356 PH/43 $value is now also set for the "else" part of a ${run expansion.
4358 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4359 being worked on, but at least Exim now implements the latest version to
4362 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4363 process to complete was treated as a delivery failure, a timeout while
4364 writing the message to the pipe was logged, but erroneously treated as a
4365 successful delivery. Such timeouts include transport filter timeouts. For
4366 consistency with the overall process timeout, these timeouts are now
4367 treated as errors, giving rise to delivery failures by default. However,
4368 there is now a new Boolean option for the pipe transport called
4369 timeout_defer, which, if set TRUE, converts the failures into defers for
4370 both kinds of timeout. A transport filter timeout is now identified in
4373 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4374 systems where "make" and "gmake" are different, calling "gmake" at top
4375 level broke things. I've arranged for the value of $(MAKE) to be passed
4376 from the Makefile to this script so that it can call the same version of
4380 A note about Exim versions 4.44 and 4.50
4381 ----------------------------------------
4383 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4384 changes of various kinds. As a consequence, a big documentation update was
4385 needed. This delayed the release for rather longer than seemed good, especially
4386 in the light of a couple of (minor) security issues. Therefore, the changes
4387 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4388 release. So 4.44 and 4.50 are in effect two different branches that both start
4391 I have left the 4.50 change log unchanged; it contains all the changes since
4392 4.43. The change log for 4.44 is below; many of its items are identical to
4393 those for 4.50. This seems to be the most sensible way to preserve the
4394 historical information.
4400 1. Minor wording change to the doc/README.SIEVE file.
4402 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4403 computation of the current number of files was incorrect.
4405 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4406 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4409 4. Give more explanation in the error message when the command for a transport
4410 filter fails to execute.
4412 5. There are several places where Exim runs a non-Exim command in a
4413 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4414 was being done only for the command run by the queryprogram router. It is
4415 now done for all such subprocesses. The other cases are: ${run, transport
4416 filters, and the commands run by the lmtp and pipe transports.
4418 6. Added CONFIGURE_GROUP build-time option.
4420 7. Some older OS have a limit of 256 on the maximum number of file
4421 descriptors. Exim was using setrlimit() to set 1000 as a large value
4422 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4423 systems. I've change it so that if it can't get 1000, it tries for 256.
4425 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4426 was an oversight, and furthermore, ever since the addition of extra
4427 controls (e.g. 4.43/32), the checks on when to allow different forms of
4428 "control" were broken. There should now be diagnostics for all cases when a
4429 control that does not make sense is encountered.
4431 9. Added the /retain_sender option to "control=submission".
4433 10. $recipients is now available in the predata ACL (oversight).
4435 11. Tidy the search cache before the fork to do a delivery from a message
4436 received from the command line. Otherwise the child will trigger a lookup
4437 failure and thereby defer the delivery if it tries to use (for example) a
4438 cached ldap connection that the parent has called unbind on.
4440 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4441 of $address_data from the recipient verification was clobbered by the
4442 sender verification.
4444 13. The value of address_data from a sender verification is now available in
4445 $sender_address_data in subsequent conditions in the ACL statement.
4447 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4449 15. Added a new option "connect=<time>" to callout options, to set a different
4452 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4453 was its contents. (It was OK if the option was not defined at all.)
4455 17. A "Completed" log line is now written for messages that are removed from
4456 the spool by the -Mrm option.
4458 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4459 information about exactly what failed.
4461 19. Added -dd to debug only the daemon process.
4463 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4464 handles timeouts, both on the server side and network timeouts. Renamed the
4465 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4467 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4468 It is now set to "smtps".
4470 22. $host_address is now set to the target address during the checking of
4471 ignore_target_hosts.
4473 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4474 being passed; this would have caused $sender_host_name to have been used if
4475 matching the list had actually called for a host name (not very likely,
4476 since this list is usually IP addresses). A host name is now passed as
4479 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4480 code that specifies a non-restarting handler (typically sigaction() in
4481 modern systems) in an attempt to fix a rare and obscure crash bug.
4483 25. Narrowed the window for a race in the daemon that could cause it to ignore
4484 SIGCHLD signals. This is not a major problem, because they are used only to
4485 wake it up if nothing else does.
4487 26. A malformed maildirsize file could cause Exim to calculate negative values
4488 for the mailbox size or file count. Odd effects could occur as a result.
4489 The maildirsize information is now recalculated if the size or filecount
4492 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4493 support for a long time. Removed HAVE_SYS_VFS_H.
4495 28. Installed the latest version of exipick from John Jetmore.
4497 29. In an address list, if the pattern was not a regular expression, an empty
4498 subject address (from a bounce message) matched only if the pattern was an
4499 empty string. Non-empty patterns were not even tested. This was the wrong
4500 because it is perfectly reasonable to use an empty address as part of a
4501 database query. An empty address is now tested by patterns that are
4502 lookups. However, all the other forms of pattern expect the subject to
4503 contain a local part and a domain, and therefore, for them, an empty
4504 address still always fails if the pattern is not itself empty.
4506 30. Exim went into a mad DNS loop when attempting to do a callout where the
4507 host was specified on an smtp transport, and looking it up yielded more
4508 than one IP address.
4510 31. Re-factored the code for checking spool and log partition space into a
4511 function that finds that data and another that does the check. The former
4512 is then used to implement four new variables: $spool_space, $log_space,
4513 $spool_inodes, and $log_inodes.
4515 32. The RFC2047 encoding function was originally intended for short strings
4516 such as real names; it was not keeping to the 75-character limit for
4517 encoded words that the RFC imposes. It now respects the limit, and
4518 generates multiple encoded words if necessary. To be on the safe side, I
4519 have increased the buffer size for the ${rfc2047: expansion operator from
4522 33. It is now permitted to omit both strings after an "if" condition; if the
4523 condition is true, the result is "true". As before, when the second string
4524 is omitted, a false condition yields an empty string. This makes it less
4525 cumbersome to write custom ACL and router conditions.
4527 34. Failure to deliver a bounce message always caused it to be frozen, even if
4528 there was an errors_to setting on the router. The errors_to setting is now
4531 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4532 canonical form (fully expanded) before being placed in
4533 $sender_host_address.
4535 36. The table in the code that translates DNS record types into text (T_A to
4536 "A" for instance) was missing entries for NS and CNAME. It is just possible
4537 that this could have caused confusion if both these types were looked up
4538 for the same domain, because the text type is used as part of Exim's
4539 per-process caching. But the chance of anyone hitting this buglet seems
4542 37. The dnsdb lookup has been extended in a number of ways.
4544 (1) There is a new type, "zns", which walks up the domain tree until it
4545 finds some nameserver records. It should be used with care.
4547 (2) There is a new type, "mxh", which is like "mx" except that it returns
4548 just the host names, not the priorities.
4550 (3) It is now possible to give a list of domains (or IP addresses) to be
4551 looked up. The behaviour when one of the lookups defers can be
4552 controlled by a keyword.
4554 (4) It is now possible to specify the separator character for use when
4555 multiple records are returned.
4557 38. The dnslists ACL condition has been extended: it is now possible to supply
4558 a list of IP addresses and/or domains to be looked up in a particular DNS
4561 39. Added log_selector=+queue_time_overall.
4563 40. When running the queue in the test harness, wait just a tad after forking a
4564 delivery process, to get repeatability of debugging output.
4566 41. Include certificate and key file names in error message when GnuTLS fails
4567 to set them up, because the GnuTLS error message doesn't include the name
4568 of the failing file when there is a problem reading it.
4570 42. Allow both -bf and -bF in the same test run.
4572 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4574 44. The "Exiscan patch" is now merged into the mainline Exim source.
4576 45. Sometimes the final signoff response after QUIT could fail to get
4577 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4578 < 0 before doing a fflush(). This bug looks as though it goes back to the
4579 introduction of TLS in release 3.20, but "sometimes" must have been rare
4580 because the tests only now provoked it.
4582 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4583 (this can affect the format of dates).
4585 47. exim_tidydb, when checking for the continued existence of a message for
4586 which it has found a message-specific retry record, was not finding
4587 messages that were in split spool directories. Consequently, it was
4588 deleting retry records that should have stayed in existence.
4590 48. Steve fixed some bugs in eximstats.
4592 49. The SPA authentication driver was not abandoning authentication and moving
4593 on to the next authenticator when an expansion was forced to fail,
4594 contradicting the general specification for all authenticators. Instead it
4595 was generating a temporary error. It now behaves as specified.
4597 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4598 (the order specifies the preference for clients). The order is now AES256,
4599 AES128, 3DES, ARCFOUR128.
4601 51. Small patch to Sieve code - explicitly set From: when generating an
4604 52. Exim crashed if a remote delivery caused a very long error message to be
4605 recorded - for instance if somebody sent an entire SpamAssassin report back
4606 as a large number of 550 error lines. This bug was coincidentally fixed by
4607 increasing the size of one of Exim's internal buffers (big_buffer) that
4608 happened as part of the Exiscan merge. However, to be on the safe side, I
4609 have made the code more robust (and fixed the comments that describe what
4612 53. Now that there can be additional text after "Completed" in log lines (if
4613 the queue_time_overall log selector is set), a one-byte patch to exigrep
4614 was needed to allow it to recognize "Completed" as not the last thing in
4617 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4618 patch that reportedly fixes this has been added. I am not expert enough to
4619 create a test for it. This is what the patch creator wrote:
4621 "I found a little strange behaviour of ldap code when working with
4622 Windows 2003 AD Domain, where users was placed in more than one
4623 Organization Units. When I tried to give exim partial DN, the exit code
4624 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4625 But simultaneously result of request was absolutely normal ldap result,
4626 so I produce this patch..."
4628 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4629 REFERENCE, so I have modified the code to exclude the patch when that macro
4632 55. Some experimental protocols are using DNS PTR records for new purposes. The
4633 keys for these records are domain names, not reversed IP addresses. The
4634 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4635 leaves it alone. Component reversal etc. now happens only for IP addresses.
4638 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4640 57. Double the size of the debug message buffer (to 2048) so that more of very
4641 long debug lines gets shown.
4643 58. The exicyclog utility now does better if the number of log files to keep
4644 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4646 59. Two changes related to the smtp_active_hostname option:
4648 (1) $smtp_active_hostname is now available as a variable.
4649 (2) The default for smtp_banner uses $smtp_active_hostname instead
4650 of $primary_hostname.
4652 60. The host_aton() function is supposed to be passed a string that is known
4653 to be a valid IP address. However, in the case of IPv6 addresses, it was
4654 not checking this. This is a hostage to fortune. Exim now panics and dies
4655 if the condition is not met. A case was found where this could be provoked
4656 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4657 components; fortuitously, this particular loophole had already been fixed
4658 by change 4.50/55 above.
4660 If there are any other similar loopholes, the new check in host_aton()
4661 itself should stop them being exploited. The report I received stated that
4662 data on the command line could provoke the exploit when Exim was running as
4663 exim, but did not say which command line option was involved. All I could
4664 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4665 running as the user.
4668 61. There was a buffer overflow vulnerability in the SPA authentication code
4669 (which came originally from the Samba project). I have added a test to the
4670 spa_base64_to_bits() function which I hope fixes it.
4673 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4674 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4676 63. The daemon start-up calls getloadavg() while still root for those OS that
4677 need the first call to be done as root, but it missed one case: when
4678 deliver_queue_load_max is set with deliver_drop_privilege. This is
4679 necessary for the benefit of the queue runner, because there is no re-exec
4680 when deliver_drop_privilege is set.
4682 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4683 This has been fixed.
4685 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4686 was in use, was not putting the data itself into the right store pool;
4687 consequently, it could be overwritten for a subsequent message in the same
4688 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4691 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4693 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4694 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4695 the function were treating the return as a boolean value, which happened to
4696 work because 0=false and not-0=true, but is not correct code.
4698 68. The host_aton() function was not handling scoped IPv6 addresses (those
4699 with, for example, "%eth0" on the end) correctly.
4701 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4702 negated items (that is, ~something) in unsigned ints. Some compilers
4703 apparently mutter when there is no cast.
4705 70. If an address verification called from an ACL failed, and did not produce a
4706 user-specific message (i.e. there was only a "system" message), nothing was
4707 put in $acl_verify_message. In this situation, it now puts the system
4710 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4711 session; change 4.31/43 added the unwanted input to the log line - except
4712 that it did not do this in the start of session case. It now does.
4714 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4715 This is wrong and can cause the other end to generate a synchronization
4716 error if it is another Exim or anything else that does the synchronization
4717 check. A QUIT command is no longer sent after a timeout.
4719 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4720 during host lookups.
4722 74. The defer_ok option of callout verification was not working if it was used
4723 when verifying addresses in header lines, that is, for this case:
4725 verify = header_sender/callout=defer_ok
4727 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4728 those file descriptors could be used for SMTP connections. If anything
4729 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4730 could be sent to the SMTP client, causing chaos. The daemon now opens
4731 stdin, stdout, and stderr to /dev/null when it puts itself into the
4734 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4735 log by default. The user can override this with suitable Perl magic.
4737 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4738 the log message when discard triggers, was not working for the DATA ACL or
4739 for the non-SMTP ACL.
4741 78. Error message wording change in sieve.c.
4743 79. If smtp_accept_max_per_host was set, the number of connections could be
4744 restricted to fewer than expected, because the daemon was trying to set up
4745 a new connection before checking whether the processes handling previous
4746 connections had finished. The check for completed processes is now done
4747 earlier. On busy systems, this bug wouldn't be noticed because something
4748 else would have woken the daemon, and it would have reaped the completed
4751 80. If a message was submitted locally by a user whose login name contained one
4752 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4753 It caused a spool format error. I have fixed the spool reading code. A
4754 related problem was that the "from" clause in the Received: line became
4755 illegal because of the space(s). It is now covered by ${quote_local_part.
4757 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4760 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4762 83. Patch from Sieve maintainer to fix unterminated string problem in
4763 "vacation" handling.
4765 84. Some minor changes to the Linux configuration files to help with other
4766 OS variants using glibc.
4768 85. One more patch for Sieve to update vacation handling to latest spec.
4771 ----------------------------------------------------
4772 See the note above about the 4.44 and 4.50 releases.
4773 ----------------------------------------------------
4779 1. Change 4.43/35 introduced a bug that caused file counts to be
4780 incorrectly computed when quota_filecount was set in an appendfile
4783 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4784 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4787 3. Give more explanation in the error message when the command for a transport
4788 filter fails to execute.
4790 4. There are several places where Exim runs a non-Exim command in a
4791 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4792 was being done only for the command run by the queryprogram router. It is
4793 now done for all such subprocesses. The other cases are: ${run, transport
4794 filters, and the commands run by the lmtp and pipe transports.
4796 5. Some older OS have a limit of 256 on the maximum number of file
4797 descriptors. Exim was using setrlimit() to set 1000 as a large value
4798 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4799 systems. I've change it so that if it can't get 1000, it tries for 256.
4801 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4802 was an oversight, and furthermore, ever since the addition of extra
4803 controls (e.g. 4.43/32), the checks on when to allow different forms of
4804 "control" were broken. There should now be diagnostics for all cases when a
4805 control that does not make sense is encountered.
4807 7. $recipients is now available in the predata ACL (oversight).
4809 8. Tidy the search cache before the fork to do a delivery from a message
4810 received from the command line. Otherwise the child will trigger a lookup
4811 failure and thereby defer the delivery if it tries to use (for example) a
4812 cached ldap connection that the parent has called unbind on.
4814 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4815 of $address_data from the recipient verification was clobbered by the
4816 sender verification.
4818 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4819 was its contents. (It was OK if the option was not defined at all.)
4821 11. A "Completed" log line is now written for messages that are removed from
4822 the spool by the -Mrm option.
4824 12. $host_address is now set to the target address during the checking of
4825 ignore_target_hosts.
4827 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4828 being passed; this would have caused $sender_host_name to have been used if
4829 matching the list had actually called for a host name (not very likely,
4830 since this list is usually IP addresses). A host name is now passed as
4833 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4834 code that specifies a non-restarting handler (typically sigaction() in
4835 modern systems) in an attempt to fix a rare and obscure crash bug.
4837 15. Narrowed the window for a race in the daemon that could cause it to ignore
4838 SIGCHLD signals. This is not a major problem, because they are used only to
4839 wake it up if nothing else does.
4841 16. A malformed maildirsize file could cause Exim to calculate negative values
4842 for the mailbox size or file count. Odd effects could occur as a result.
4843 The maildirsize information is now recalculated if the size or filecount
4846 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4847 support for a long time. Removed HAVE_SYS_VFS_H.
4849 18. Updated exipick to current release from John Jetmore.
4851 19. Allow an empty sender to be matched against a lookup in an address list.
4852 Previously the only cases considered were a regular expression, or an
4855 20. Exim went into a mad DNS lookup loop when doing a callout where the
4856 host was specified on the transport, if the DNS lookup yielded more than
4859 21. The RFC2047 encoding function was originally intended for short strings
4860 such as real names; it was not keeping to the 75-character limit for
4861 encoded words that the RFC imposes. It now respects the limit, and
4862 generates multiple encoded words if necessary. To be on the safe side, I
4863 have increased the buffer size for the ${rfc2047: expansion operator from
4866 22. Failure to deliver a bounce message always caused it to be frozen, even if
4867 there was an errors_to setting on the router. The errors_to setting is now
4870 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4871 canonical form (fully expanded) before being placed in
4872 $sender_host_address.
4874 24. Updated eximstats to version 1.33
4876 25. Include certificate and key file names in error message when GnuTLS fails
4877 to set them up, because the GnuTLS error message doesn't include the name
4878 of the failing file when there is a problem reading it.
4880 26. Expand error message when OpenSSL has problems setting up cert/key files.
4883 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4884 (this can affect the format of dates).
4886 28. exim_tidydb, when checking for the continued existence of a message for
4887 which it has found a message-specific retry record, was not finding
4888 messages that were in split spool directories. Consequently, it was
4889 deleting retry records that should have stayed in existence.
4891 29. eximstats updated to version 1.35
4892 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4893 1.35 - bugfix such that pie charts by volume are generated correctly
4895 30. The SPA authentication driver was not abandoning authentication and moving
4896 on to the next authenticator when an expansion was forced to fail,
4897 contradicting the general specification for all authenticators. Instead it
4898 was generating a temporary error. It now behaves as specified.
4900 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4901 (the order specifies the preference for clients). The order is now AES256,
4902 AES128, 3DES, ARCFOUR128.
4904 31. Small patch to Sieve code - explicitly set From: when generating an
4907 32. Exim crashed if a remote delivery caused a very long error message to be
4908 recorded - for instance if somebody sent an entire SpamAssassin report back
4909 as a large number of 550 error lines. This bug was coincidentally fixed by
4910 increasing the size of one of Exim's internal buffers (big_buffer) that
4911 happened as part of the Exiscan merge. However, to be on the safe side, I
4912 have made the code more robust (and fixed the comments that describe what
4915 33. Some experimental protocols are using DNS PTR records for new purposes. The
4916 keys for these records are domain names, not reversed IP addresses. The
4917 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4918 leaves it alone. Component reversal etc. now happens only for IP addresses.
4921 34. The host_aton() function is supposed to be passed a string that is known
4922 to be a valid IP address. However, in the case of IPv6 addresses, it was
4923 not checking this. This is a hostage to fortune. Exim now panics and dies
4924 if the condition is not met. A case was found where this could be provoked
4925 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4926 components; fortuitously, this particular loophole had already been fixed
4927 by change 4.50/55 or 4.44/33 above.
4929 If there are any other similar loopholes, the new check in host_aton()
4930 itself should stop them being exploited. The report I received stated that
4931 data on the command line could provoke the exploit when Exim was running as
4932 exim, but did not say which command line option was involved. All I could
4933 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4934 running as the user.
4937 35. There was a buffer overflow vulnerability in the SPA authentication code
4938 (which came originally from the Samba project). I have added a test to the
4939 spa_base64_to_bits() function which I hope fixes it.
4942 36. The daemon start-up calls getloadavg() while still root for those OS that
4943 need the first call to be done as root, but it missed one case: when
4944 deliver_queue_load_max is set with deliver_drop_privilege. This is
4945 necessary for the benefit of the queue runner, because there is no re-exec
4946 when deliver_drop_privilege is set.
4948 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4949 was in use, was not putting the data itself into the right store pool;
4950 consequently, it could be overwritten for a subsequent message in the same
4951 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4954 38. Sometimes the final signoff response after QUIT could fail to get
4955 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4956 < 0 before doing a fflush(). This bug looks as though it goes back to the
4957 introduction of TLS in release 3.20, but "sometimes" must have been rare
4958 because the tests only now provoked it.
4964 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4965 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4966 Now it returns an integer. A number of calls were still expecting a T/F
4967 return. Fortuitously, in all cases, the tests worked in OK situations,
4968 which is the norm. However, things would have gone wrong on any write
4969 failures on the smtp file descriptor. This function is used when sending
4970 messages over SMTP and also when doing verify callouts.
4972 2. When Exim is called to do synchronous delivery of a locally submitted
4973 message (the -odf or -odi options), it no longer closes stderr before doing
4976 3. Implemented the mua_wrapper option.
4978 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4980 5. Implemented the functions header_remove(), header_testname(),
4981 header_add_at_position(), and receive_remove_recipient(), and exported them
4984 6. If an ACL "warn" statement specified the addition of headers, Exim already
4985 inserted X-ACL-Warn: at the start if there was no header name. However, it
4986 was not making this test for the second and subsequent header lines if
4987 there were newlines in the string. This meant that an invalid header could
4988 be inserted if Exim was badly configured.
4990 7. Allow an ACL "warn" statement to add header lines at the start or after all
4991 the Received: headers, as well as at the end.
4993 8. Added the rcpt_4xx retry error code.
4995 9. Added postmaster_mailfrom=xxx to callout verification option.
4997 10. Added mailfrom=xxxx to the callout verification option, for verify=
5000 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5001 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5003 12. Inserted some casts to stop certain compilers warning when using pointer
5004 differences as field lengths or precisions in printf-type calls (mostly
5005 affecting debugging statements).
5007 13. Added optional readline() support for -be (dynamically loaded).
5009 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5010 same clock tick as a message's arrival, so that its received time was the
5011 same as the "first fail" time on the retry record, and that message
5012 remained on the queue past the ultimate address timeout, every queue runner
5013 would try a delivery (because it was past the ultimate address timeout) but
5014 after another failure, the ultimate address timeout, which should have then
5015 bounced the address, did not kick in. This was a "< instead of <=" error;
5016 in most cases the first failure would have been in the next clock tick
5017 after the received time, and all would be well.
5019 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5020 being recognized when the domain list was tested by the match_domain
5021 condition in an expansion string.
5023 16. Added the ${str2b64: operator.
5025 17. Exim was always calling setrlimit() to set a large limit for the number of
5026 processes, without checking whether the existing limit was already
5027 adequate. (It did check for the limit on file descriptors.) Furthermore,
5028 errors from getrlimit() and setrlimit() were being ignored. Now they are
5029 logged to the main and panic logs, but Exim does carry on, to try to do its
5030 job under whatever limits there are.
5032 18. Imported PCRE 5.0.
5034 19. Trivial typo in log message " temporarily refused connection" (the leading
5037 20. If the log selector return_path_on_delivery was set and an address was
5038 redirected to /dev/null, the delivery process crashed because it assumed
5039 that a return path would always be set for a "successful" delivery. In this
5040 case, the whole delivery is bypassed as an optimization, and therefore no
5043 21. Internal re-arrangement: the function for sending a challenge and reading
5044 a response while authentication was assuming a zero-terminated challenge
5045 string. It's now changed to take a pointer and a length, to allow for
5046 binary data in such strings.
5048 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5050 23. Exim was not respecting finduser_retries when seeking the login of the
5051 uid under which it was called; it was always trying 10 times. (The default
5052 setting of finduser_retries is zero.) Also, it was sleeping after the final
5053 failure, which is pointless.
5055 24. Implemented tls_on_connect_ports.
5057 25. Implemented acl_smtp_predata.
5059 26. If the domain in control=submission is set empty, Exim assumes that the
5060 authenticated id is a complete email address when it generates From: or
5061 Sender: header lines.
5063 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5064 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5065 chown and chgrp in /bin and hostname in /usr/bin.
5067 28. Exim was keeping the "process log" file open after each use, just as it
5068 does for the main log. This opens the possibility of it remaining open for
5069 long periods when the USR1 signal hits a daemon. Occasional processlog
5070 errors were reported, that could have been caused by this. Anyway, it seems
5071 much more sensible not to leave this file open at all, so that is what now
5074 29. The long-running daemon process does not normally write to the log once it
5075 has entered its main loop, and it closes the log before doing so. This is
5076 so that log files can straightforwardly be renamed and moved. However,
5077 there are a couple of unusual error situations where the daemon does write
5078 log entries, and I had neglected to close the log afterwards.
5080 30. The text of an SMTP error response that was received during a remote
5081 delivery was being truncated at 512 bytes. This is too short for some of
5082 the long messages that one sometimes sees. I've increased the limit to
5085 31. It is now possible to make retry rules that apply only when a message has a
5086 specific sender, in particular, an empty sender.
5088 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5089 it possible to be selective about when SMTP synchronization is enforced.
5091 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5093 32. Implemented hosts_connection_nolog.
5095 33. Added an ACL for QUIT.
5097 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5100 35. Added mailbox_size and mailbox_filecount to appendfile.
5102 36. Added control = no_multiline_responses to ACLs.
5104 37. There was a bug in the logic of the code that waits for the clock to tick
5105 in the case where the clock went backwards by a substantial amount such
5106 that the microsecond fraction of "now" was more than the microsecond
5107 fraction of "then" (but the whole seconds number was less).
5109 38. Added support for the libradius Radius client library this is found on
5110 FreeBSD (previously only the radiusclient library was supported).
5116 1. When certain lookups returned multiple values in the form name=value, the
5117 quoting of the values was not always being done properly. Specifically:
5118 (a) If the value started with a double quote, but contained no whitespace,
5120 (b) If the value contained whitespace other than a space character (i.e.
5121 tabs or newlines or carriage returns) it was not quoted.
5122 This fix has been applied to the mysql and pgsql lookups by writing a
5123 separate quoting function and calling it from the lookup code. The fix
5124 should probably also be applied to nisplus, ibase and oracle lookups, but
5125 since I cannot test any of those, I have not disturbed their existing code.
5127 2. A hit in the callout cache for a specific address caused a log line with no
5128 reason for rejecting RCPT. Now it says "Previous (cached) callout
5129 verification failure".
5131 3. There was an off-by-one bug in the queryprogram router. An over-long
5132 return line was truncated at 256 instead of 255 characters, thereby
5133 overflowing its buffer with the terminating zero. As well as fixing this, I
5134 have increased the buffer size to 1024 (and made a note to document this).
5136 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5137 when Exim is waiting for an SMTP response from a remote server, Exim
5138 restarts its select() call on the socket, thereby resetting its timeout.
5139 This is not a problem when such interrupts are rare. Somebody set up a cron
5140 job to run exiwhat every 2 minutes, which is less than the normal select()
5141 timeout (5 or 10 minutes). This meant that the select() timeout never
5142 kicked in because it was always reset. I have fixed this by comparing the
5143 time when an interrupt arrives with the time at the start of the first call
5144 to select(). If more time than the timeout has elapsed, the interrupt is
5145 treated as a timeout.
5147 5. Some internal re-factoring in preparation for the addition of Sieve
5148 extensions (by MH). In particular, the "personal" test is moved to a
5149 separate function, and given an option for scanning Cc: and Bcc: (which is
5150 not set for Exim filters).
5152 6. When Exim created an email address using the login of the caller as the
5153 local part (e.g. when creating a From: or Sender: header line), it was not
5154 quoting the local part when it contained special characters such as @.
5156 7. Installed new OpenBSD configuration files.
5158 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5159 try to make them clearer.
5161 9. Callout options, other than the timeout value, were being ignored when
5162 verifying sender addresses in header lines. For example, when using
5164 verify = header_sender/callout=no_cache
5166 the cache was (incorrectly) being used.
5168 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5169 only the Cygwin environment.
5171 11. When return_path_on_delivery was set as a log selector, if different remote
5172 addresses in the same message used different return paths and parallel
5173 remote delivery occurred, the wrong values would sometimes be logged.
5174 (Whenever a remote delivery process finished, the return path value from
5175 the most recently started remote delivery process was logged.)
5177 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5178 lines when AUTH and/or TLS are in use. This is the "received protocol"
5179 field. Exim used to use "asmtp" for authenticated SMTP, without any
5180 indication (in the protocol name) for TLS use. Now it follows the RFC and
5181 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5182 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5183 names appear in log lines as well as in Received: header lines.
5185 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5186 extensions, and comparison tests, and to fix some bugs.
5188 14. Changes to the "personal" filter test:
5190 (1) The test was buggy in that it was just doing the equivalent of
5191 "contains" tests on header lines. For example, if a user's address was
5192 anne@some.where, the "personal" test would incorrectly be true for
5194 To: susanne@some.where
5196 This test is now done by extracting each address from the header in turn,
5197 and checking the entire address. Other tests that are part of "personal"
5198 are now done using regular expressions (for example, to check local parts
5199 of addresses in From: header lines).
5201 (2) The list of non-personal local parts in From: addresses has been
5202 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5203 taken from the Sieve specification recommendations.
5205 (3) If the message contains any header line starting with "List-" it is
5206 treated as non-personal.
5208 (4) The test for "circular" in the Subject: header line has been removed
5209 because it now seems ill-conceived.
5211 15. Minor typos in src/EDITME comments corrected.
5213 16. Installed latest exipick from John Jetmore.
5215 17. If headers_add on a router specified a text string that was too long for
5216 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5217 of string_sprintf() is now avoided.
5219 18. $message_body_size was not set (it was always zero) when running the DATA
5220 ACL and the local_scan() function.
5222 19. For the "mail" command in an Exim filter, no default was being set for
5223 the once_repeat time, causing a random time value to be used if "once" was
5224 specified. (If the value happened to be <= 0, no repeat happened.) The
5225 default is now 0s, meaning "never repeat". The "vacation" command was OK
5226 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5227 (I found it when inspecting the code).
5229 20. There is now an overall timeout for performing a callout verification. It
5230 defaults to 4 times the callout timeout, which applies to individual SMTP
5231 commands during the callout. The overall timeout applies when there is more
5232 than one host that can be tried. The timeout is checked before trying the
5233 next host. This prevents very long delays if there are a large number of
5234 hosts and all are timing out (e.g. when the network connections are timing
5235 out). The value of the overall timeout can be changed by specifying an
5236 additional sub-option for "callout", called "maxwait". For example:
5238 verify = sender/callout=5s,maxwait=20s
5240 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5241 to the end before writing, but this should make it even safer).
5243 22. Exim was forgetting that it had advertised PIPELINING for the second and
5244 subsequent messages on an SMTP connection. It was also not resetting its
5245 memory on STARTTLS and an internal HELO.
5247 23. When Exim logs an SMTP synchronization error within a session, it now
5248 records whether PIPELINING has been advertised or not.
5250 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5251 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5252 rather than long int.
5254 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5256 26. Added the never_mail option to autoreply.
5262 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5263 crash if the getsockname() call failed; this can happen if a connection is
5264 closed very soon after it is established. The problem was simply in the
5265 order in which certain operations were done, causing Exim to try to write
5266 to the SMTP stream before it had set up the file descriptor. The bug has
5267 been fixed by making things happen in the correct order.
5273 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5274 before the connection was closed, thus losing the rejection response.
5276 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5277 some early Solaris releases, but causes trouble in current releases where
5278 socklen_t is defined.
5280 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5283 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5286 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5287 that releasing the top of it at the end releases what was used for sub-
5288 expansions (unless the block got too big). However, discard this block if
5289 the first thing is a variable or header, so that we can use its block when
5290 it is dynamic (useful for very large $message_headers, for example).
5292 6. Lookups now cache *every* query, not just the most recent. A new, separate
5293 store pool is used for this. It can be recovered when all lookup caches are
5294 flushed. Lookups now release memory at the end of their result strings.
5295 This has involved some general refactoring of the lookup sources.
5297 7. Some code has been added to the store_xxx() functions to reduce the amount
5298 of flapping under certain conditions.
5300 8. log_incoming_interface used to affect only the <= reception log lines. Now
5301 it causes the local interface and port to be added to several more SMTP log
5302 lines, for example "SMTP connection from", and rejection lines.
5304 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5306 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5308 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5309 could be overwritten at the end of the current message (or the start of a
5310 new message if it was set in a HELO ACL). The value is now preserved for
5311 the duration of the SMTP connection.
5313 12. If a transport had a headers_rewrite setting, and a matching header line
5314 contained an unqualified address, that address was qualified, even if it
5315 did not match any rewriting rules. The underlying bug was that the values
5316 of the flags that permit the existence of unqualified sender and recipient
5317 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5318 non-local messages, and by -bnq for local messages) were not being
5319 preserved with the message after it was received.
5321 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5322 "next input=" as part of the text comprising the host identity instead of
5323 the correct text. The code was using the same buffer for two different
5324 strings. However, depending on which order the printing function evaluated
5325 its arguments, the bug did not always show up. Under Linux, for example, my
5326 test suite worked just fine.
5328 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5329 doesn't work with some older versions of Perl. It has been changed to "my",
5330 which in any case is probably the better facility to use.
5332 15. A really picky compiler found some instances of statements for creating
5333 error messages that either had too many or two few arguments for the format
5336 16. The size of the buffer for calls to the DNS resolver has been increased
5337 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5338 for addresses that have a lot of PTR records. This alleviates a problem; it
5339 does not fully solve it.
5341 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5342 buffer now truncates the list and logs the incident, which is the same
5343 action as happens when Exim is looking up a host name and its aliases.
5344 Previously in this situation something unpredictable would happen;
5345 sometimes it was "internal error: store_reset failed".
5347 18. If a server dropped the connection unexpectedly when an Exim client was
5348 using GnuTLS and trying to read a response, the client delivery process
5349 crashed while trying to generate an error log message.
5351 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5352 string, for example:
5354 warn message = H1: something\nH2: something
5356 the text was added as a single header line from Exim's point of view
5357 though it ended up OK in the delivered message. However, searching for the
5358 second and subsequent header lines using $h_h2: did not work. This has been
5359 fixed. Similarly, if a system filter added multiple headers in this way,
5360 the routers could not see them.
5362 20. Expanded the error message when iplsearch is called with an invalid key to
5363 suggest using net-iplsearch in a host list.
5365 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5366 ACLs are no longer actually imposed (and a message to that effect is
5369 22. If a "gecos" field in a passwd entry contained escaped characters, in
5370 particular, if it contained a \" sequence, Exim got it wrong when building
5371 a From: or a Sender: header from that name. A second bug also caused
5372 incorrect handling when an unquoted " was present following a character
5373 that needed quoting.
5375 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5376 was not being matched caselessly.
5378 24. Arranged for all hyphens in the exim.8 source to be escaped with
5381 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5382 properly. Recipient callout cache records were still being keyed to include
5383 the sender, even when use_sender was set false. This led to far more
5384 callouts that were necessary. The sender is no longer included in the key
5385 when use_sender is false.
5387 26. Added "control = submission" modifier to ACLs.
5389 27. Added the ${base62d: operator to decode base 62 numbers.
5391 28. dnsdb lookups can now access SRV records.
5393 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5394 the configuration file.
5396 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5397 (-v) mode. This makes the output for a verbose queue run more intelligible.
5399 31. Added a use_postmaster feature to recipient callouts.
5401 32. Added the $body_zerocount variable, containing the number of binary zero
5402 bytes in the message body.
5404 33. The time of last modification of the "new" subdirectory is now used as the
5405 "mailbox time last read" when there is a quota error for a maildir
5408 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5410 35. Added +ignore_unknown as a special item in host lists.
5412 36. Code for decoding IPv6 addresses in host lists is now included, even if
5413 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5414 address was recognized as an IP address, but was then not correctly decoded
5415 into binary, causing unexpected and incorrect effects when compared with
5422 1. Very minor rewording of debugging text in manualroute to say "list of
5423 hosts" instead of "hostlist".
5425 2. If verify=header_syntax was set, and a header line with an unqualified
5426 address (no domain) and a large number of spaces between the end of the
5427 name and the colon was received, the reception process suffered a buffer
5428 overflow, and (when I tested it) crashed. This was caused by some obsolete
5429 code that should have been removed. The fix is to remove it!
5431 3. When running in the test harness, delay a bit after writing a bounce
5432 message to get a bit more predictability in the log output.
5434 4. Added a call to search_tidyup() just before forking a reception process. In
5435 theory, someone could use a lookup in the expansion of smtp_accept_max_
5436 per_host which, without the tidyup, could leave open a database connection.
5438 5. Added the variables $recipient_data and $sender_data which get set from a
5439 lookup success in an ACL "recipients" or "senders" condition, or a router
5440 "senders" option, similar to $domain_data and $local_part_data.
5442 6. Moved the writing of debug_print from before to after the "senders" test
5445 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5446 problems for message scanning, either using a data ACL, or using
5447 local_scan() because the Received: header was not generated till after they
5448 were called (in order to set the time as the time of reception completion).
5449 I have revised the way this works. The header is now generated after the
5450 body is received, but before the ACL or local_scan() are called. After they
5451 are run, the timestamp in the header is updated.
5457 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5458 before starting a queue runner without re-exec. This happened only when
5459 deliver_drop_privilege was set or when the Exim user was set to root. The
5460 effect of the bug was that timeouts during subsequent deliveries caused
5461 crashes instead of being properly handled. The handler is now left at its
5462 default (and expected) setting.
5464 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5465 message, again when deliver_drop_privilege is set or Exim is run as root.
5466 The bug described in (1) was not present in this case, but the tidying up
5467 of the other signals was missing. I have made the two cases consistent.
5469 3. The ignore_target_hosts setting on a manualroute router was being ignored
5470 for hosts that were looked up using the /MX notation.
5472 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5475 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5476 operated on the sender address. After changing the $sender_address to <>
5477 for the sender address verify, Exim was re-instated it as the original
5478 (before rewriting) address, but remembering that it had rewritten it, so it
5479 wasn't rewriting it again. This bug also had the effect of breaking the
5480 sender address verification caching when the sender address was rewritten.
5482 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5483 This has been changed so that if the ip literal address matches
5484 ignore_target_hosts, the router declines.
5486 7. Added expansion conditions match_domain, match_address, and match_local_
5487 part (NOT match_host).
5489 8. The placeholder for the Received: header didn't have a length field set.
5491 9. Added code to Exim itself and to exim_lock to test for a specific race
5492 condition that could lead to file corruption when using MBX delivery. The
5493 issue is with the lockfile that is created in /tmp. If this file is removed
5494 after a process has opened it but before that process has acquired a lock,
5495 there is the potential for a second process to recreate the file and also
5496 acquire a lock. This could lead to two Exim processes writing to the file
5497 at the same time. The added code performs the same test as UW imapd; it
5498 checks after acquiring the lock that its file descriptor still refers to
5499 the same named file.
5501 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5502 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5505 11. Added the smtp_active_hostname option. If used, this will typically be made
5506 to depend on the incoming interface address. Because $interface_address is
5507 not set up until the daemon has forked a reception process, error responses
5508 that can happen earlier (such as "too many connections") no longer contain
5511 12. If an expansion in a condition on a "warn" statement fails because a lookup
5512 defers, the "warn" statement is abandoned, and the next ACL statement is
5513 processed. Previously this caused the whole ACL to be aborted.
5515 13. Added the iplsearch lookup type.
5517 14. Added ident_timeout as a log selector.
5519 15. Added tls_certificate_verified as a log selector.
5521 16. Added a global option tls_require_ciphers (compare the smtp transport
5522 option of the same name). This controls incoming TLS connections.
5524 17. I finally figured out how to make tls_require_ciphers do a similar thing
5525 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5526 before starting the TLS session.
5528 18. Tabs are now shown as \t in -bP output.
5530 19. If the log selector return_path_on_delivery was set, Exim crashed when
5531 bouncing a message because it had too many Received: header lines.
5533 20. If two routers both had headers_remove settings, and the first one included
5534 a superfluous trailing colon, the final name in the first list and the
5535 first name in the second list were incorrectly joined into one item (with a
5536 colon in the middle).
5542 1. Added -C and -D options to the exinext utility, mainly to make it easier
5543 to include in the automated testing, but these could be helpful when
5544 multiple configurations are in use.
5546 2. The exinext utility was not formatting the output nicely when there was
5547 an alternate port involved in the retry record key, nor when there was a
5548 message id as well (for retries that were specific to a specific message
5549 and a specific host). It was also confused by IPv6 addresses, because of
5550 the additional colons they contain. I have fixed the IPv4 problem, and
5551 patched it up to do a reasonable job for IPv6.
5553 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5554 delivery, the log line now contains "pipelined" if PIPELINING was used.
5556 4. An SMTP transport process used to panic and die if the bind() call to set
5557 an explicit outgoing interface failed. This has been changed; it is now
5558 treated in the same way as a connect() failure.
5560 5. A reference to $sender_host_name in the part of a conditional expansion
5561 that was being skipped was still causing a DNS lookup. This no longer
5564 6. The def: expansion condition was not recognizing references to header lines
5565 that used bh_ and bheader_.
5567 7. Added the _cache feature to named lists.
5569 8. The code for checking quota_filecount in the appendfile transport was
5570 allowing one more file than it should have been.
5572 9. For compatibility with Sendmail, the command line option
5580 and sets the incoming protocol and host name (for trusted callers). The
5581 host name and its colon can be omitted when only the protocol is to be set.
5582 Note the Exim already has two private options, -pd and -ps, that refer to
5583 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5584 "s", but I don't think that's a major issue.
5586 10. A number of refactoring changes to the code, none of which should affect
5589 (a) The number of logging options was getting close to filling up the
5590 32-bit word that was used as a bit map. I have split them into two classes:
5591 those that are passed in the argument to log_write(), and those that are
5592 only ever tested independently outside of that function. These are now in
5593 separate 32-bit words, so there is plenty of room for expansion again.
5594 There is no change in the user interface or the logging behaviour.
5596 (b) When building, for example, log lines, the code previously used a
5597 macro that called string_cat() twice, in order to add two strings. This is
5598 not really sufficiently general. Furthermore, there was one instance where
5599 it was actually wrong because one of the argument was used twice, and in
5600 one call a function was used. (As it happened, calling the function twice
5601 did not affect the overall behaviour.) The macro has been replaced by a
5602 function that can join an arbitrary number of extra strings onto a growing
5605 (c) The code for expansion conditions now uses a table and a binary chop
5606 instead of a serial search (which was left over from when there were very
5607 few conditions). Also, it now recognizes conditions like "pam" even when
5608 the relevant support is not compiled in: a suitably worded error message is
5609 given if an attempt is made to use such a condition.
5611 11. Added ${time_interval:xxxxx}.
5613 12. A bug was causing one of the ddress fields not to be passed back correctly
5614 from remote delivery subprocesses. The field in question was not being
5615 subsequently used, so this caused to problems in practice.
5617 13. Added new log selectors queue_time and deliver_time.
5619 14. Might have fixed a bug in maildirsizefile handling that threw up
5620 "unexpected character" debug warnings, and recalculated the data
5621 unnecessarily. In any case, I expanded the warning message to give more
5624 15. Added the message "Restricted characters in address" to the statements in
5625 the default ACL that block characters like @ and % in local parts.
5627 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5628 Three changes have been made:
5630 (a) There was a serious bug; a negative response to MAIL caused the whole
5631 recipient domain to be cached as invalid, thereby blocking all messages
5632 to all local parts at the same domain, from all senders. This bug has
5633 been fixed. The domain is no longer cached after a negative response to
5634 MAIL if the sender used is not empty.
5636 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5639 (c) A new callout option, "use_sender" has been added for people who want
5640 the modified behaviour.
5646 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5649 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5650 indeed breaks things for older releases.
5652 3. Added additional logging to the case where there is a problem reading data
5653 from a filter that is running in a subprocess using a pipe, in order to
5654 try to track down a specific problem.
5656 4. Testing facility fudge: when running in the test harness and attempting
5657 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5658 getting "No route to host". Convert this to a timeout.
5660 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5663 6. Some OS don't have socklen_t but use size_t instead. This affects the
5664 fifth argument of getsockopt() amongst other things. This is now
5665 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5666 can be set for individual OS. I have set it for SunOS5, OSF1, and
5667 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5668 some earlier ones do not.
5670 7. Change 4.30/15 was not doing the test caselessly.
5672 8. The standard form for an IPv6 address literal was being rejected by address
5673 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5674 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5675 this, as well as the form without the "IPv6" on the front (but only when
5676 address literals are enabled, of course).
5678 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5680 10. Exim crashed if a message with an empty sender address specified by -f
5681 encountered a router with an errors_to setting. This could be provoked only
5682 by a command such as
5686 where an empty string was supplied; "<>" did not hit this bug.
5688 11. Installed PCRE release 4.5.
5690 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5691 remained set. It is now erased.
5693 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5694 times from message ids (which are base 36 rather than the normal 62).
5696 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5697 were being counted as actual protocol errors, and logged if the log
5698 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5699 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5700 and DATA following a set of rejected RCPTs do not count as protocol errors.
5701 In other words, Exim assumes they were pipelined, though this may not
5702 actually be the case. Of course, in all cases the client gets an
5703 appropriate error code.
5705 15. If a lookup fails in an ACL condition, a message about the failure may
5706 be available; it is used if testing the ACL cannot continue, because most
5707 such messages specify what the cause of the deferral is. However, some
5708 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5709 that caused an old message to be retained and used if a later statement
5710 caused a defer, replacing the real cause of the deferral.
5712 16. If an IP address had so many PTR records that the DNS lookup buffer
5713 was not large enough to hold them, Exim could crash while trying to process
5714 the truncated data. It now detects and logs this case.
5716 17. Further to 4.21/58, another change has been made: if (and only if) the
5717 first line of a message (the first header line) ends with CRLF, a bare LF
5718 in a subsequent header line has a space inserted after it, so as not to
5719 terminate the header.
5721 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5722 unnecessarily, used atoi() instead of strtol(), and didn't check the
5723 termination when getting file sizes from file names by regex.
5725 19. Completely re-implemented the support for maildirsize files, in the light
5726 of a number of problems with the previous contributed implementation
5727 (4.30/29). In particular:
5729 . If the quota is zero, the maildirsize file is maintained, but no quota is
5732 . If the maildir directory does not exist, it is created before any attempt
5733 to write a maildirsize file.
5735 . The quota value in the file is just a cache; if the quota is changed in
5736 the transport, the new value overrides.
5738 . A regular expression is available for excluding directories from the
5741 20. The autoreply transport checks the characters in options that define the
5742 message's headers; it allows continued headers, but it was checking with
5743 isspace() after an embedded newline instead of explicitly looking for a
5746 21. If all the "regular" hosts to which an address was routed had passed their
5747 expiry times, and had not reached their retry times, the address was
5748 bounced, even if fallback hosts were defined. Now Exim should go on to try
5751 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5752 equivalent code in the SMTP transport. Some hosts send humungous responses
5753 to HELO/EHLO, more than 1024 it seems.
5755 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5756 gives compiler warnings in some environments. I've now done it "properly",
5759 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5760 (because of problems with the built-in one) was declared to return uschar *
5761 instead of char *, causing compiler failure.
5763 25. Fixed a file descriptor leak when processing alias/forward files.
5765 26. Fixed a minor format string issue in dbfn.c.
5767 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5769 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5770 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5773 29. When a sender address is verified, it is cached, to save repeating the test
5774 when there is more than one recipient in a message. However, when the
5775 verification involves a callout, it is possible for different callout
5776 options to be set for different recipients. It is too complicated to keep
5777 track of this in the cache, so now Exim always runs a verification when a
5778 callout is required, relying on the callout cache for the optimization.
5779 The overhead is duplication of the address routing, but this should not be
5782 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5783 to be verified with callout=postmaster, and the main callout worked but the
5784 postmaster check failed, the verification correctly failed. However, if a
5785 subsequent RCPT command asked for sender verification *without* the
5786 postmaster check, incorrect caching caused this verification also to fail,
5789 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5790 it was not caching the DNS options (qualify_single, search_parents) that
5791 were used when the lookup failed. A subsequent lookup with different
5792 options therefore always gave the same answer, though there were cases
5793 where it should not have. (Example: a "domains = !$mx_any" option on a
5794 dnslookup router: the "domains" option is always processed without any
5795 widening, but the router might have qualify_single set.) Now Exim uses the
5796 cached value only when the same options are set.
5798 32. Added John Jetmore's "exipick" utility to the distribution.
5800 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5801 than a timeout (e.g. a certificate is required, and is not provided), an
5802 Exim server now closes the connection immediately. Previously it waited for
5803 the client to close - but if the client is SSL, it seems that they each
5804 wait for each other, leading to a delay before one of them times out.
5806 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5807 maintained 0.8.x compatibility because I don't think many are using it, and
5808 it is clearly obsolete.
5810 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5813 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5814 client certificate was expired. A simple patch fixes this, though I don't
5815 understand the full logic of why the verify callback is called multiple
5818 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5819 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5820 which causes problems with some clients (such as the Certicom SSL Plus
5821 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5822 disables the coutermeasure allowing Eudora to connect."
5824 38. Exim was not checking that a write() to a log file succeeded. This could
5825 lead to Bad Things if a log got too big, in particular if it hit a file
5826 size limit. Exim now panics and dies if it cannot write to a log file, just
5827 as it does if it cannot open a log file.
5829 39. Modified OS/Makefile-Linux so that it now contains
5831 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5833 The two -D definitions ensure that Exim is compiled with large file
5834 support, which makes it possible to handle log files that are bigger than
5837 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5838 instance) a domain was checked against a named list that involved a lookup,
5839 causing $domain_data to be set, then another domain was checked against the
5840 same list, then the first domain was re-checked, the value of $domain_data
5841 after the final check could be wrong. In particular, if the second check
5842 failed, it could be set empty. This bug probably also applied to
5845 41. The strip_trailing_dot option was not being applied to the address given
5846 with the -f command-line option.
5848 42. The code for reading a message's header from the spool was incrementing
5849 $received_count, but never initializing it. This meant that the value was
5850 incorrect (doubled) while delivering a message in the same process in which
5851 it was received. In the most common configuration of Exim, this never
5852 happens - a fresh exec is done - but it can happen when
5853 deliver_drop_privilege is set.
5855 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5856 it now includes up to 150 characters of the unexpected data in the log
5859 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5860 and building data strings. The size of both of these buffers was 10 000
5861 bytes - far larger than anybody would *ever* want, thought I. Needless to
5862 say, somebody hit the limit. I have increased the maximum line length to
5863 20 000 and the maximum data length of concatenated lines to 100 000. I have
5864 also fixed two bugs, because there was no checking on these buffers. Tsk,
5865 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5866 buffer is too small.
5868 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5869 lsearch lookups. Now it does.
5871 46. When parsing a route_list item in a manualroute router, a fixed-length
5872 buffer was used for the list of hosts. I made this 1024 bytes long,
5873 thinking that nobody would ever have a list of hosts that long. Wrong.
5874 Somebody had a whole pile of complicated expansion conditions, and the
5875 string was silently truncated, leading to an expansion error. It turns out
5876 that it is easier to change to an unlimited length (owing to other changes
5877 that have happened since this code was originally written) than to build
5878 structure for giving a limitation error. The length of the item that
5879 expands into the list of hosts is now unlimited.
5881 47. The lsearch lookup could not handle data where the length of text line was
5882 more than 4095 characters. Such lines were truncated, leading to shortened
5883 data being returned. It should now handle lines of any length.
5885 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5886 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5889 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5890 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5891 no longer happens in comment lines. A list of replacements is now placed
5892 at the head of all of the source files, except those whose only change is
5893 to replace PERL_COMMAND in the very first #! line.
5895 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5896 messages on the queue, with a bottom-up merge sort, using code contributed
5897 by Michael Haardt. This should make operations like -bp somewhat faster on
5898 large queues. It won't affect queue runners, except when queue_run_in_order
5901 51. Installed eximstats 1.31 in the distribution.
5903 52. Added support for SRV lookups to the dnslookup router.
5905 53. If an ACL referred to $message_body or $message_body_end, the value was not
5906 reset for any messages that followed in the same SMTP session.
5908 54. The store-handling optimization for building very long strings was not
5909 differentiating between the different store pools. I don't think this
5910 actually made any difference in practice, but I've tidied it.
5912 55. While running the routers to verify a sender address, $sender_address
5913 was still set to the sender address. This is wrong, because when routing to
5914 send a bounce to the sender, it would be empty. Therefore, I have changed
5915 it so that, while verifying a sender address, $sender_address is set to <>.
5916 (There is no change to what happens when verifying a recipient address.)
5918 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5919 target A or AAAA records (if not already returned) without resetting the
5920 qualify_single or search_parents options of the DNS resolver. These are
5921 inappropriate in this case because the targets of MX and SRV records must
5922 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5923 target that, when qualified, matched something in the local domain. These
5924 two options are now turned off when doing these lookups.
5926 57. It seems that at least some releases of Reiserfs (which does not have the
5927 concept of a fixed number of inodes) returns zero and not -1 for the
5928 number of available inodes. This interacted badly with check_spool_inodes,
5929 which assumed that -1 was the "no such thing" setting. What I have done is
5930 to check that the total number of inodes is greater than zero before doing
5931 the test of how many are available.
5933 58. When a "warn" ACL statement has a log_message modifier, the message is
5934 remembered, and not repeated. This is to avoid a lot of repetition when a
5935 message has many recipients that cause the same warning to be written.
5936 However, Exim was preserving the list of already written lines for an
5937 entire SMTP session, which doesn't seem right. The memory is now reset if a
5938 new message is started.
5940 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5941 debugging output unless log_rewrite was also set.
5943 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5944 of (int)(handle) when we know that handle contains (void *)(-1).
5946 61. The Exim daemon panic-logs an error return when it closes the incoming
5947 connection. However "connection reset by peer" seems to be common, and
5948 isn't really an error worthy of noting specially, so that particular error
5951 62. When Exim is trying to find all the local interfaces, it used to panic and
5952 die if the ioctl to get the interface flags failed. However, it seems that
5953 on at least one OS (Solaris 9) it is possible to have an interface that is
5954 included in the list of interfaces, but for which you get a failure error
5955 for this call. This happens when the interface is not "plumbed" into a
5956 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5957 failure of the "get flags" call assumes that the interface is down.
5959 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5960 makes life easier for people who are doing arithmetic on fields extracted
5961 from dates, where you often get leading zeros that should not be
5962 interpreted as octal.
5964 64. Added qualify_domain to the redirect router, to override the global
5967 65. If a pathologically long header line contained very many addresses (the
5968 report of this problem mentioned 10 000) and each of them was rewritten,
5969 Exim could use up a very large amount of memory. (It kept on making new
5970 copies of the header line as it rewrote, and never released the old ones.)
5971 At the expense of a bit more processing, the header rewriting function has
5972 been changed so that it no longer eats memory in this way.
5974 66. The generation of the Received: header has been moved from the time that a
5975 message starts to be received, to the time that it finishes. The timestamp
5976 in the Received: header should now be very close to that of the <= log
5977 line. There are two side-effects of this change:
5979 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5980 logged header lines no longer include the local Received: line, because
5981 it has not yet been created. The same applies to a copy of the message
5982 that is returned to a non-SMTP sender when a message is rejected.
5984 (b) When a filter file is tested using -bf, no additional Received: header
5985 is added to the test message. After some thought, I decided that this
5988 This change does not affect the value of $received_for. It is still set
5989 after address rewriting, but before local_scan() is called.
5991 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5993 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5994 gave an unhelpful panic error message, and a defer error. I have managed to
5995 change this behaviour so that it now rejects any supplied certificate,
5996 which seems right, as the list of acceptable certificates is empty.
5998 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5999 gave an unhelpful defer error. I have not managed to make this reject any
6000 supplied certificates, but the error message it gives is "no certificate
6001 supplied", which is not helpful.
6003 70. exigrep's output now also includes lines that are not associated with any
6004 message, but which match the given pattern. Implemented by a patch from
6005 Martin Sluka, which also tidied up the Perl a bit.
6007 71. Recipient callout verification, like sender verification, was using <> in
6008 the MAIL FROM command. This isn't really the right thing, since the actual
6009 sender may affect whether the remote host accepts the recipient or not. I
6010 have changed it to use the actual sender in the callout; this means that
6011 the cache record is now keyed on a recipient/sender pair, not just the
6012 recipient address. There doesn't seem to be a real danger of callout loops,
6013 since a callout by the remote host to check the sender would use <>.
6014 [SEE ABOVE: changed after hitting problems.]
6016 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6017 temporary errors. However, in the case of such a code being given after
6018 the end of a data transmission (i.e. after ".") Exim was failing to write
6019 a retry record for the message. (Yes, there was some broken host that was
6020 actually sending 8xx at this point.)
6022 73. An unknown lookup type in a host list could cause Exim to panic-die when
6023 the list was checked. (An example that provoked this was putting <; in the
6024 middle of a list instead of at the start.) If this happened during a DATA
6025 ACL check, a -D file could be left lying around. This kind of configuration
6026 error no longer causes Exim to die; instead it causes a defer error. The
6027 incident is still logged to the main and panic logs.
6029 74. Buglet left over from Exim 3 conversion. The message "too many messages
6030 in one connection" was written to the rejectlog but not the mainlog, except
6031 when address rewriting (yes!) was being logged.
6033 75. Added write_rejectlog option.
6035 76. When a system filter was run not as root (that is, when system_filter_user
6036 was set), the values of the $n variables were not being returned to the
6037 main process; thus, they were not subsequently available in the $sn
6040 77. Added +return_path_on_delivery log selector.
6042 78. A connection timeout was being treated differently from recipients deferred
6043 when testing hosts_max_try with a message that was older than the host's
6044 retry timeout. (The host should not be counted, thus allowing all hosts to
6045 be tried at least once before bouncing.) This may have been the cause of an
6046 occasionally reported bug whereby a message would remain on the queue
6047 longer than the retry timeout, but would be bounced if a delivery was
6048 forced. I say "may" because I never totally pinned down the problem;
6049 setting up timeout/retry tests is difficult. See also the next item.
6051 79. The ultimate address timeout was not being applied to errors that involved
6052 a combination of host plus message (for example, a timeout on a MAIL
6053 command). When an address resolved to a number of possible hosts, and they
6054 were not all tried for each delivery (e.g. because of hosts_max_try), a
6055 message could remain on the queue longer than the retry timeout.
6057 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6060 81. Fixed an obscure SMTP outgoing bug which required at least the following
6061 conditions: (a) there was another message waiting for the same server;
6062 (b) the server returned 5xx to all RCPT commands in the first message so
6063 that the message was not completed; (c) the server dropped the connection
6064 or gave a negative response to the RSET that Exim sends to abort the
6065 transaction. The observed case was a dropped connection after DATA that had
6066 been sent in pipelining mode. That is, the server had advertised PIPELINING
6067 but was not implementing it correctly. The effect of the bug was incorrect
6068 behaviour, such as trying another host, and this could lead to a crash.
6074 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6075 and daemon.c were passed as pointers to ints; they should have been
6076 pointers to socklen_t variables (which are typically unsigned ints).
6078 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6081 3. Fixed a really odd bug that affected only the testing scheme; patching a
6082 certain fixed string in the binary changed the value of another string that
6083 happened to be identical to the end of the original first string.
6085 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6086 name", it returns that address as the IP address. On some operating
6087 systems (e.g. Solaris), it also passes back the IP address string as the
6088 "host name". However, on others (e.g. Linux), it passes back an empty
6089 string. Exim wasn't checking for this, and was changing the host name to an
6090 empty string, assuming it had been canonicalized.
6092 5. Although rare, it is permitted to have more than one PTR record for a given
6093 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6094 all the names associated with an address, because they do in Solaris.
6095 However, it seems that they do not in Linux for data that comes from the
6096 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6097 I found this out when I moved to a new Linux workstation and tried to run
6098 the Exim test suite.
6100 To get round this problem I have changed the code so that it now does its
6101 own call to the DNS to look up PTR records when searching for a host name.
6102 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6103 addresses that are only in /etc/hosts are still found.
6105 This behaviour is, however, controlled by an option called host_lookup_
6106 order, which defaults to "bydns:byaddr". If people want to use the other
6107 order, or indeed, just use one or the other means of lookup, they can
6108 specify it in this variable.
6110 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6111 some operating systems, this comes back from gethostbyaddr() as an empty
6112 string, and this is what Exim used to test for. However, it seems that in
6113 other systems, "." is yielded. Exim now tests for this case too.
6115 7. The values of check_spool_space and check_log_space are now held internally
6116 as a number of kilobytes instead of an absolute number of bytes. If a
6117 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6118 kilobyte. This means that much larger values can be stored.
6120 8. Exim monitor: an attempt to get the action menu when not actually pointing
6121 at a message produces an empty menu entitled "No message selected". This
6122 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6123 no entries in it ("Shell widget menu has zero width and/or height"). So I
6124 have added a single, blank menu entry in this case.
6126 9. Added ${quote_local_part.
6128 10. MIME decoding is now applied to the contents of Subject: header lines when
6131 11. Now that a reference to $sender_host_address automatically causes a reverse
6132 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6133 host lookup before query-style lookups in lists that might use this
6134 variable. This has therefore been abolished, and the "net-" prefix is no
6135 longer necessary for query-style lookups.
6137 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6138 have been a typo for LFLAGS, so it has been changed.
6140 13. The install script calls Exim with "-C /dev/null" in order to find the
6141 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6142 to be output. However, since Exim outputs its version number before the
6143 error, it didn't break the script. It just looked ugly. I fixed this by
6144 always allowing "-C /dev/null" if the caller is root.
6146 14. Ignore overlarge ACL variable number when reading spool file - insurance
6147 against a later release with more variables having written the file.
6149 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6150 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6151 this, as well as the form without the "IPv6" on the front.
6153 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6154 OS/Makefile-Darwin file.
6156 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6157 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6159 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6160 variable. However, some OS use ints for inodes. I've added cast to long int
6161 to get rid of the compiler warning.
6163 19. I had forgotten to lock out "/../" in configuration file names when
6164 ALT_CONFIG_PREFIX was set.
6166 20. Routers used for verification do not need to specify transports. However,
6167 if such a router generated a host list, and callout was configured, Exim
6168 crashed, because it could not find a port number from the (non-existent)
6169 transport. It now assumes port 25 in this circumstance.
6171 21. Added the -t option to exigrep.
6173 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6174 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6175 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6176 them is set, LOOKUP_LSEARCH is forced.
6178 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6180 24. Added sender and host information to the "rejected by local_scan()" log
6181 line; previously there was no indication of these.
6183 25. Added .include_if_exists.
6185 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6186 Linux. It turns out that not all file systems support this. Apparently some
6187 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6188 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6189 not supported on the file descriptor, is now ignored when Exim is trying to
6190 sync a directory. This applies only to Linux.
6192 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6194 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6195 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6198 29. Added support for maildirsize files from supplied patch (modified a bit).
6200 30. The use of :fail: followed by an empty string could lead Exim to respond to
6201 sender verification failures with (e.g.):
6203 550 Verification failed for <xxx>
6204 550 Sender verify failed
6206 where the first response line was missing the '-' that indicates it is not
6207 the final line of the response.
6209 31. The loop for finding the name of the user that called Exim had a hardwired
6210 limit of 10; it now uses the value of finduser_retries, which is used for
6211 all other user lookups.
6213 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6216 33. Exim was neglecting to zero errno before one call of strtol() when
6217 expanding a string and expecting an integer value. On some systems this
6218 resulted in spurious "integer overflow" errors. Also, it was casting the
6219 result into an int without checking.
6221 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6222 did not work. The code looks as if it has *never* worked, though it appears
6223 to have been documented since at least release 1.62. I have made it work.
6225 35. The "timeout_DNS" error in retry rules, also documented since at least
6226 1.62, also never worked. As it isn't clear exactly what this means, and
6227 clearly it isn't a major issue, I have abolished the feature by treating it
6228 as "timeout", and writing a warning to the main and panic logs.
6230 36. The display of retry rules for -brt wasn't always showing the error code
6233 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6234 timeout_connect_A, timeout_connect_MX.
6236 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6237 to the empty sender.
6239 39. The daemon was not analysing the content of -oX till after it had closed
6240 stderr and disconnected from the controlling terminal. This meant that any
6241 syntax errors were only noted on the panic log, and the return code from
6242 the command was 0. By re-arranging the code a little, I've made the
6243 decoding happen first, so such errors now appear on stderr, and the return
6244 code is 1. However, the actual setting up of the sockets still happens in
6245 the disconnected process, so errors there are still only recorded on the
6248 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6249 connections (as happens on some IP stacks) was logged at start up time as
6250 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6251 it from "IPv6 and IPv4", which means that two separate sockets are being
6254 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6255 says whether AF_INET or AF_INET6 was passed as an argument.
6257 42. Exiwhat output was messed up when time zones were included in log
6260 43. Exiwhat now gives more information about the daemon's listening ports,
6261 and whether -tls-on-connect was used.
6263 44. The "port" option of the smtp transport is now expanded.
6265 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6266 silently ignored. Now an error message is written to the main and panic
6269 46. There's a new ACL modifier called "logwrite" which writes to a log file
6270 as soon as it is encountered.
6272 47. Added $local_user_uid and $local_user_gid at routing time.
6274 48. Exim crashed when trying to verify a sender address that was being
6277 49. Exim was recognizing only a space character after ".include". It now also
6278 recognizes a tab character.
6280 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6281 extracting the relevant information from the specification. The man page no
6282 longer contains scrambled data for the -d option, and I've added a section
6283 at the front about calling Exim under different names.
6285 51. Added "extra_headers" argument to the "mail" command in filter files.
6287 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6290 53. Installed eximstats 1.29.
6292 54. Added transport_filter_timeout as a generic transport option.
6294 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6295 Cc: header lines. This was required by RFC 822, but it not required by RFC
6298 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6299 incoming messages that did not have them. Now it does so only if the
6300 message originates locally, that is, if there is no associated remote host
6301 address. When Resent- header lines are present, this applies to the Resent-
6302 lines rather than the non-Resent- lines.
6304 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6305 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6307 58. Messages for configuration errors now include the name of the main
6308 configuration file - useful now that there may be more than one file in a
6309 list (.included file names were always shown).
6311 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6312 for those rare installations that do not start the daemon as root or run it
6313 setuid root. I've cut out the call to initgroups() if the daemon is not
6316 60. The Exim user and group can now be bound into the binary as text strings
6317 that are looked up at the start of Exim's processing.
6319 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6321 62. Added $mailstore_basename variable.
6323 63. Installed patch to sieve.c from Michael Haardt.
6325 64. When Exim failed to open the panic log after failing to open the main log,
6326 the original message it was trying to log was written to stderr and debug
6327 output, but if they were not available (the usual case in production), it
6328 was lost. Now it is written to syslog before the two lines that record the
6329 failures to open the logs.
6331 65. Users' Exim filters run in subprocesses under the user's uid. It is
6332 possible for a "deliver" command or an alias in a "personal" command to
6333 provoke an address rewrite. If logging of address rewriting is configured,
6334 this fails because the process is not running as root or exim. There may be
6335 a better way of dealing with this, but for the moment (because 4.30 needs
6336 to be released), I have disabled address rewrite logging when running a
6337 filter in a non-root, non-exim process.
6343 1. The buildconfig auxiliary program wasn't quoting the value set for
6344 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6345 not defined. This bug was masked in 4.22 by the effect that was fixed in
6348 2. Some messages that were rejected after a message id was allocated were
6349 shown as "incomplete" by exigrep. It no longer does this for messages that
6350 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6352 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6353 have allow_domain_literals set, the ID did not get logged in the <= line.
6354 Domain literals are now always recognized in Message-ID: header lines.
6356 4. The first argument for a ${extract expansion item is the key name or field
6357 number. Leading and trailing spaces in this item were not being ignored,
6358 causing some misleading effects.
6360 5. When deliver_drop_privilege was set, single queue runner processes started
6361 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6362 same command in the process it spins off) were not dropping privilege.
6364 6. When the daemon running as "exim" started a queue runner, it always
6365 re-executed Exim in the spun-off process. This is a waste of effort when
6366 deliver_drop_privilege is set. The new process now just calls the
6367 queue-runner function directly.
6373 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6376 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6377 ignored. Though the use of -oP was forcing the writing of a pid file, it
6378 was always written to the default place.
6380 3. If the message "no IP address found for host xxxx" is generated during
6381 incoming verification, it is now followed by identification of the incoming
6382 connection (so you can more easily find what provoked it).
6384 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6386 5. Added some features to "harden" Exim a bit more against certain attacks:
6388 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6389 be put in Local/Makefile. This is like the never_users runtime option,
6390 but it cannot be overridden. The default setting is "root".
6392 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6393 prefix string with which any file named in a -C command line option
6396 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6397 is retained for -C and -D only if the caller of Exim is root. Without
6398 it, the exim user may also use -C and -D and retain privilege.
6400 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6401 command line option is disabled.
6403 6. Macro names set by the -D option must start with an upper case letter, just
6404 like macro names defined in the configuration file.
6406 7. Added "dereference=" facility to LDAP.
6408 8. Two instances of the typo "uknown" in the source files are fixed.
6410 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6411 the Configure-Makefile script screwed up while processing it.
6413 10. Incorporated PCRE 4.4.
6415 11. The SMTP synchronization check was not operating right at the start of an
6416 SMTP session. For example, it could not catch a HELO sent before the client
6417 waited for the greeting. There is now a check for outstanding input at the
6418 point when the greeting is written. Because of the duplex, asynchronous
6419 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6420 way, but not yet received, when the check is performed.
6422 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6423 on TCP/IP sockets, because this apparently causes some broken clients to
6426 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6427 unchanged) from the Cygwin maintainer.
6429 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6430 was supported instead of testing for mbx. Effectively a typo.
6432 15. The spa authenticator server code was not checking that the input it
6433 received was valid base64.
6435 16. The debug output line for the "set" modifier in ACLs was not showing the
6436 name of the variable that was being set.
6438 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6440 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6441 lookup on its own. Something else was needed to trigger the lookup. For
6442 example, a match in host_lookup or the need for a host name in a host list.
6443 Now, if $sender_host_name is referenced and the host name has not yet been
6444 looked up, a lookup is performed. If the lookup fails, the variable remains
6445 empty, and $host_lookup_failed is set to "1".
6447 19. Added "eqi" as a case-independent comparison operator.
6449 20. The saslauthd authentication condition could segfault if neither service
6450 nor realm was specified.
6452 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6453 error message that was logged was misleading, and incoming SMTP
6454 connections were dropped. The message is now more accurate, and temporary
6455 errors are given to SMTP connections.
6457 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6458 (except RSET) with a 421 error, until QUIT is received. However, it was
6459 failing to send a response to QUIT.
6461 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6462 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6463 ACL is now run after the helo_try_verify_hosts code.
6465 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6466 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6467 case-independent, so other case variants are also recognized). Apparently
6468 some systems use these upper case variants.
6470 25. If more than two messages were waiting for the same host, and a transport
6471 filter was specified for the transport, Exim sent two messages over the
6472 same TCP/IP connection, and then failed with "socket operation on non-
6473 socket" when it tried to send the third.
6475 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6477 27. The extern definition of crypt16() in expand.c was not being excluded when
6478 the OS had its own crypt16() function.
6480 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6481 as a preferred synonym for return_size_limit, both as an option and as an
6484 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6486 30. Changed the default configuration ACL to relax the local part checking rule
6487 for addresses that are not in any local domains. For these addresses,
6488 slashes and pipe symbols are allowed within local parts, but the sequence
6489 /../ is explicitly forbidden.
6491 31. SPA server authentication was not clearing the challenge buffer before
6494 32. log_message in a "warn" ACL statement was writing to the reject log as
6495 well as to the main log, which contradicts the documentation and doesn't
6496 seem right (because no rejection is happening). So I have stopped it.
6498 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6499 However, I am unable to do any testing of this.
6501 34. Fixed an infelicity in the appendfile transport. When checking directories
6502 for a mailbox, to see if any needed to be created, it was accidentally
6503 using path names with one or more superfluous leading slashes; tracing
6504 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6506 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6507 contents are added to the log line that is written for every discarded
6508 recipient. (Previously a log_message setting was ignored.)
6510 36. The ${quote: operator now quotes the string if it is empty.
6512 37. The install script runs exim in order to find its version number. If for
6513 some reason other than non-existence or emptiness, which it checks, it
6514 could not run './exim', it was installing it with an empty version number,
6515 i.e. as "exim-". This error state is now caught, and the installation is
6518 38. An argument was missing from the function that creates an error message
6519 when Exim fails to connect to the socket for saslauthd authentication.
6520 This could cause Exim to crash, or give a corrupted message.
6522 39. Added isip, isip4, and isip6 to ${if conditions.
6524 40. The ACL variables $acl_xx are now saved with the message, and can be
6525 accessed later in routers, transports, and filters.
6527 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6528 strings in the file are not string-expanded.
6530 42. If a MAIL command specified a SIZE value that was too large to fit into an
6531 int variable, the check against message_size_limit failed. Such values are
6532 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6533 day this will have to be increased, but I don't think I want to be around
6534 when emails are that large.
6541 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6542 iconv() is not standard in FreeBSD.
6544 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6545 IPv6 enabled. The observed symptom was a segmentation fault on return from
6546 the function os_common_find_running_interfaces() in src/os.c.
6548 3. In the check_special_case() function in daemon.c I had used "errno" as an
6549 argument name, which causes warnings on some systems. This was basically a
6550 typo, since it was named "eno" in the comments!
6552 4. The code that waits for the clock to tick (at a resolution of some fraction
6553 of a second) so as to ensure message-id uniqueness was always waiting for
6554 at least one whole tick, when it could have waited for less. [This is
6555 almost certainly not relevant at current processor speeds, where it is
6556 unlikely to ever wait at all. But we try to future-proof.]
6558 5. The function that sleeps for a time interval that includes fractions of a
6559 second contained a race. It did not block SIGALRM between setting the
6560 timer, and suspending (a couple of lines later). If the interval was short
6561 and the sigsuspend() was delayed until after it had expired, the suspension
6562 never ended. On busy systems this could lead to processes getting stuck for
6565 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6566 process, before it forks any delivery processes. The open lookup caching
6567 mechanism meant that the open file or database connection was passed into
6568 the delivery process. The problem was that delivery processes always tidy
6569 up cached lookup data. This could cause a problem for the next delivery
6570 process started by the queue runner, because the external queue runner
6571 process does not know about the closure. So the next delivery process
6572 still has data in the lookup cache. In the case of a file lookup, there was
6573 no problem because closing a file descriptor in a subprocess doesn't affect
6574 the parent. However, if the lookup was caching a connection to a database,
6575 the connection was closed, and the second delivery process was likely to
6576 see errors such as "PGSQL: query failed: server closed the connection
6577 unexpectedly". The problem has been fixed by closing all cached lookups
6578 in a queue runner before running a delivery process.
6580 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6581 seem to have the "const" qualifier which it has on other OS. I've
6584 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6585 *of the same type* (client or server) with the same public name that an
6586 error should be diagnosed.
6588 9. When Exim looked up a host name for an IP address, but failed to find the
6589 original IP address when looking up the host name (a safety check), it
6590 output the message "<ip address> does not match any IP for NULL", which was
6591 confusing, to say the least. The bug was that the host name should have
6592 appeared instead of "NULL".
6594 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6595 user that is built into the binary, and the -C or -D options is used, root
6596 privilege is dropped before the configuration file is read. In addition,
6597 logging is switched to stderr instead of the normal log files. If the
6598 configuration then re-defines the Exim user, the unprivileged environment
6599 is probably not what is expected, so Exim logs a panic warning message (but
6602 However, if deliver_drop_privilege is set, the unprivileged state may well
6603 be exactly what is intended, so the warning has been cut out in that case,
6604 and Exim is allowed to try to write to its normal log files.
6610 1. smtp_return_error_details was not giving details for temporary sender
6611 or receiver verification errors.
6613 2. Diagnose a configuration error if two authenticators have the same public
6616 3. Exim used not to create the message log file for a message until the first
6617 delivery attempt. This could be confusing when incoming messages were held
6618 for policy or load reasons. The message log file is now created at the time
6619 the message is received, and an initial "Received" line is written to it.
6621 4. The automatically generated man page for command line options had a minor
6622 bug that caused no ill effects; however, a more serious problem was that
6623 the procedure for building the man page automatically didn't always
6624 operate. Consequently, release 4.20 contains an out-of-date version. This
6625 shouldn't happen again.
6627 5. When building Exim with embedded Perl support, the script that builds the
6628 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6629 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6631 6. The freeze_tell option was not being used for messages that were frozen on
6632 arrival, either by an ACL or by local_scan().
6634 7. Added the smtp_incomplete_transaction log selector.
6636 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6637 was accepting AUTH without a new EHLO.
6639 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6640 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6643 10. Logging of TCP/IP connections (when configured) now happens in the main
6644 daemon process instead of the child process, so that the TCP/IP connection
6645 count is more accurate (but it can never be perfect).
6647 11. The use of "drop" in a nested ACL was not being handled correctly in the
6648 outer ACL. Now, if condition failure induced by the nested "drop" causes
6649 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6650 or "require"), the connection is dropped.
6652 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6653 that yield "discard" can now be used with an "accept" or a "discard" verb,
6654 but an error is generated for any others (because I can't see a useful way
6655 to define what should happen).
6657 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6658 are now processed in the same way as lines in the Exim configuration file.
6659 In particular, continuation lines are supported.
6661 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6663 15. Added -ti meaning -t -i.
6665 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6666 domains, and warn by logging if others are found.
6668 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6669 returned from GIFCONF when Exim is trying to find the list of interfaces on
6670 a host. The code in os.c has been modified to copy each ifreq to an aligned
6671 structure in all cases.
6673 Also, in some cases, the returned ifreq's were being copied to a 'struct
6674 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6675 means the last couple of bytes of an IPv6 address could be chopped if the
6676 ifreq contained only a normal sockaddr (14 bytes storage).
6678 18. Named domain lists were not supported in the hosts_treat_as_local option.
6679 An entry such as +xxxx was not recognized, and was treated as a literal
6682 19. Ensure that header lines added by a DATA ACL are included in the reject log
6683 if the ACL subsequently rejects the message.
6685 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6686 MD5 (which is deprecated).
6688 21. When testing a filter file using -bf, Exim was writing a message when it
6689 took the sender from a "From " line in the message, but it was not doing so
6690 when it took $return_path from a Return-Path: header line. It now does.
6692 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6693 with a valid header line field name (a series of printing characters
6694 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6696 23. Changed "disc" in the source to "disk" to conform to the documentation and
6697 the book and for uniformity.
6699 24. Ignore Sendmail's -Ooption=value command line item.
6701 25. When execve() failed while trying to run a command in a pipe transport,
6702 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6703 could be confused with a return value of 69 from the command itself. This
6704 has been changed to 127, the value the shell returns if it is asked to run
6705 a non-existent command. The wording for the related log line suggests a
6706 non-existent command as the problem.
6708 26. If received_header_text expands to an empty string, do not add a Received:
6709 header line to the message. (Well, it adds a token one on the spool, but
6710 marks it "old" so that it doesn't get used or transmitted.)
6712 27. Installed eximstats 1.28 (addition of -nt option).
6714 28. There was no check for failure on the call to getsockname() in the daemon
6715 code. This can fail if there is a shortage of resources on the system, with
6716 ENOMEM, for example. A temporary error is now given on failure.
6718 29. Contrary to the C standard, it seems that in some environments, the
6719 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6720 program. Exim now does this explicitly; it affects the formatting of
6721 timestamps using strftime().
6723 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6724 complaints. I've now initialized all the variables, to avoid this.
6726 32. Header lines added by a system filter were not being "seen" during
6727 transport-time rewrites.
6729 33. The info_callback() function passed to OpenSSL is set up with type void
6730 (*)(SSL *, int, int), as described somewhere. However, when calling the
6731 function (actually a macro) that sets it up, the type void(*)() is
6732 expected. I've put in a cast to prevent warnings from picky compilers.
6734 34. If a DNS black list lookup found a CNAME record, but there were no A
6735 records associated with the domain it pointed at, Exim crashed.
6737 35. If a DNS black list lookup returned more than one A record, Exim ignored
6738 all but the first. It now scans all returned addresses if a particular IP
6739 value is being sought. In this situation, the contents of the
6740 $dnslist_value variable are a list of all the addresses, separated by a
6743 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6744 to accept a host name and all its aliases if the forward lookup for any of
6745 them yielded the IP address of the incoming connection. Now it accepts only
6746 those names whose forward lookup yields the correct IP address. Any other
6747 names are discarded. This closes a loophole whereby a rogue DNS
6748 administrator could create reverse DNS records to break through a
6749 wildcarded host restriction in an ACL.
6751 37. If a user filter or a system filter that ran in a subprocess used any of
6752 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6753 the wrong values were passed to the pipe command ($thisaddress had the
6754 value of $0, $0 had the value of $1, etc). This bug was introduced by
6755 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6757 38. Improved the line breaking for long SMTP error messages from ACLs.
6758 Previously, if there was no break point between 40 and 75 characters, Exim
6759 left the rest of the message alone. Two changes have been made: (a) I've
6760 reduced the minimum length to 35 characters; (b) if it can't find a break
6761 point between 35 and 75 characters, it looks ahead and uses the first one
6762 that it finds. This may give the occasional overlong line, but at least the
6763 remaining text gets split now.
6765 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6766 file descriptors might be low, and that setting 1000 would always raise it.
6767 It turns out that in some environments, the limit is already over 1000 and
6768 that lowering it causes trouble. So now Exim takes care not to decrease it.
6770 40. When delivering a message, the value of $return_path is set to $sender_
6771 address at the start of routing (routers may change the value). By an
6772 oversight, this default was not being set up when an address was tested by
6773 -bt or -bv, which affected the outcome if any router or filter referred to
6776 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6777 log only when "message" or "log_message" are set. However, if one of the
6778 conditions was an address verification, or a call to a nested ACL, the
6779 messages generated by the underlying test were being passed through. This
6780 no longer happens. The underlying message is available in $acl_verify_
6781 message for both "message" and "log_message" expansions, so it can be
6782 passed through if needed.
6784 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6785 new expansion $bh_ to give the encoded byte string without charset
6786 translation. Translation happens only if iconv() is available; HAVE_ICONV
6787 indicates this at build time. HEADERS_CHARSET gives the charset to
6788 translate to; headers_charset can change it in the configuration, and
6789 "headers charset" can change it in an individual filter file.
6791 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6792 that creates RFC 2047 encoded "words" labels them as that charset instead
6793 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6794 expansion operator; (ii) when Exim creates a From: line for a local
6795 message; (iii) when a header line is rewritten to include a "phrase" part.
6797 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6798 buggy, causing it to skip the first lines of messages whose message ID
6799 ended in 'D'. This would not have bitten before Exim release 4.14, because
6800 message IDs were unlikely to end in 'D' before then. The effect was to have
6801 incorrect size information for certain domains.
6803 45. #include "config.h" was missing at the start of the crypt16.c module. This
6804 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6807 46. If there was a timeout during a "random" callout check, Exim treated it as
6808 a failure of the random address, and carried on sending RSET and the real
6809 address. If the delay was just some slowness somewhere, the response to the
6810 original RCPT would be taken as a response to RSET and so on, causing
6811 mayhem of various kinds.
6813 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6814 when I implemented it. It didn't allow for the fact that some option values
6815 may legitimately be negative (e.g. size_addition), and it didn't even do
6816 the right test for positive values.
6818 48. Domain names in DNS records are case-independent. Exim always looks them up
6819 in lower case. Some resolvers return domain names in exactly the case they
6820 appear in the zone file, that is, they may contain uppercase letters. Not
6821 all resolvers do this - some return always lower case. Exim was treating a
6822 change of case by a resolver as a change of domain, similar to a widening
6823 of a domain abbreviation. This triggered its re-routing code and so it was
6824 trying to route what was effectively the same domain again. This normally
6825 caused routing to fail (because the router wouldn't handle the domain
6826 twice). Now Exim checks for this case specially, and just changes the
6827 casing of the domain that it ultimately uses when it transmits the message
6830 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6833 50. If a filter generated a file delivery with a non-absolute name (possible if
6834 no home directory exists for the router), the forbid_file option was not
6837 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6838 the existing equality matching.
6840 52. Exim was using ints instead of ino_t variables in some places where it was
6841 dealing with inode numbers.
6843 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6844 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6845 called TMPDIR, and if it finds it is different, it changes its value.
6847 54. The smtp_printf() function is now made available to local_scan() so
6848 additional output lines can be written before returning. There is also an
6849 smtp_fflush() function to enable the detection of a dropped connection.
6850 The variables smtp_input and smtp_batched_input are exported to
6853 55. Changed the default runtime configuration: the message "Unknown user"
6854 has been removed from the ACL, and instead placed on the localuser router,
6855 using the cannot_route_message feature. This means that any verification
6856 failures that generate their own messages won't get overridden. Similarly,
6857 the "Unrouteable address" message that was in the ACL for unverifiable
6858 relay addresses has also been removed.
6860 56. Added hosts_avoid_esmtp to the smtp transport.
6862 57. The exicyclog script was not checking for the esoteric option
6863 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6864 will work only if exicyclog is run under the appropriate euid.
6866 58. Following a discussion on the list, the rules by which Exim recognises line
6867 endings on incoming messages have been changed. The -dropcr and drop_cr
6868 options are now no-ops, retained only for backwards compatibility. The
6869 following line terminators are recognized: LF CRLF CR. However, special
6870 processing applies to CR:
6872 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6873 nor a local message in the state where . is a terminator.
6875 (ii) If a bare CR is encountered in a header line, an extra space is added
6876 after the line terminator so as not to end the header. The reasoning
6877 behind this is that bare CRs in header lines are most likely either
6878 to be mistakes, or people trying to play silly games.
6880 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6881 was being incorrectly given as 18 bytes larger than it should have been.
6882 This is a VOB (very old bug).
6884 60. This may never have affected anything current, but just in case it has:
6885 When the local host is found other than at the start of a list of hosts,
6886 the local host, those with the same MX, and any that follow, are discarded.
6887 When the list in question was part of a longer list of hosts, the following
6888 hosts (not currently being processed) were also being discarded. This no
6889 longer happens. I'm not sure if this situation could ever has previously
6892 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6895 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6896 header. This is something that is recommended in a new Internet Draft, and
6897 is something that is documented as being done by Sendmail. There are two
6898 possible values. For messages generated by the autoreply transport, Exim
6901 Auto-Submitted: auto-replied
6903 whereas for all other generated messages (e.g. bounces) it adds
6905 Auto-Submitted: auto-generated
6907 63. The "personal" condition in filters now includes a test for the
6908 Auto-Submitted: header. If it contains the string "auto-" the message it
6909 not considered personal.
6911 64. Added rcpt_include_affixes as a generic transport option.
6913 65. Added queue_only_override (default true).
6915 66. Added the syslog_duplication option.
6917 67. If what should have been the first header line of a message consisted of
6918 a space followed by a colon, Exim was mis-interpreting it as a header line.
6919 It isn't of course - it is syntactically invalid and should therefore be
6920 treated as the start of the message body. The misbehaviour could have
6921 caused a number of strange effects, including loss of data in subsequent
6922 header lines, and spool format errors.
6924 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6925 client host had authenticated. This control can now be exercised by an ACL
6926 for more flexibility.
6928 69. By default, callouts do not happen when testing with -bh. There is now a
6929 variant, -bhc, which does actually run the callout code, including
6930 consulting and updating the callout cache.
6932 70. Added support for saslauthd authentication, courtesy of Alexander
6935 71. If statvfs() failed on the spool or log directories while checking their
6936 size for availability, Exim confusingly gave the error "space shortage".
6937 Furthermore, in debugging mode it crashed with a floating point exception.
6938 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6939 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6940 serious problem, Exim now writes to the main and panic logs when this
6941 happens, with details of the failure. It then refuses to accept the
6942 incoming message, giving the message "spool directory problem" or "log
6943 directory problem" with a 421 code for SMTP messages.
6945 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6946 0, 1, and 2 exist, because some OS complain for execs without them (see
6947 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6948 descriptors. However, the code omitted to check that the open succeeded,
6949 causing mysterious errors if for some reason the permissions on /dev/null
6950 got screwed. Now Exim writes a message to the main and panic logs, and
6951 bombs out if it can't open /dev/null.
6953 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6954 interact so that it is all more flexible. It is supposed to remain
6955 backwards compatible. Also added extra_local_interfaces.
6957 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6958 to bomb out with an assertion failure - to the client this appears as a
6959 connection drop. This problem occurs in the part of the code that was taken
6960 from the Samba project. Fortunately, the assertion is in a very simple
6961 function, so I have fixed this by reproducing the function inline in the
6962 one place where it is called, and arranging for authentication to fail
6963 instead of killing the process with assert().
6965 75. The SPA client code was not working when the server requested OEM rather
6966 than Unicode encoding.
6968 76. Added code to make require_files with a specific uid setting more usable in
6969 the case where statting the file as root fails - usually a non-root-mounted
6970 NFS file system. When this happens and the failure is EACCES, Exim now
6971 forks a subprocess and does the per-uid checking as the relevant uid.
6973 77. Added process_log_path.
6975 78. If log_file_path was not explicitly set, a setting of check_log_space or
6976 check_log_inodes was ignored.
6978 79. If a space check for the spool or log partitions fails, the incident is now
6979 logged. Of course, in the latter case the data may get lost...
6981 80. Added the %p formatting code to string_format() so that it can be used to
6982 print addresses in debug_print(). Adjusted all the address printing in the
6983 debugging in store.c to use %p rather than %d.
6985 81. There was a concern that a line of code in smtp_in.c could overflow a
6986 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6987 initially expressed, the concern was not well-founded, because trailing
6988 spaces are removed early. However, if the trailing spaces were followed by
6989 a NULL, they did not get removed, so the overflow was possible. Two fixes
6992 (a) I re-wrote the offending code in a cleaner fashion.
6993 (b) If an incoming SMTP command contains a NULL character, it is rejected
6996 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6997 runs initgroups(), so that if the Exim user is in any additional groups,
6998 they will be used during message reception.
7004 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob