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.
44 JH/01 Rework error string handling in TLS interface so that the caller in
45 more cases is responsible for logging. This permits library-sourced
46 string to be attached to addresses during delivery, and collapses
47 pairs of long lines into single ones.
49 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
50 during configuration. Wildcards are allowed and expanded.
52 JH/02 Rework error string handling in DKIM to pass more info back to callers.
53 This permits better logging.
55 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
56 do not close it down and have the child transport start it up again on
57 the passed-on TCP connection. Instead, proxy the child (and any
58 subsequent ones) for TLS via a unix-domain socket channel. Logging is
59 affected: the continued delivery log lines do not have any DNSSEC, TLS
60 Certificate or OCSP information. TLS cipher information is still logged.
62 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
63 identical IP addresses on different listening ports. Will also affect
66 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
67 add noisy ifdef guards to special-case this sillyness.
68 Patch from Bernd Kuhls.
70 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
71 than 255 are no longer allowed.
73 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
74 Disable the session-cache too, which might reduce our load. Since we
75 currrectly use a new context for every connection, both as server and
76 client, there is no benefit for these.
77 GnuTLS appears to not support tickets server-side by default (we don't
78 call gnutls_session_ticket_enable_server()) but client side is enabled
79 by default on recent versions (3.1.3 +) unless the PFS priority string
82 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
83 <https://reproducible-builds.org/specs/source-date-epoch/>.
85 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
86 the check for any unsuccessful recipients did not notice the limit, and
87 erroneously found still-pending ones.
89 JH/08 Pipeline CHUNKING command and data together, on kernels that support
90 MSG_MORE. Only in-clear (not on TLS connections).
92 JH/09 Avoid using a temporary file during transport using dkim. Unless a
93 transport-filter is involved we can buffer the headers in memory for
94 creating the signature, and read the spool data file once for the
95 signature and again for transmission.
97 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
98 4.77 as the kernel support then wasn't solid, having issues in 64bit
99 mode. Now, it's been long enough. Add support for FreeBSD also.
101 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
102 case where the routing stage had gathered several addresses to send to
103 a host before calling the transport for the first, we previously failed
104 to close down TLS in the old transport process before passing the TCP
105 connection to the new process. The new one sent a STARTTLS command
106 which naturally failed, giving a failed delivery and bloating the retry
107 database. Investigation and fix prototype from Wolfgang Breyha.
109 JH/12 Fix check on SMTP command input synchronisation. Previously there were
110 false-negatives in the check that the sender had not preempted a response
111 or prompt from Exim (running as a server), due to that code's lack of
112 awareness of the SMTP input buffering.
114 PP/04 Add commandline_checks_require_admin option.
115 Exim drops privileges sanely, various checks such as -be aren't a
116 security problem, as long as you trust local users with access to their
117 own account. When invoked by services which pass untrusted data to
118 Exim, this might be an issue. Set this option in main configuration
119 AND make fixes to the calling application, such as using `--` to stop
122 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
123 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
124 responses to those, into a single TLS record each way (this usually means
125 a single packet). As a side issue, smtp_enforce_sync now works on TLS
128 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
129 affects you only if you're dancing at the edge of the param size limits.
130 If you are, and this message makes sense to you, then: raise the
131 configured limit or use OpenSSL 1.1. Nothing we can do for older
134 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
135 cmdline element to get the documented default one. Previously it was
138 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
139 are unsuable for non-message-specific reasons - eg. connection timeout,
140 banner-time rejection.
142 JH/16 Fix logging of delivery remote port, when specified by router, under
145 PP/06 Repair manualroute's ability to take options in any order, even if one
146 is the name of a transport.
149 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
151 JH/17 Change the list-building routines interface to use the expanding-string
152 triplet model, for better allocation and copying behaviour.
154 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
155 Previously it was constructed the first time a possibly-matching string
156 was met in the configuration file input during startup; now it is done
159 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
160 compatible one, to avoid the (poorly documented) possibility of a config
161 file in the working directory redirecting the DB files, possibly correpting
162 some existing file. CVE-2017-10140 assigned for BDB.
164 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
165 cache-hot. Previously, although the result was properly cached, the
166 initial verify call returned a defer.
168 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
169 the main verify for receipient in uncached-mode.
171 JH/22 Retire historical build files to an "unsupported" subdir. These are
172 defined as "ones for which we have no current evidence of testing".
174 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
175 if present. Previously it was ignored.
177 JH/24 Start using specified-initialisers in C structure init coding. This is
178 a C99 feature (it's 2017, so now considered safe).
180 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
181 if was a fixed-sized field and bitmask ops via macros; it is now more
184 PP/07 GitHub PR 56: Apply MariaDB build fix.
185 Patch provided by Jaroslav Škarvada.
187 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
188 during Coverity cleanups [4.87 JH/47]
189 Diagnosis and fix provided by Michael Fischer v. Mollard.
191 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
192 the right size to place the terminating semicolon on its own folded
193 line, the header hash was calculated to an incorrect value thanks to
194 the (relaxed) space the fold became.
196 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
197 and confused the parent.
199 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
200 which could crash as a result. This could lead to undeliverable messages.
202 JH/28 Logging: "next input sent too soon" now shows where input was truncated
205 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
206 matters on fast-turnover and PID-randomising systems, which were getting
207 out-of-order delivery.
209 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
210 a possibly-overlapping copy. The symptom was that "Remote host closed
211 connection in response to HELO" was logged instead of the actual 4xx
214 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
215 Previously only that bufferd was discarded, resulting in SYMTP command
218 JH/32 DKIM: when a message has multiple signatures matching an identity given
219 in dkim_verify_signers, run the dkim acl once for each. Previously only
220 one run was done. Bug 2189.
222 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
223 "panic the current process" to "deliberately defer". The panic log is
224 still written with the problem list name; the mail and reject logs now
225 get a temp-reject line for the message that was being handled, saying
226 something like "domains check lookup or other defer". The SMTP 451
227 message is still "Temporary local problem".
229 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
230 A crafted sequence of BDAT commands could result in in-use memory beeing
231 freed. CVE-2017-16943.
233 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
234 from SMTP input. Previously it was always done; now only done for DATA
235 and not BDAT commands. CVE-2017-16944.
237 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
238 to the message (such as an overlong header line). Previously this was
239 not done and we did not exit BDAT mode. Followon from the previous item
240 though a different problem.
246 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
247 than -2003 did; needs libidn2 in addition to libidn.
249 JH/02 The path option on a pipe transport is now expanded before use.
251 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
252 Patch provided by "Björn", documentation fix added too.
254 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
255 missing a wire-to-host endian conversion.
257 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
258 close after a BDAT command line could be taken as a following command,
259 giving a synch failure. Fix by only checking for synch immediately
260 before acknowledging the chunk.
262 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
263 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
264 macro. Patches provided by Josh Soref.
266 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
267 Previously we did not; the RFC seems ambiguous and VRFY is not listed
268 by IANA as a service extension. However, John Klensin suggests that we
271 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
272 the dkim code may be unix-mode line endings rather than smtp wire-format
273 CRLF, so prepend a CR to any bare LF.
275 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
276 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
278 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
279 the first were themselves being wrongly included in the feed into dkim
280 processing; with most chunk sizes in use this resulted in an incorrect
281 body hash calculated value.
283 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
284 DKIM signature block, for verification. Although advised against by
285 standards it is specifically not ruled illegal.
287 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
289 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
290 missing a body hash (the bh= tag).
292 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
293 It seems that HAProxy sends the Proxy Protocol information in clear and
294 only then does a TLS startup, so do the same.
296 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
297 TCP connections (such as for Spamd) unless the daemon successfully set
298 Fast Open mode on its listening sockets. This fixes breakage seen on
299 too-old kernels or those not configured for Fast Open, at the cost of
300 requiring both directions being enabled for TFO, and TFO never being used
301 by non-daemon-related Exim processes.
303 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
304 endings, at least on the first header line. Try to canonify any that get
305 past that check, despite the cost.
307 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
308 now limited to an arbitrary five deep, while parsing addresses with the
309 strip_excess_angle_brackets option enabled.
311 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
312 instead leave the unprompted TLS handshake in socket buffer for the
313 TLS library to consume.
315 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
317 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
319 JH/16 Drop variables when they go out of scope. Memory management drops a whole
320 region in one operation, for speed, and this leaves assigned pointers
321 dangling. Add checks run only under the testsuite which checks all
322 variables at a store-reset and panics on a dangling pointer; add code
323 explicitly nulling out all the variables discovered. Fixes one known
324 bug: a transport crash, where a dangling pointer for $sending_ip_address
325 originally assigned in a verify callout, is re-used.
327 PP/06 Drop '.' from @INC in various Perl scripts.
329 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
331 PP/08 Reduce a number of compilation warnings under clang; building with
332 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
333 should be warning-free.
335 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
337 HS/01 Fix portability problems introduced by PP/08 for platforms where
338 realloc(NULL) is not equivalent to malloc() [SunOS et al].
340 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
341 chunk. This allows us to accept broken chunked messages. We need a more
342 general solution here.
344 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
345 already-broken messages in the queue.
347 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
349 JH/19 Fix reference counting bug in routing-generated-address tracking.
355 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
356 supports it and a size is available (ie. the sending peer gave us one).
358 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
359 years of being deprecated). The replacements are the ACLs
360 acl_smtp_mime and acl_not_smtp_mime.
362 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
363 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
364 or even an in-clear connection were permitted. Now, if the host lookup
365 was dnssec and dane was requested then the host is only used if the
366 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
367 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
368 if one fails this test.
369 This means that a poorly-configured remote DNS will make it incommunicado;
370 but it protects against a DNS-interception attack on it.
372 JH/04 Bug 1810: make continued-use of an open smtp transport connection
373 non-noisy when a race steals the message being considered.
375 JH/05 If main configuration option tls_certificate is unset, generate a
376 self-signed certificate for inbound TLS connections.
378 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
379 in rewrites and routers.
381 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
382 and logged a warning sing 4.83; now they are a configuration file error.
384 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
385 (lacking @domain). Apply the same qualification processing as RCPT.
387 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
389 JH/10 Support ${sha256:} applied to a string (as well as the previous
392 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
393 a cutthrough deliver is pending, as we always want to make a connection.
394 This also avoids re-routing the message when later placing the cutthrough
395 connection after a verify cache hit.
396 Do not update it with the verify result either.
398 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
399 when routing results in more than one destination address.
401 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
402 signing (which inhibits the cutthrough capability). Previously only
403 the presence of an option was tested; now an expansion evaluating as
404 empty is permissible (obviously it should depend only on data available
405 when the cutthrough connection is made).
407 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
408 the relevant preceding SMTP command did not note the pipelining mode.
410 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
411 Previously they were not counted.
413 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
414 as one having no matching records. Previously we deferred the message
415 that needed the lookup.
417 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
418 distinguished as "(=".
420 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
421 for missing MX records. Previously it only worked for missing A records.
423 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
425 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
426 after the data-go-ahead and data-ack. Patch from Jason Betts.
428 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
429 even for a "none" policy. Patch from Tony Meyer.
431 JH/22 Fix continued use of a connection for further deliveries. If a port was
432 specified by a router, it must also match for the delivery to be
435 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
436 When one of the recipients of a message was unsuitable for the connection
437 (has no matching addresses), we lost track of needing to mark it
438 deferred. As a result mail would be lost.
440 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
442 JH/25 Decoding ACL controls is now done using a binary search; the source code
443 takes up less space and should be simpler to maintain. Merge the ACL
444 condition decode tables also, with similar effect.
446 JH/26 Fix problem with one_time used on a redirect router which returned the
447 parent address unchanged. A retry would see the parent address marked as
448 delivered, so not attempt the (identical) child. As a result mail would
451 JH/27 Fix a possible security hole, wherein a process operating with the Exim
452 UID can gain a root shell. Credit to http://www.halfdog.net/ for
453 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
456 JH/28 Enable {spool,log} filesystem space and inode checks as default.
457 Main config options check_{log,spool}_{inodes,space} are now
458 100 inodes, 10MB unless set otherwise in the configuration.
460 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
461 Previously it only applied to the main-section connection policy
464 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
466 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
467 by me. Added RFC7919 DH primes as an alternative.
469 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
470 are not in the system include path.
472 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
473 GnuTLS, when a session startup failed (eg because the client disconnected)
474 Exim did stdio operations after fclose. This was exposed by a recent
475 change which nulled out the file handle after the fclose.
477 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
478 signed directly by the cert-signing cert, rather than an intermediate
479 OCSP-signing cert. This is the model used by LetsEncrypt.
481 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
483 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
484 an incoming connection.
486 HS/02 Bug 1802: Do not half-close the connection after sending a request
489 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
490 fallback to "prime256v1".
492 JH/34 SECURITY: Use proper copy of DATA command in error message.
493 Could leak key material. Remotely exploitable. CVE-2016-9963.
499 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
500 and 3.4.4 - once the server is enabled to respond to an OCSP request
501 it does even when not requested, resulting in a stapling non-aware
502 client dropping the TLS connection.
504 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
505 support variable-length bit vectors. No functional change.
507 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
508 The I= interface field on outgoing lines is now after the H= remote
509 host field, same as incoming lines. There is a separate
510 outgoing_interface log selector which allows you to disable the
513 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
514 If not running log_selector +smtp_connection the mainlog would be held
515 open indefinitely after a "too many connections" event, including to a
516 deleted file after a log rotate. Leave the per net connection logging
517 leaving it open for efficiency as that will be quickly detected by the
518 check on the next write.
520 HS/01 Bug 1671: Fix post transport crash.
521 Processing the wait-<transport> messages could crash the delivery
522 process if the message IDs didn't exist for some reason. When
523 using 'split_spool_directory=yes' the construction of the spool
524 file name failed already, exposing the same netto behaviour.
526 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
527 mime_regex ACL conditions.
529 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
530 to DSN fail messages (bounces): remote IP, remote greeting, remote response
531 to HELO, local diagnostic string.
533 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
534 log line to debug. Even when configured with a tls authenticator many
535 client connections are expected to not authenticate in this way, so
536 an authenticate fail is not an error.
538 HS/02 Add the Exim version string to the process info. This way exiwhat
539 gives some more detail about the running daemon.
541 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
542 matter for fast-change records such as DNSBLs.
544 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
545 retry record, even if constant. There may be multiple transports with
546 different interface settings and the retry behaviour needs to be kept
549 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
551 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
553 JH/10 Bug 840: fix log_defer_output option of pipe transport
555 JH/11 Bug 830: use same host for all RCPTS of a message, even under
556 hosts_randomize. This matters a lot when combined with mua_wrapper.
558 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
559 ${quote_pgsql:<string>} operator.
561 JH/13 Bug 1708: avoid misaligned access in cached lookup.
563 JH/14 Change header file name for freeradius-client. Relevant if compiling
564 with Radius support; from the Gentoo tree and checked under Fedora.
566 JH/15 Bug 1712: Introduce $prdr_requested flag variable
568 JH/16 Bug 1714: Permit an empty string as expansion result for transport
569 option transport_filter, meaning no filtering.
571 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
573 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
574 defaults to "*" (all hosts). The variable is now available when not built
575 with TLS, default unset, mainly to enable keeping the testsuite sane.
576 If a server certificate is not supplied (via tls_certificate) an error is
577 logged, and clients will find TLS connections fail on startup. Presumably
578 they will retry in-clear.
579 Packagers of Exim are strongly encouraged to create a server certificate
580 at installation time.
582 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
583 with the $config_file variable.
585 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
586 in transport context, after the attempt, and per-recipient. The latter type
587 is per host attempted. The event data is the error message, and the errno
588 information encodes the lookup type (A vs. MX) used for the (first) host,
589 and the trailing two digits of the smtp 4xx response.
591 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
592 to write to mainlog (or rejectlog, paniclog) in the window between file
593 creation and permissions/ownership being changed. Particularly affects
594 installations where exicyclog is run as root, rather than exim user;
595 result is that the running daemon panics and dies.
597 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
599 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
600 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
601 "pri" and "weight". Note that the previous implicit priority given by the
602 list order is no longer honoured.
604 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
607 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
608 by defining SUPPORT_SOCKS.
610 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
611 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
612 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
613 variables are renamed to proxy_{local,external}_{address,port}.
615 JH/27 Move Internationalisation support from Experimental to mainline, enabled
616 for a build by defining SUPPORT_I18N
618 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
619 of the query string, and make ${quote_redis:} do that quoting.
621 JH/29 Move Events support from Experimental to mainline, enabled by default
622 and removable for a build by defining DISABLE_EVENT.
624 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
626 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
627 cached by the daemon.
629 JH/32 Move Redis support from Experimental to mainline, enabled for a build
630 by defining LOOKUP_REDIS. The libhiredis library is required.
632 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
633 keys are given for lookup.
635 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
636 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
637 only supported when built with TLS support. The PolarSSL SHA routines
638 are still used when the TLS library is too old for convenient support.
640 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
641 openssl_options), for security. OpenSSL forces this from version 1.1.0
642 server-side so match that on older versions.
644 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
645 allocation for $value could be released as the expansion processing
646 concluded, but leaving the global pointer active for it.
648 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
649 and to use the domains and local_parts ACL conditions.
651 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
652 incorrectly not doubled on cutthrough transmission, hence seen as a
653 body-termination at the receiving system - resulting in truncated mails.
654 Commonly the sender saw a TCP-level error, and retransmitted the message
655 via the normal store-and-forward channel. This could result in duplicates
656 received - but deduplicating mailstores were liable to retain only the
657 initial truncated version.
659 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
661 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
663 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
664 we're in there, support oversigning also; bug 1309.
666 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
668 HS/04 Add support for keep_environment and add_environment options.
670 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
671 either intentional arithmetic overflow during PRNG, or testing config-
674 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
675 delivery resulted in actual delivery. Cancel cutthrough before DATA
678 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
679 rejects a recipient: pass the reject to the originator.
681 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
682 Many were false-positives and ignorable, but it's worth fixing the
685 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
686 for the new environment-manipulation done at startup. Move the routines
687 from being local to tls.c to being global via the os.c file.
689 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
690 an extract embedded as result-arg for a map, the first arg for extract
691 is unavailable so we cannot tell if this is a numbered or keyed
692 extraction. Accept either.
698 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
701 JH/02 The smtp transport option "multi_domain" is now expanded.
703 JH/03 The smtp transport now requests PRDR by default, if the server offers
706 JH/04 Certificate name checking on server certificates, when exim is a client,
707 is now done by default. The transport option tls_verify_cert_hostnames
708 can be used to disable this per-host. The build option
709 EXPERIMENTAL_CERTNAMES is withdrawn.
711 JH/05 The value of the tls_verify_certificates smtp transport and main options
712 default to the word "system" to access the system default CA bundle.
713 For GnuTLS, only version 3.0.20 or later.
715 JH/06 Verification of the server certificate for a TLS connection is now tried
716 (but not required) by default. The verification status is now logged by
717 default, for both outbound TLS and client-certificate supplying inbound
720 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
723 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
724 Status Notification (bounce) messages are now MIME format per RFC 3464.
725 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
726 under the control of the dsn_advertise_hosts option, and routers may
727 have a dsn_lasthop option.
729 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
730 default, modifiable by a malware= option. The list separator for
731 the options can now be changed in the usual way. Bug 68.
733 JH/10 The smtp_receive_timeout main option is now expanded before use.
735 JH/11 The incoming_interface log option now also enables logging of the
736 local interface on delivery outgoing connections.
738 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
739 if the interface and destination host and port all match.
741 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
744 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
745 Patch from Andrew Lewis.
747 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
748 now supports optional time-restrictions, weighting, and priority
749 modifiers per server. Patch originally by <rommer@active.by>.
751 JH/16 The spamd_address main option now supports a mixed list of local
752 and remote servers. Remote servers can be IPv6 addresses, and
753 specify a port-range.
755 JH/17 Bug 68: The spamd_address main option now supports an optional
756 timeout value per server.
758 JH/18 Bug 1581: Router and transport options headers_add/remove can
759 now have the list separator specified.
761 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
764 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
767 JH/21 Support for the A6 type of dns record is withdrawn.
769 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
770 rather than the verbs used.
772 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
773 from 255 to 1024 chars.
775 JH/24 Verification callouts now attempt to use TLS by default.
777 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
778 are generic router options now. The defaults didn't change.
780 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
781 Original patch from Alexander Shikoff, worked over by JH.
783 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
784 files if ZCAT_COMMAND is not executable.
786 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
788 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
790 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
791 Normally benign, it bites when the pair was led to by a CNAME;
792 modern usage is to not canonicalize the domain to a CNAME target
793 (and we were inconsistent anyway for A-only vs AAAA+A).
795 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
797 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
798 when evaluating $sender_host_dnssec.
800 JH/31 Check the HELO verification lookup for DNSSEC, adding new
801 $sender_helo_dnssec variable.
803 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
805 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
807 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
809 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
810 documented as working, but never had. Support all but $spam_report.
812 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
813 added for tls authenticator.
815 HS/03 Add perl_taintmode main config option
821 TL/01 When running the test suite, the README says that variables such as
822 no_msglog_check are global and can be placed anywhere in a specific
823 test's script, however it was observed that placement needed to be near
824 the beginning for it to behave that way. Changed the runtest perl
825 script to read through the entire script once to detect and set these
826 variables, reset to the beginning of the script, and then run through
827 the script parsing/test process like normal.
829 TL/02 The BSD's have an arc4random API. One of the functions to induce
830 adding randomness was arc4random_stir(), but it has been removed in
831 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
832 function when detected.
834 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
835 cause callback expansion.
837 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
838 syntax errors in an expansion can be treated as a string instead of
839 logging or causing an error, due to the internal use of bool_lax
840 instead of bool when processing it.
842 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
843 server certificates when making smtp deliveries.
845 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
847 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
849 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
851 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
852 Merged patch from Sebastian Wiedenroth.
854 JH/05 Fix results-pipe from transport process. Several recipients, combined
855 with certificate use, exposed issues where response data items split
856 over buffer boundaries were not parsed properly. This eventually
857 resulted in duplicates being sent. This issue only became common enough
858 to notice due to the introduction of connection certificate information,
859 the item size being so much larger. Found and fixed by Wolfgang Breyha.
861 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
862 size buffer was used, resulting in syntax errors when an expansion
865 JH/07 Add support for directories of certificates when compiled with a GnuTLS
866 version 3.3.6 or later.
868 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
869 is EXPERIMENTAL_EVENT, the main-configuration and transport options
870 both become "event_action", the variables become $event_name, $event_data
871 and $event_defer_errno. There is a new variable $verify_mode, usable in
872 routers, transports and related events. The tls:cert event is now also
873 raised for inbound connections, if the main configuration event_action
876 TL/06 In test suite, disable OCSP for old versions of openssl which contained
877 early OCSP support, but no stapling (appears to be less than 1.0.0).
879 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
880 server certificate names available under the smtp transport option
881 "tls_verify_cert_hostname" now do not permit multi-component wildcard
884 JH/10 Time-related extraction expansions from certificates now use the main
885 option "timezone" setting for output formatting, and are consistent
886 between OpenSSL and GnuTLS compilations. Bug 1541.
888 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
889 encoded parameter in the incoming message. Bug 1558.
891 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
892 include certificate info, eximon was claiming there were spoolfile
895 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
897 JH/14 Log delivery-related information more consistently, using the sequence
898 "H=<name> [<ip>]" wherever possible.
900 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
901 are problematic for Debian distribution, omit them from the release
904 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
906 JH/16 Fix string representation of time values on 64bit time_t architectures.
909 JH/17 Fix a null-indirection in certextract expansions when a nondefault
910 output list separator was used.
915 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
916 checkers that were complaining about end of non-void function with no
919 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
920 This was a regression introduced in 4.83 by another bugfix.
922 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
924 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
925 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
931 TF/01 Correctly close the server side of TLS when forking for delivery.
933 When a message was received over SMTP with TLS, Exim failed to clear up
934 the incoming connection properly after forking off the child process to
935 deliver the message. In some situations the subsequent outgoing
936 delivery connection happened to have the same fd number as the incoming
937 connection previously had. Exim would try to use TLS and fail, logging
938 a "Bad file descriptor" error.
940 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
941 utilities have not been installed.
943 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
944 temporary space as the ACL may create new global variables.
946 TL/01 LDAP support uses per connection or global context settings, depending
947 upon the detected version of the libraries at build time.
949 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
950 to extract and use the src ip:port in logging and expansions as if it
951 were a direct connection from the outside internet. PPv2 support was
952 updated based on HAProxy spec change in May 2014.
954 JH/02 Add ${listextract {number}{list}{success}{fail}}.
956 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
957 Properly escape header and check for NULL return.
959 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
962 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
964 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
965 characters in header names, implemented as a verify condition.
966 Contributed by Michael Fischer v. Mollard.
968 TL/05 Rename SPF condition results err_perm and err_temp to standardized
969 results permerror and temperror. Previous values are deprecated but
970 still accepted. In a future release, err_perm and err_temp will be
971 completely removed, which will be a backward incompatibility if the
972 ACL tests for either of these two old results. Patch contributed by
973 user bes-internal on the mailing list.
975 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
977 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
978 selectors, in both main and reject logs.
980 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
983 JH/07 Add malware type "sock" for talking to simple daemon.
985 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
987 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
988 routers/transports under cutthrough routing.
990 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
991 numbers. Touch up "bool" conditional to keep the same definition.
993 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
995 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
997 JH/12 Expand items in router/transport headers_add or headers_remove lists
998 individually rather than the list as a whole. Bug 1452.
1000 Required for reasonable handling of multiple headers_ options when
1001 they may be empty; requires that headers_remove items with embedded
1002 colons must have them doubled (or the list-separator changed).
1004 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1005 view the policy declared in the DMARC record. Currently, $dmarc_status
1006 is a combined value of both the record presence and the result of the
1009 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1011 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1012 dnslookup router and the smtp transport (applying to the forward
1015 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1016 of ldap servers used for a specific lookup. Patch provided by Heiko
1019 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1020 New variable $lookup_dnssec_authenticated for observability.
1022 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1023 Patch submitted by Lars Timman.
1025 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1027 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1028 Requires trusted mode and valid format message id, aborts otherwise.
1029 Patch contributed by Heiko Schlichting.
1031 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1032 certextract with support for various fields. Bug 1358.
1034 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1035 is requested by default, modifiable by smtp transport option
1038 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1039 operate on certificate variables to give certificate fingerprints
1040 Also new ${sha256:cert_variable}.
1042 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1044 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1045 Christian Aistleitner.
1047 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1049 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1050 file. Patch from Wolfgang Breyha.
1052 JH/25 Expand the coverage of the delivery $host and $host_address to
1053 client authenticators run in verify callout. Bug 1476.
1055 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1056 align with daemon_smtp_ports. Bug 72.
1058 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1059 support and error reporting did not work properly.
1061 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1062 and is readable. Patch from Andrew Colin Kissa.
1064 TL/14 Enhance documentation of ${run expansion and how it parses the
1065 commandline after expansion, particularly in the case when an
1066 unquoted variable expansion results in an empty value.
1068 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1070 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1071 by renaming away from it. Addresses GH issue 15, reported by
1074 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1076 TL/15 SECURITY: prevent double expansion in math comparison functions
1077 (can expand unsanitized data). Not remotely exploitable.
1084 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1086 PP/02 Make -n do something, by making it not do something.
1087 When combined with -bP, the name of an option is not output.
1089 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1092 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1093 $sender_host_name and config options to manage this, and basic check
1096 PP/05 DSCP support for outbound connections and control modifier for inbound.
1098 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1099 (Only plugin which currently uses this is kerberos4, which nobody should
1100 be using, but we should make it available and other future plugins might
1101 conceivably use it, even though it would break NAT; stuff *should* be
1102 using channel bindings instead).
1104 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1105 name; added for Sendmail compatibility; requires admin caller.
1106 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1107 just ignore it); requires trusted caller.
1108 Also parse but ignore: -Ac -Am -X<logfile>
1111 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1113 TL/02 Add +smtp_confirmation as a default logging option.
1115 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1116 Patch by Magnus Holmgren from 2007-02-20.
1118 TL/04 Bugzilla 1281 - Spec typo.
1119 Bugzilla 1283 - Spec typo.
1120 Bugzilla 1290 - Spec grammar fixes.
1122 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1124 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1126 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1127 mailing list by by Dmitry Isaikin.
1129 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1131 JH/02 Support "G" suffix to numbers in ${if comparisons.
1133 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1135 NM/01 Bugzilla 1197 - Spec typo
1136 Bugzilla 1196 - Spec examples corrections
1138 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1140 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1141 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1144 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1145 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1147 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1148 "acl {{name}{arg}...}", and optional args on acl condition
1151 JH/05 Permit multiple router/transport headers_add/remove lines.
1153 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1155 JH/07 Avoid using a waiting database for a single-message-only transport.
1156 Performance patch from Paul Fisher. Bugzilla 1262.
1158 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1161 JH/09 Add $headers_added variable, with content from use of ACL modifier
1162 add_header (but not yet added to the message). Bugzilla 199.
1164 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1165 Pulled from Bugzilla 817 by Wolfgang Breyha.
1167 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1169 (nb: this is the same fix as in Exim 4.80.1)
1171 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1174 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1177 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1179 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1180 advertise SMTP AUTH mechanism to us, instead of a generic
1181 protocol violation error. Also, make Exim more robust to bad
1182 data from the Dovecot auth socket.
1184 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1186 When a queue runner is handling a message, Exim first routes the
1187 recipient addresses, during which it prunes them based on the retry
1188 hints database. After that it attempts to deliver the message to
1189 any remaining recipients. It then updates the hints database using
1192 So if a recipient address works intermittently, it can get repeatedly
1193 deferred at routing time. The retry hints record remains fresh so the
1194 address never reaches the final cutoff time.
1196 This is a fairly common occurrence when a user is bumping up against
1197 their storage quota. Exim had some logic in its local delivery code
1198 to deal with this. However it did not apply to per-recipient defers
1199 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1201 This change adds a proper retry rule check during routing so that the
1202 final cutoff time is checked against the message's age. We only do
1203 this check if there is an address retry record and there is not a
1204 domain retry record; this implies that previous attempts to handle
1205 the address had the retry_use_local_parts option turned on. We use
1206 this as an approximation for the destination being like a local
1207 delivery, as in LMTP.
1209 I suspect this new check makes the old local delivery cutoff check
1210 redundant, but I have not verified this so I left the code in place.
1212 TF/02 Correct gecos expansion when From: is a prefix of the username.
1214 Test 0254 submits a message to Exim with the header
1218 When I ran the test suite under the user fanf2, Exim expanded
1219 the header to contain my full name, whereas it should have added
1220 a Resent-Sender: header. It erroneously treats any prefix of the
1221 username as equal to the username.
1223 This change corrects that bug.
1225 GF/01 DCC debug and logging tidyup
1226 Error conditions log to paniclog rather than rejectlog.
1227 Debug lines prefixed by "DCC: " to remove any ambiguity.
1229 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1231 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1232 Bug spotted by Jeremy Harris; was flawed since initial commit.
1233 Would have resulted in OCSP responses post-SNI triggering an Exim
1234 NULL dereference and crash.
1236 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1238 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1239 Bug detection, analysis and fix by Samuel Thibault.
1240 Bugzilla 1331, Debian bug #698092.
1242 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1244 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1245 Server implementation by Todd Lyons, client by JH.
1246 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1247 config variable "prdr_enable" controls whether the server
1248 advertises the facility. If the client requests PRDR a new
1249 acl_data_smtp_prdr ACL is called once for each recipient, after
1250 the body content is received and before the acl_smtp_data ACL.
1251 The client is controlled by both of: a hosts_try_prdr option
1252 on the smtp transport, and the server advertisement.
1253 Default client logging of deliveries and rejections involving
1254 PRDR are flagged with the string "PRDR".
1256 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1257 fclose(). Diagnosis by Todd Lyons.
1259 PP/17 Update configure.default to handle IPv6 localhost better.
1260 Patch by Alain Williams (plus minor tweaks).
1263 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1264 This is now consistent with GnuTLS, and is now documented: the
1265 previous undocumented portable approach to treating the option as
1266 unset was to force an expansion failure. That still works, and
1267 an empty string is now equivalent.
1269 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1270 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1271 not performing validation itself.
1273 PP/20 Added force_command boolean option to pipe transport.
1274 Patch from Nick Koston, of cPanel Inc.
1276 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1279 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1281 PP/21 Fix eximon continuous updating with timestamped log-files.
1282 Broken in a format-string cleanup in 4.80, missed when I repaired the
1283 other false fix of the same issue.
1284 Report and fix from Heiko Schlichting.
1287 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1288 Report from Prashanth Katuri.
1290 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1291 It's SecureTransport, so affects any MacOS clients which use the
1292 system-integrated TLS libraries, including email clients.
1294 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1295 using a MIME ACL for non-SMTP local injection.
1296 Report and assistance in diagnosis by Warren Baker.
1298 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1300 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1302 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1303 last id that failed so it may be referenced in subsequent ACL's.
1305 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1308 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1309 ldap library initialization, allowing self-signed CA's to be
1310 used. Also properly sets require_cert option later in code by
1311 using NULL (global ldap config) instead of ldap handle (per
1312 session). Bug diagnosis and testing by alxgomz.
1314 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1315 the src/util/ subdirectory.
1317 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1318 renamed to Transport Post Delivery Action by Jeremy Harris, as
1321 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1322 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1323 redis_servers = needs to be configured which will be used by the redis
1324 lookup. Patch from Warren Baker, of The Packet Hub.
1326 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1328 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1329 hostname or reverse DNS when processing a host list. Used suggestions
1330 from multiple comments on this bug.
1332 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1334 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1335 Missed a few lines, added it to make the runtest require no keyboard
1338 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1339 contains upper case chars. Make router use caseful_local_part.
1341 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1342 support when GnuTLS has been built with p11-kit.
1348 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1350 This, or similar/improved, will also be change PP/11 of 4.82.
1356 PP/01 Handle short writes when writing local log-files.
1357 In practice, only affects FreeBSD (8 onwards).
1358 Bugzilla 1053, with thanks to Dmitry Isaikin.
1360 NM/01 Bugzilla 949 - Documentation tweak
1362 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1365 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1367 PP/02 Implemented gsasl authenticator.
1369 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1371 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1372 `pkg-config foo` for cflags/libs.
1374 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1375 with rest of GSASL and with heimdal_gssapi.
1377 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1378 `pkg-config foo` for cflags/libs for the TLS implementation.
1380 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1381 properties get this fed in as external SSF. A number of robustness
1382 and debugging improvements to the cyrus_sasl authenticator.
1384 PP/08 cyrus_sasl server now expands the server_realm option.
1386 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1387 Patch by Jeremy Harris.
1389 PP/10 Added dbmjz lookup type.
1391 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1393 PP/12 MAIL args handles TAB as well as SP, for better interop with
1394 non-compliant senders.
1395 Analysis and variant patch by Todd Lyons.
1397 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1398 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1399 Patch from Dirk Mueller <dmueller@suse.com>
1401 PP/13 tls_peerdn now print-escaped for spool files.
1402 Observed some $tls_peerdn in wild which contained \n, which resulted
1403 in spool file corruption.
1405 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1406 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1407 or write after TLS renegotiation, which otherwise led to messages
1410 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1411 as a tracking header (ie: a signed header comes before the signature).
1412 Patch from Wolfgang Breyha.
1414 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1415 comma-sep list; embedded commas doubled.
1417 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1419 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1421 Report and patch from Dmitry Banschikov.
1423 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1424 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1425 failures. We appear to now support TLS1.1+ with Exim.
1427 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1428 lets Exim select keys and certificates based upon TLS SNI from client.
1429 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1430 before an outbound SMTP session. New log_selector, +tls_sni.
1432 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1433 NULL dereference. Report and patch from Alun Jones.
1435 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1436 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1437 Not seeing resolver debug output on NetBSD, but suspect this is a
1438 resolver implementation change.
1440 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1441 Left warnings. Added "eximon gdb" invocation mode.
1443 PP/21 Defaulting "accept_8bitmime" to true, not false.
1445 PP/22 Added -bw for inetd wait mode support.
1447 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1448 locate the relevant includes and libraries. Made this the default.
1450 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1451 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1453 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1454 This may cause build issues on older platforms.
1456 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1457 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1458 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1459 Added SNI support via GnuTLS too.
1460 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1462 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1464 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1465 Applied second patch from Janne, implementing suggestion to default
1466 multiple-strings-in-record handling to match SPF spec.
1468 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1470 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1471 read-only, out of scope).
1472 Patch from Wolfgang Breyha, report from Stuart Northfield.
1474 PP/29 Fix three issues highlighted by clang analyser static analysis.
1475 Only crash-plausible issue would require the Cambridge-specific
1476 iplookup router and a misconfiguration.
1477 Report from Marcin Mirosław.
1479 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1481 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1482 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1483 As part of this, removing so much warning spew let me fix some minor
1484 real issues in debug logging.
1486 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1487 assignment on my part. Fixed.
1489 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1490 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1491 Janne Snabb (who went above and beyond: thank you).
1493 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1494 string otherwise requires a connection and a bunch more work and it's
1495 relatively easy to get wrong. Should also expose TLS library linkage
1498 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1499 64-bit ${eval} (JH/03).
1501 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1502 GNU libc to support some of the 64-bit stuff, should not lead to
1503 conflicts. Defined before os.h is pulled in, so if a given platform
1504 needs to override this, it can.
1506 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1507 protection layer was required, which is not implemented.
1508 Bugzilla 1254, patch from Wolfgang Breyha.
1510 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1511 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1512 tls_dhparam take prime identifiers. Also unbreak combination of
1513 OpenSSL+DH_params+TLSSNI.
1515 PP/39 Disable SSLv2 by default in OpenSSL support.
1521 PP/01 Solaris build fix for Oracle's LDAP libraries.
1522 Bugzilla 1109, patch from Stephen Usher.
1524 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1526 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1529 TF/02 Fix a couple more cases where we did not log the error message
1530 when unlink() failed. See also change 4.74-TF/03.
1532 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1533 lock up or crash if it happened to be inside a call to libc when it
1534 got a SIGUSR1 from exiwhat.
1536 The SIGUSR1 handler appends the current process status to the process
1537 log which is later printed by exiwhat. It used to use the general
1538 purpose logging code to do this, but several functions it calls are
1539 not safe for signals.
1541 The new output code in the SIGUSR1 handler is specific to the process
1542 log, and simple enough that it's easy to inspect for signal safety.
1543 Removing some special cases also simplifies the general logging code.
1544 Removing the spurious timestamps from the process log simplifies
1547 TF/04 Improved ratelimit ACL condition.
1549 The /noupdate option has been deprecated in favour of /readonly which
1550 has clearer semantics. The /leaky, /strict, and /readonly update modes
1551 are mutually exclusive. The update mode is no longer included in the
1552 database key; it just determines when the database is updated. (This
1553 means that when you upgrade Exim will forget old rate measurements.)
1555 Exim now checks that the per_* options are used with an update mode that
1556 makes sense for the current ACL. For example, when Exim is processing a
1557 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1558 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1559 must specify per_mail/readonly. If you omit the update mode it defaults to
1560 /leaky where that makes sense (as before) or /readonly where required.
1562 The /noupdate option is now undocumented but still supported for
1563 backwards compatibility. It is equivalent to /readonly except that in
1564 ACLs where /readonly is required you may specify /leaky/noupdate or
1565 /strict/noupdate which are treated the same as /readonly.
1567 A useful new feature is the /count= option. This is a generalization
1568 of the per_byte option, so that you can measure the throughput of other
1569 aggregate values. For example, the per_byte option is now equivalent
1570 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1572 The per_rcpt option has been generalized using the /count= mechanism
1573 (though it's more complicated than the per_byte equivalence). When it is
1574 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1575 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1576 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1577 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1578 Note that using per_rcpt with a non-readonly update mode in more than
1579 one ACL will cause the recipients to be double-counted. (The per_mail
1580 and per_byte options don't have this problem.)
1582 The handling of very low rates has changed slightly. If the computed rate
1583 is less than the event's count (usually one) then this event is the first
1584 after a long gap. In this case the rate is set to the same as this event's
1585 count, so that the first message of a spam run is counted properly.
1587 The major new feature is a mechanism for counting the rate of unique
1588 events. The new per_addr option counts the number of different
1589 recipients that someone has sent messages to in the last time period. It
1590 behaves like per_rcpt if all the recipient addresses are different, but
1591 duplicate recipient addresses do not increase the measured rate. Like
1592 the /count= option this is a general mechanism, so the per_addr option
1593 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1594 example, measure the rate that a client uses different sender addresses
1595 with the options per_mail/unique=$sender_address. There are further
1596 details in the main documentation.
1598 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1600 TF/06 Removed a few PCRE remnants.
1602 TF/07 Automatically extract Exim's version number from tags in the git
1603 repository when doing development or release builds.
1605 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1606 Bugzilla 879. Patch from Paul Fisher.
1608 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1609 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1612 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1614 PP/05 Variable $av_failed, true if the AV scanner deferred.
1615 Bugzilla 1078. Patch from John Horne.
1617 PP/06 Stop make process more reliably on build failure.
1618 Bugzilla 1087. Patch from Heiko Schlittermann.
1620 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1621 Bugzilla 1089. Patch from Heiko Schlittermann.
1623 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1624 Bugzilla 1131. Patch from Holger Weiß.
1626 PP/09 Handle IPv6 addresses with SPF.
1627 Bugzilla 860. Patch from Wolfgang Breyha.
1629 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1631 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1634 PP/11 match_* no longer expand right-hand-side by default.
1635 New compile-time build option, EXPAND_LISTMATCH_RHS.
1636 New expansion conditions, "inlist", "inlisti".
1638 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1640 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1642 PP/14 fix log_write() format string regression from TF/03.
1643 Bugzilla 1152. Patch from Dmitry Isaikin.
1649 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1651 PP/02 Harmonised TLS library version reporting; only show if debugging.
1652 Layout now matches that introduced for other libraries in 4.74 PP/03.
1654 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1656 PP/04 New "dns_use_edns0" global option.
1658 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1661 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1662 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1664 TK/01 Updated PolarSSL code to 0.14.2.
1665 Bugzilla 1097. Patch from Andreas Metzler.
1667 PP/07 Catch divide-by-zero in ${eval:...}.
1668 Fixes bugzilla 1102.
1670 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1673 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1674 format-string attack -- SECURITY: remote arbitrary code execution.
1676 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1677 time unintentionally subject to list matching rules, letting the header
1678 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1679 arbitrary string expansion). This allowed for information disclosure.
1681 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1682 INT_MIN/-1 -- value coerced to INT_MAX.
1688 NM/01 Workaround for PCRE version dependency in version reporting
1691 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1692 This fixes portability to compilers other than gcc, notably
1693 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1695 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1696 makefiles for portability to HP-UX and POSIX correctness.
1698 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1699 Also via indented variable definition in the Makefile.
1700 (Debugging by Oliver Heesakkers).
1702 PP/02 Restore caching of spamd results with expanded spamd_address.
1703 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1705 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1706 Improves build reliability. Fix from: Frank Elsner
1708 NM/02 Fix wide character breakage in the rfc2047 coding
1709 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1711 NM/03 Allow underscore in dnslist lookups
1712 Fixes bug 1026. Patch from Graeme Fowler
1714 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1715 Code patches from Adam Ciarcinski of NetBSD.
1717 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1720 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1721 is logged, to avoid truncation. Patch from John Horne.
1723 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1724 Patch from Jakob Hirsch.
1726 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1727 SQL string expansion failure details.
1728 Patch from Andrey Oktyabrski.
1730 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1731 Patch from Simon Arlott.
1733 PP/09 New lookups functionality failed to compile on old gcc which rejects
1734 extern declarations in function scope.
1735 Patch from Oliver Fleischmann
1737 PP/10 Use sig_atomic_t for flags set from signal handlers.
1738 Check getgroups() return and improve debugging.
1739 Fixed developed for diagnosis in bug 927 (which turned out to be
1742 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1743 Patch from Mark Zealey.
1745 PP/12 Bugzilla 1056: Improved spamd server selection.
1746 Patch from Mark Zealey.
1748 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1749 Based on patch from Heiko Schlittermann.
1751 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1752 Patch from Uwe Doering, sign-off by Michael Haardt.
1754 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1755 variable declaration deep within a block. Bug and patch from
1758 PP/15 lookups-Makefile IRIX compatibility coercion.
1760 PP/16 Make DISABLE_DKIM build knob functional.
1762 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1763 Patch by Simon Arlott
1765 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1766 variable argument macros. Our copy now differs from upstream.
1772 TF/01 Failure to get a lock on a hints database can have serious
1773 consequences so log it to the panic log.
1775 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1776 controlled using the smtp_confirmation log selector.
1778 TF/03 Include the error message when we fail to unlink a spool file.
1780 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1781 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1782 for maintaining out-of-tree patches for some time.
1784 PP/01 Bugzilla 139: Documentation and portability issues.
1785 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1786 Handle per-OS dynamic-module compilation flags.
1788 PP/02 Let /dev/null have normal permissions.
1789 The 4.73 fixes were a little too stringent and complained about the
1790 permissions on /dev/null. Exempt it from some checks.
1791 Reported by Andreas M. Kirchwitz.
1793 PP/03 Report version information for many libraries, including
1794 Exim version information for dynamically loaded libraries. Created
1795 version.h, now support a version extension string for distributors
1796 who patch heavily. Dynamic module ABI change.
1798 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1799 privilege escalation vulnerability whereby the Exim run-time user
1800 can cause root to append content of the attacker's choosing to
1803 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1806 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1807 If dropping privileges for untrusted macros, we disabled normal logging
1808 on the basis that it would fail; for the Exim run-time user, this is not
1809 the case, and it resulted in successful deliveries going unlogged.
1810 Fixed. Reported by Andreas Metzler.
1816 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1817 only prepend for the Resent-* case. Fixes regression introduced in
1818 Exim 4.70 by NM/22 for Bugzilla 607.
1820 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1821 increasing numbers of administrators be bitten by this.
1823 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1825 PP/03 Bugzilla 994: added openssl_options main configuration option.
1827 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1829 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1831 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1833 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1834 without a peer certificate, leading to a segfault because of an
1835 assumption that peers always have certificates. Be a little more
1836 paranoid. Problem reported by Martin Tscholak.
1838 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1839 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1840 NB: ClamAV planning to remove STREAM in "middle of 2010".
1841 CL also introduces -bmalware, various -d+acl logging additions and
1842 more caution in buffer sizes.
1844 PP/09 Implemented reverse_ip expansion operator.
1846 PP/10 Bugzilla 937: provide a "debug" ACL control.
1848 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1850 PP/12 Bugzilla 973: Implement --version.
1852 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1854 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1856 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1858 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1859 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1860 ignore trailing whitespace.
1862 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1864 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1867 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1868 Notification from Dr Andrew Aitchison.
1870 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1871 ExtendedDetectionInfo response format.
1872 Notification from John Horne.
1874 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1877 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1878 XSL and documented dependency on system catalogs, with examples of how
1881 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1884 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1885 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1886 configuration file which is writeable by the Exim user or group.
1888 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1889 of configuration files to cover files specified with the -C option if
1890 they are going to be used with root privileges, not just the default
1893 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1894 option (effectively making it always true).
1896 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1897 files to be used while preserving root privileges.
1899 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1900 that rogue child processes cannot use them.
1902 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1903 run-time user, instead of root.
1905 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1906 Exim run-time user without dropping privileges.
1908 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1909 result string, instead of calling string_vformat() twice with the same
1912 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1913 for other users. Others should always drop root privileges if they use
1914 -C on the command line, even for a whitelisted configure file.
1916 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1918 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1924 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1925 $data_path, and $header_path variables; fixed documentation bugs and
1928 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1929 exipick to access non-standard spools, including the "frozen" queue
1932 NM/01 Bugzilla 965: Support mysql stored procedures.
1933 Patch from Alain Williams
1935 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1937 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1938 Patch from Andreas Metzler
1940 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1941 Patch from Kirill Miazine
1943 NM/05 Bugzilla 671: Added umask to procmail example.
1945 JJ/03 installed exipick 20100323.0, fixing doc bug
1947 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1948 directory. Notification and patch from Dan Rosenberg.
1950 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1952 TK/02 Improve log output when DKIM signing operation fails.
1954 MH/01 Treat the transport option dkim_domain as a colon separated
1955 list, not as a single string, and sign the message with each element,
1956 omitting multiple occurences of the same signer.
1958 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1959 Bugzilla 985, 986. Patch by Simon Arlott
1961 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1962 Patch by Simon Arlott
1964 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1965 MBX locking. Notification from Dan Rosenberg.
1971 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1973 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1975 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1977 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1979 NM/04 Bugzilla 915: Flag broken perl installation during build.
1985 TK/01 Added patch by Johannes Berg that expands the main option
1986 "spamd_address" if it starts with a dollar sign.
1988 TK/02 Write list of recipients to X-Envelope-Sender header when building
1989 the mbox-format spool file for content scanning (suggested by Jakob
1992 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1993 (http://www.dcc-servers.net/) support via dccifd. Activated by
1994 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1996 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1997 by Mark Daniel Reidel <mr@df.eu>.
1999 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2000 When building exim an external PCRE library is now needed -
2001 PCRE is a system library on the majority of modern systems.
2002 See entry on PCRE_LIBS in EDITME file.
2004 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2005 conversation. Added nologin parameter to request.
2006 Patch contributed by Kirill Miazine.
2008 TF/01 Do not log submission mode rewrites if they do not change the address.
2010 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2012 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2013 log files in place. Contributed by Roberto Lima.
2015 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2017 TF/03 Bugzilla 615: When checking the local_parts router precondition
2018 after a local_part_suffix or local_part_prefix option, Exim now
2019 does not use the address's named list lookup cache, since this
2020 contains cached lookups for the whole local part.
2022 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2023 Robert Millan. Documentation is in experimental-spec.txt.
2025 TF/04 Bugzilla 668: Fix parallel build (make -j).
2027 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2029 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2030 Patch provided by Jan Srzednicki.
2032 TF/05 Leading white space used to be stripped from $spam_report which
2033 wrecked the formatting. Now it is preserved.
2035 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2036 that they are available at delivery time.
2038 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2040 TF/08 TLS error reporting now respects the incoming_interface and
2041 incoming_port log selectors.
2043 TF/09 Produce a more useful error message if an SMTP transport's hosts
2044 setting expands to an empty string.
2046 NM/06 Bugzilla 744: EXPN did not work under TLS.
2047 Patch provided by Phil Pennock.
2049 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2050 Patch provided by Richard Godbee.
2052 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2053 acl_smtp_notquit, added index entry.
2055 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2056 Patch provided by Eugene Bujak.
2058 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2059 accept(). Patch provided by Maxim Dounin.
2061 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2062 Patch provided by Phil Pennock.
2064 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2066 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2067 Patch provided by Brad "anomie" Jorsch.
2069 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2070 Patch provided by Dean Brooks.
2072 TK/05 Add native DKIM support (does not depend on external libraries).
2074 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2075 Patch provided by Graeme Fowler.
2077 NM/16 Bugzilla 851: Documentation example syntax fix.
2079 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2081 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2084 NM/19 Bugzilla 745: TLS version reporting.
2085 Patch provided by Phil Pennock.
2087 NM/20 Bugzilla 167: bool: condition support.
2088 Patch provided by Phil Pennock.
2090 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2091 clients. Patch provided by Phil Pennock.
2093 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2094 Patch provided by Brad "anomie" Jorsch.
2096 NM/23 Bugzilla 687: Fix misparses in eximstats.
2097 Patch provided by Heiko Schlittermann.
2099 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2100 Patch provided by Heiko Schlittermann.
2102 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2103 plus update to original patch.
2105 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2107 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2108 Patch provided by David Brownlee.
2110 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2112 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2114 NM/30 Bugzilla 888: TLS documentation bugfixes.
2116 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2118 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2119 Unlike the original bugzilla I have changed all shell scripts in src tree.
2121 NM/33 Bugzilla 898: Transport filter timeout fix.
2122 Patch by Todd Rinaldo.
2124 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2125 Patch by Serge Demonchaux.
2127 NM/35 Bugzilla 39: Base64 decode bug fixes.
2128 Patch by Jakob Hirsch.
2130 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2132 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2134 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2136 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2142 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2143 ALT-N's libdkim that I have put here:
2144 http://duncanthrax.net/exim-experimental/
2146 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2147 were called 'true' and it seems that C99 defines that as a reserved
2148 keyword to be used with 'bool' variable types. That means you could
2149 not include C99-style headers which use bools without triggering
2150 build errors in sieve.c.
2152 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2153 as mailq or other aliases. Changed the --help handling significantly
2154 to do whats expected. exim_usage() emits usage/help information.
2156 SC/01 Added the -bylocaldomain option to eximstats.
2158 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2160 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2162 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2168 PH/01 Another patch from the Sieve maintainer.
2170 PH/02 When an IPv6 address is converted to a string for single-key lookup
2171 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2172 dots are used instead of colons so that keys in lsearch files need not
2173 contain colons. This was done some time before quoting was made available
2174 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2175 (notated using the quote facility) so as to distinguish them from IPv4
2176 keys. This meant that lookups for IP addresses in host lists did not work
2177 for iplsearch lookups.
2179 This has been fixed by arranging for IPv6 addresses to be expressed with
2180 colons if the lookup type is iplsearch. This is not incompatible, because
2181 previously such lookups could never work.
2183 The situation is now rather anomalous, since one *can* have colons in
2184 ordinary lsearch keys. However, making the change in all cases is
2185 incompatible and would probably break a number of configurations.
2187 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2190 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2191 conversion specification without a maximum field width, thereby enabling
2192 a rogue spamd server to cause a buffer overflow. While nobody in their
2193 right mind would setup Exim to query an untrusted spamd server, an
2194 attacker that gains access to a server running spamd could potentially
2195 exploit this vulnerability to run arbitrary code as the Exim user.
2197 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2198 $primary_hostname instead of what libspf2 thinks the hosts name is.
2200 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2201 a directory entry by the name of the lookup key. Previously, if a
2202 symlink pointed to a non-existing file or a file in a directory that
2203 Exim lacked permissions to read, a lookup for a key matching that
2204 symlink would fail. Now it is enough that a matching directory entry
2205 exists, symlink or not. (Bugzilla 503.)
2207 PH/03 The body_linecount and body_zerocount variables are now exported in the
2210 PH/04 Added the $dnslist_matched variable.
2212 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2213 This means they are set thereafter only if the connection becomes
2216 PH/06 Added the client_condition to authenticators so that some can be skipped
2217 by clients under certain conditions.
2219 PH/07 The error message for a badly-placed control=no_multiline_responses left
2220 "_responses" off the end of the name.
2222 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2224 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2225 (without spaces) instead of just copying the configuration text.
2227 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2229 PH/11 Added $max_received_linelength.
2231 PH/12 Added +ignore_defer and +include_defer to host lists.
2233 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2234 way in which PCRE > 7.0 is built.
2236 PH/14 Implemented queue_only_load_latch.
2238 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2239 MAIL command. The effect was to mangle the value on 64-bit systems.
2241 PH/16 Another patch from the Sieve maintainer.
2243 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2245 PH/18 If a system quota error occurred while trying to create the file for
2246 a maildir delivery, the message "Mailbox is full" was not appended to the
2247 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2248 only to a quota excession during the actual writing of the file.
2250 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2251 characters?) which causes problems in log lines. The DN values are now
2252 passed through string_printing() before being added to log lines.
2254 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2255 and InterBase are left for another time.)
2257 PH/21 Added message_body_newlines option.
2259 PH/22 Guard against possible overflow in moan_check_errorcopy().
2261 PH/23 POSIX allows open() to be a macro; guard against that.
2263 PH/24 If the recipient of an error message contained an @ in the local part
2264 (suitably quoted, of course), incorrect values were put in $domain and
2265 $local_part during the evaluation of errors_copy.
2271 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2272 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2275 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2276 issue a MAIL command.
2278 PH/02 In an ACL statement such as
2280 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2282 if a client was not listed at all, or was listed with a value other than
2283 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2284 the condition was not true (as it should be), so access was not denied.
2285 The bug was that the ! inversion was incorrectly passed on to the second
2286 item. This has been fixed.
2288 PH/03 Added additional dnslists conditions == and =& which are different from
2289 = and & when the dns lookup returns more than one IP address.
2291 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2292 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2294 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2295 FSYNC, which compiles an option called disable_fsync that allows for
2296 bypassing fsync(). The documentation is heavily laced with warnings.
2298 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2300 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2301 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2302 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2303 including adding "make clean"; (3) Added -fPIC when compiling the test
2304 dynamically loaded module, to get rid of a warning.
2306 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2307 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2308 = 0s. The bug is otherwise harmless.
2310 PH/07 There was a bug in the dovecot authenticator such that the value of
2311 $auth1 could be overwritten, and so not correctly preserved, after a
2312 successful authentication. This usually meant that the value preserved by
2313 the server_setid option was incorrect.
2315 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2317 PH/09 Installed PCRE release 7.0.
2319 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2320 run for batched SMTP input. It is now run at the start of every message
2321 in the batch. While fixing this I discovered that the process information
2322 (output by running exiwhat) was not always getting set for -bs and -bS
2323 input. This is fixed, and it now also says "batched" for BSMTP.
2325 PH/11 Added control=no_pipelining.
2327 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2328 patch, slightly modified), and move the expansion of helo_data till after
2329 the connection is made in the smtp transport (so it can use these
2332 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2334 PH/14 Added log_selector = +pid.
2336 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2338 PH/16 Add ${if forany and ${if forall.
2340 PH/17 Added dsn_from option to vary the From: line in DSNs.
2342 PH/18 Flush SMTP output before performing a callout, unless control =
2343 no_callout_flush is set.
2345 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2346 was true (the default) a successful delivery failed to delete the retry
2347 item, thus causing premature timeout of the address. The bug is now
2350 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2352 PH/21 Long custom messages for fakedefer and fakereject are now split up
2353 into multiline responses in the same way that messages for "deny" and
2354 other ACL rejections are.
2356 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2357 with slight modification.
2359 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2360 draft, changing the syntax and factoring some duplicate code".
2362 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2363 for deliveries of the second and subsequent messages over the same SMTP
2366 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2367 ${reduce, with only minor "tidies".
2369 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2371 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2372 expansion side effects.
2374 PH/27 When a message times out after an over-quota error from an Exim-imposed
2375 quota, the bounce message says "mailbox is full". This message was not
2376 being given when it was a system quota that was exceeded. It now should
2379 MH/03 Made $recipients available in local_scan(). local_scan() already has
2380 better access to the recipient list through recipients_list[], but
2381 $recipients can be useful in postmaster-provided expansion strings.
2383 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2384 in the case of a MAIL command with additional options following the
2385 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2386 were accidentally chopped off.
2388 PH/29 SMTP synchronization checks are implemented when a command is read -
2389 there is a check that no more input is waiting when there shouldn't be
2390 any. However, for some commands, a delay in an ACL can mean that it is
2391 some time before the response is written. In this time, more input might
2392 arrive, invalidly. So now there are extra checks after an ACL has run for
2393 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2394 pipelining has not been advertised.
2396 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2398 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2399 returned in the SMTP response when smtp_return_error_details was false.
2400 This has been fixed.
2402 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2403 instead of the C I/O that was originally supplied, because problems were
2404 reported on Solaris.
2406 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2407 Exim which did not show up earlier: it was assuming that a call to
2408 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2409 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2410 was a macro that became an assignment, so it seemed to work. This has
2411 changed to a proper function call with a void return, hence the compile
2412 error. Exim's code has been fixed.
2414 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2417 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2419 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2421 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2422 because it turns out to be too big or there is a timeout) there is no
2423 "Completed" line in the log. When some messages of this type were
2424 selected by exigrep, they were listed as "not completed". Others were
2425 picked up by some special patterns. I have improved the selection
2426 criteria to be more general.
2428 PH/38 The host_find_failed option in the manualroute router can now be set
2429 to "ignore", to completely ignore a host whose IP address cannot be
2430 found. If all hosts are ignored, the behaviour is controlled by the new
2431 host_all_ignored option.
2433 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2434 homing or because of multiple MX records with /mx) generated more than
2435 one IP address, and the following item turned out to be the local host,
2436 all the secondary addresses of the first item were incorrectly removed
2437 from the list, along with the local host and any following hosts (which
2438 is what is supposed to happen).
2440 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2441 whoever called Exim into the -H file. In the case of the daemon it was
2442 behaving confusingly. When first started, it used values for whoever
2443 started the daemon, but after a SIGHUP it used the Exim user (because it
2444 calls itself on a restart). I have changed the code so that it now always
2447 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2448 message are rejected with the same error (e.g. no authentication or bad
2449 sender address), and a DATA command is nevertheless sent (as can happen
2450 with PIPELINING or a stupid MUA), the error message that was given to the
2451 RCPT commands is included in the rejection of the DATA command. This is
2452 intended to be helpful for MUAs that show only the final error to their
2455 PH/42 Another patch from the Sieve maintainer.
2457 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2458 Eximstats - Fixed some broken HTML links and added missing column headers
2460 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2461 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2463 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2465 PH/43 Yet another patch from the Sieve maintainer.
2467 PH/44 I found a way to check for a TCP/IP connection going away before sending
2468 the response to the final '.' that terminates a message, but only in the
2469 case where the client has not sent further data following the '.'
2470 (unfortunately, this is allowed). However, in many cases there won't be
2471 any further data because there won't be any more messages to send. A call
2472 to select() can be used: if it shows that the input is "ready", there is
2473 either input waiting, or the socket has been closed. An attempt to read
2474 the next input character can distinguish the two cases. Previously, Exim
2475 would have sent an OK response which the client would never have see.
2476 This could lead to message repetition. This fix should cure that, at
2477 least in a lot of common cases.
2479 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2480 advertised in response to EHLO.
2486 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2487 fixed by 4.65/MH/01 (is this a record?) are fixed:
2489 (i) An empty string was always treated as zero by the numeric comparison
2490 operators. This behaviour has been restored.
2492 (ii) It is documented that the numeric comparison operators always treat
2493 their arguments as decimal numbers. This was broken in that numbers
2494 starting with 0 were being interpreted as octal.
2496 While fixing these problems I realized that there was another issue that
2497 hadn't been noticed. Values of message_size_limit (both the global option
2498 and the transport option) were treated as octal if they started with 0.
2499 The documentation was vague. These values are now always treated as
2500 decimal, and I will make that clear in the documentation.
2506 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2507 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2510 MH/01 Don't check that the operands of numeric comparison operators are
2511 integers when their expansion is in "skipping" mode (fixes bug
2512 introduced by 4.64-PH/07).
2514 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2515 child addresses, Exim now panics and dies. Previously, because the count
2516 is held in a short int, deliveries were likely to be lost. As such a
2517 large number of recipients for a single message is ridiculous
2518 (performance will be very, very poor), I have chosen to impose a limit
2519 rather than extend the field.
2525 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2526 leftover -K file (the existence of which was triggered by #402).
2527 While we were at it, introduced process PID as part of the -K
2528 filename. This should rule out race conditions when creating
2531 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2532 processing considerably. Previous code took too long for large mails,
2533 triggering a timeout which in turn triggers #401.
2535 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2536 in the DK code in transports.c. sendfile() is not really portable,
2537 hence the _LINUX specificness.
2539 TF/01 In the add_headers option to the mail command in an Exim filter,
2540 there was a bug that Exim would claim a syntax error in any
2541 header after the first one which had an odd number of characters
2544 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2545 callout verification, Exim cached a "reject" for the entire domain. This
2546 is correct for most verifications, but it is not correct for a recipient
2547 verification with use_sender or use_postmaster set, because in that case
2548 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2549 case of MAIL FROM:<> rejection from other early rejections (e.g.
2550 rejection of HELO). When verifying a recipient using a non-null MAIL
2551 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2552 Whatever the result of the callout, the value of the domain cache is
2553 left unchanged (for any other kind of callout, getting as far as trying
2554 RCPT means that the domain itself is ok).
2556 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2559 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2560 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2561 session, the daemon ignores ECONNECT errors and logs others; it now
2562 ignores EPIPE as well.
2564 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2565 (quoted-printable decoding).
2567 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2568 later the small subsequent patch to fix an introduced bug.
2570 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2572 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2574 PH/08 An error is now given if message_size_limit is specified negative.
2576 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2577 to be given (somewhat) arbitrary names.
2579 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2582 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2583 miscellaneous code fixes
2585 PH/10 Added the log_reject_target ACL modifier to specify where to log
2588 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2589 hostname. This is wrong, because it relates to the incoming message (and
2590 probably the interface on which it is arriving) and not to the outgoing
2591 callout (which could be using a different interface). This has been
2592 changed to use the value of the helo_data option from the smtp transport
2593 instead - this is what is used when a message is actually being sent. If
2594 there is no remote transport (possible with a router that sets up host
2595 addresses), $smtp_active_hostname is used.
2597 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2598 tweaks were necessary in order to get it to work (see also 21 below):
2599 (a) The code assumed that strncpy() returns a negative number on buffer
2600 overflow, which isn't the case. Replaced with Exim's string_format()
2602 (b) There were several signed/unsigned issues. I just did the minimum
2603 hacking in of casts. There is scope for a larger refactoring.
2604 (c) The code used strcasecmp() which is not a standard C function.
2605 Replaced with Exim's strcmpic() function.
2606 (d) The code set only $1; it now sets $auth1 as well.
2607 (e) A simple test gave the error "authentication client didn't specify
2608 service in request". It would seem that Dovecot has changed its
2609 interface. Fortunately there's a specification; I followed it and
2610 changed what the client sends and it appears to be working now.
2612 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2615 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2616 address A is aliased to B and C, where B exists and C does not. Without
2617 -v the output is "A verified" because verification stops after a
2618 successful redirection if more than one address is generated. However,
2619 with -v the child addresses are also verified. Exim was outputting "A
2620 failed to verify" and then showing the successful verification for C,
2621 with its parentage. It now outputs "B failed to verify", showing B's
2622 parentage before showing the successful verification of C.
2624 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2625 look up a TXT record in a specific list after matching in a combined
2628 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2629 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2630 they consult the DNS. I had assumed they would set it the way they
2631 wanted; and indeed my experiments on Linux seem to show that in some
2632 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2633 To be on the safe side, however, I have now made the interface to
2634 host_find_byname() similar to host_find_bydns(), with an argument
2635 containing the DNS resolver options. The host_find_byname() function now
2636 sets these options at its start, just as host_find_bydns() does. The smtp
2637 transport options dns_qualify_single and dns_search_parents are passed to
2638 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2639 of host_find_byname() use the default settings of RES_DEFNAMES
2640 (qualify_single) but not RES_DNSRCH (search_parents).
2642 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2643 spool_read_header() do less string testing, by means of a preliminary
2644 switch on the second character of optional "-foo" lines. (This is
2645 overdue, caused by the large number of possibilities that now exist.
2646 Originally there were few.) While I was there, I also converted the
2647 str(n)cmp tests so they don't re-test the leading "-" and the first
2648 character, in the hope this might squeeze out yet more improvement.
2650 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2651 flag allowing group syntax was set by the header_syntax check but not
2652 turned off, possible causing trouble later; (2) The flag was not being
2653 set at all for the header_verify test, causing "group"-style headers to
2654 be rejected. I have now set it in this case, and also caused header_
2655 verify to ignore an empty address taken from a group. While doing this, I
2656 came across some other cases where the code for allowing group syntax
2657 while scanning a header line wasn't quite right (mostly, not resetting
2658 the flag correctly in the right place). These bugs could have caused
2659 trouble for malformed header lines. I hope it is now all correct.
2661 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2662 with the "reply" argument non-NULL. The code, however (which originally
2663 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2664 but it didn't always do it. This confused somebody who was copying the
2665 code for some other use. I have removed all the tests.
2667 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2668 feature that was used to support insecure browsers during the U.S. crypto
2669 embargo. It requires special client support, and Exim is probably the
2670 only MTA that supported it -- and would never use it because real RSA is
2671 always available. This code has been removed, because it had the bad
2672 effect of slowing Exim down by computing (never used) parameters for the
2673 RSA_EXPORT functionality.
2675 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2676 authenticator to fail if there's a tab character in the incoming data
2677 (there should never be unless someone is messing about, as it's supposed
2678 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2679 if the connection is using TLS or if the remote IP is the same as the
2680 local IP, and the "valid-client-cert option" if a client certificate has
2683 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2684 authenticators. This can be used for authorization after authentication
2685 succeeds. (In the case of plaintext, it servers for both authentication
2688 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2689 if any retry times were supplied.
2691 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2692 connection, where there is no client IP address to check. In this
2693 situation, the verify now always succeeds.
2695 PH/25 Applied John Jetmore's -Mset patch.
2697 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2699 PH/27 In a string expansion for a processed (not raw) header when multiple
2700 headers of the same name were present, leading whitespace was being
2701 removed from all of them, but trailing whitespace was being removed only
2702 from the last one. Now trailing whitespace is removed from each header
2703 before concatenation. Completely empty headers in a concatenation (as
2704 before) are ignored.
2706 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2707 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2709 PH/29 [Removed. This was a change that I later backed out, and forgot to
2710 correct the ChangeLog entry (that I had efficiently created) before
2711 committing the later change.]
2713 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2714 address errors (4xx response to RCPT) over the same connection as other
2715 messages routed to the same hosts. Such deliveries are always "forced",
2716 so retry times are not inspected. This resulted in far too many retries
2717 for the affected addresses. The effect occurred only when there were more
2718 hosts than the hosts_max_try setting in the smtp transport when it had
2719 the 4xx errors. Those hosts that it had tried were not added to the list
2720 of hosts for which the message was waiting, so if all were tried, there
2721 was no problem. Two fixes have been applied:
2723 (i) If there are any address or message errors in an SMTP delivery, none
2724 of the hosts (tried or untried) are now added to the list of hosts
2725 for which the message is waiting, so the message should not be a
2726 candidate for sending over the same connection that was used for a
2727 successful delivery of some other message. This seems entirely
2728 reasonable: after all the message is NOT "waiting for some host".
2729 This is so "obvious" that I'm not sure why it wasn't done
2730 previously. Hope I haven't missed anything, but it can't do any
2731 harm, as the worst effect is to miss an optimization.
2733 (ii) If, despite (i), such a delivery is accidentally attempted, the
2734 routing retry time is respected, so at least it doesn't keep
2735 hammering the server.
2737 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2738 in ${readsocket because some servers need this prod.
2740 PH/32 Added some extra debug output when updating a wait-xxx database.
2742 PH/33 The hint "could be header name not terminated by colon", which has been
2743 given for certain expansion errors for a long time, was not being given
2744 for the ${if def:h_colon_omitted{... case.
2746 PH/34 The spec says: "With one important exception, whenever a domain list is
2747 being scanned, $domain contains the subject domain." There was at least
2748 one case where this was not true.
2750 PH/35 The error "getsockname() failed: connection reset by peer" was being
2751 written to the panic log as well as the main log, but it isn't really
2752 panic-worthy as it just means the connection died rather early on. I have
2753 removed the panic log writing for the ECONNRESET error when getsockname()
2756 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2757 runs only) independently of the message's sender address. This meant
2758 that, if the 4xx error was in fact related to the sender, a different
2759 message to the same recipient with a different sender could confuse
2760 things. In particular, this can happen when sending to a greylisting
2761 server, but other circumstances could also provoke similar problems.
2762 I have changed the default so that the retry time for these errors is now
2763 based a combination of the sender and recipient addresses. This change
2764 can be overridden by setting address_retry_include_sender=false in the
2767 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2768 remote server are returned as part of bounce messages. This was not
2769 happening for LMTP over a pipe (the lmtp transport), but now it is the
2770 same for both kinds of LMTP.
2772 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2773 in header lines that were in fact CNAMEs. This is no longer the case.
2775 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2776 and queue runs started by the daemon processed all messages. This has
2777 been fixed so that -R and -S can now usefully be given with -q<time>.
2779 PH/40 Import PCRE release 6.7 (fixes some bugs).
2781 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2783 PH/42 Give an error if -q is specified more than once.
2785 PH/43 Renamed the variables $interface_address and $interface_port as
2786 $received_ip_address and $received_port, to make it clear that these
2787 values apply to message reception, and not to the outgoing interface when
2788 a message is delivered. (The old names remain recognized, of course.)
2790 PH/44 There was no timeout on the connect() call when using a Unix domain
2791 socket in the ${readsocket expansion. There now is.
2793 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2794 be meaningful with "accept".
2796 SC/01 Eximstats V1.43
2797 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2799 SC/02 Eximstats V1.44
2800 Use a glob alias rather than an array ref in the generated
2801 parser. This improves both readability and performance.
2803 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2804 Collect SpamAssassin and rejection statistics.
2805 Don't display local sender or destination tables unless
2806 there is data to show.
2807 Added average volumes into the top table text output.
2809 SC/04 Eximstats V1.46
2810 Collect data on the number of addresses (recipients)
2811 as well as the number of messages.
2813 SC/05 Eximstats V1.47
2814 Added 'Message too big' to the list of mail rejection
2815 reasons (thanks to Marco Gaiarin).
2817 SC/06 Eximstats V1.48
2818 Mainlog lines which have GMT offsets and are too short to
2819 have a flag are now skipped.
2821 SC/07 Eximstats V1.49 (Alain Williams)
2822 Added the -emptyok flag.
2824 SC/08 Eximstats V1.50
2825 Fixes for obtaining the IP address from reject messages.
2827 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2828 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2829 whitespace changes from 4.64-PH/27
2831 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2834 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2835 are found, allow negative numbers in numeric criteria)
2837 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2839 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2840 to match changes made in 4.64-PH/43
2842 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2844 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2845 "domain" is actually an IP address, and force a failure. This locks out
2846 those revolvers/nameservers that support "A-for-A" lookups, in
2847 contravention of the specifications.
2849 PH/48 When a host name was looked up from an IP address, and the subsequent
2850 forward lookup of the name timed out, the host name was left in
2851 $sender_host_name, contrary to the specification.
2853 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2854 restricted to single-key lookups, Exim was not diagnosing an error if
2855 * or *@ was used with a query-style lookup.
2857 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2859 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2860 long ago, but noone interested enough thought of it. Let's just say that
2861 the "1.1" means that there are some new functions that weren't there at
2862 some point in the past.
2864 PH/51 Error processing for expansion failure of helo_data from an smtp
2865 transport during callout processing was broken.
2867 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2868 tested/used via the -bh/-bhc/-bs options.
2870 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2871 bug, fixed in subsequent PCRE releases).
2873 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2874 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2876 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2882 SC/01 Use a glob alias rather than an array ref in eximstats generated
2883 parser. This improves both readability and performance.
2885 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2886 Don't display local sender or destination tables in eximstats unless
2887 there is data to show.
2888 Added average volumes into the eximstats top table text output.
2890 SC/03 Collect data on the number of addresses (recipients) as well
2891 as the number of messages in eximstats.
2893 TF/01 Correct an error in the documentation for the redirect router. Exim
2894 does (usually) call initgroups() when daemonizing.
2896 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2897 with consistent privilege compared to when running as a daemon.
2899 TF/03 Note in the spec that $authenticated_id is not set for local
2900 submissions from trusted users.
2902 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2903 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2905 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2906 by adding some example configuration directives to the default
2907 configuration file. A little bit of work is required to uncomment the
2908 directives and define how usernames and passwords are checked, but
2909 there is now a framework to start from.
2911 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2912 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2913 without this. I don't know how relevant this is to other LDAP libraries.
2915 PH/02 Add the verb name to the "unknown ACL verb" error.
2917 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2919 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2921 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2922 directory not expanded when it should be if an expanded home directory
2923 was set for the address (which is overridden by the transport).
2925 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2928 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2929 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2930 because it is too late at that time, and has no effect.
2932 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2933 security issue with \' (bugzilla #107). I could not use the
2934 PQescapeStringConn() function, because it needs a PGconn value as one of
2937 PH/08 When testing addresses using -bt, indicate those final addresses that
2938 are duplicates that would not cause an additional delivery. At least one
2939 person was confused, thinking that -bt output corresponded to deliveries.
2940 (Suppressing duplicates isn't a good idea as you lose the information
2941 about possibly different redirections that led to the duplicates.)
2943 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2944 systems where poll() doesn't work, in particular OS X.
2946 PH/10 Added more information to debugging output for retry time not reached.
2948 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2949 operations in malware.c.
2951 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2954 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2955 syslog_duplication set false, log lines that would normally be written
2956 both the the main log and to the reject log were not written to syslog at
2959 PH/14 In the default configuration, change the use of "message" in ACL warn
2960 statements to "add_header".
2962 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2963 not followed by a command (e.g. "seen endif").
2965 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2966 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2969 PH/17 Added extra conditions to the default value of delay_warning_condition
2973 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2974 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2975 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2978 The Auto-Submitted: and various List- headers are standardised, whereas I
2979 don't think Precedence: ever was.
2981 PH/18 Refactored debugging code in route_finduser() to show more information,
2982 in particular, the error code if getpwnam() issues one.
2984 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2985 This is apparently needed in addition to the PH/07 change above to avoid
2986 any possible encoding problems.
2988 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2989 but not after initializing Perl.
2991 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2992 output them only if debugging. By default they are written stderr,
2993 apparently, which is not desirable.
2995 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2998 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3001 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3003 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3004 authenticated or an ident call has been made. Suppress the default
3005 values for $authenticated_id and $authenticated_sender (but permit -oMai
3006 and -oMas) when testing with -bh.
3008 PH/24 Re-jigged the order of the tests in the default configuration so that the
3009 tests for valid domains and recipients precede the DNS black list and CSA
3010 tests, on the grounds that those ones are more expensive.
3012 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3013 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3014 This bug exists in every version of Exim that I still have, right back to
3017 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3018 However, an attempt to turn on case-sensitivity in a regex key by
3019 including (?-i) didn't work because the subject string was already
3020 lowercased, and the effects were non-intuitive. It turns out that a
3021 one-line patch can be used to allow (?-i) to work as expected.
3027 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3028 other effects) broke the use of negated acl sub-conditions.
3030 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3033 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3034 "Deny" causes Exim to reject the incoming connection with a 554 error.
3035 Unfortunately, if there is a major crisis, such as a disk failure,
3036 tcp-wrappers gives "deny", whereas what one would like would be some
3037 kind of temporary error. A kludge has been added to help with this.
3038 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3039 554 error is used if errno is still zero or contains ENOENT (which occurs
3040 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3043 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3045 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3046 errors. Otherwise a message that provokes a temporary error (when other
3047 messages do not) can cause a whole host to time out.
3049 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3050 addresses were routed directly to files or pipes from a redirect router.
3051 File deliveries just didn't batch; pipe deliveries might have suffered
3054 PH/06 A failure to get a lock for a hints database would erroneously always say
3055 "Failed to get write lock", even when it was really a read lock.
3057 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3058 of 0600. This has been changed to use the value of the lockfile_mode
3059 option (which defaults to 0600).
3061 PH/08 Applied small patch from the Sieve maintainer.
3063 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3064 folder from quota calculations, a direct delivery into this folder messed
3065 up the contents of the maildirsize file. This was because the regex was
3066 used only to exclude .Trash (or whatever) when the size of the mailbox
3067 was calculated. There was no check that a delivery was happening into an
3068 excluded directory. This bug has been fixed by ignoring all quota
3069 processing for deliveries into excluded directories.
3071 PH/10 Added the maildirfolder_create_regex option to appendfile.
3077 PH/01 The code for finding all the local interface addresses on a FreeBSD
3078 system running IPv6 was broken. This may well have applied to all BSD
3079 systems, as well as to others that have similar system calls. The broken
3080 code found IPv4 interfaces correctly, but gave incorrect values for the
3081 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3082 that it would not match correctly against @[] and not recognize the IPv6
3085 PH/02 The ipliteral router was not recognizing addresses of the form user@
3086 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3088 PH/03 Added disable_ipv6.
3090 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3091 decoded form, because it is most often used to construct To: headers
3092 lines in autoreplies, and the decoded form may well be syntactically
3093 invalid. However, $reply_address has leading white space removed, and all
3094 newlines turned into spaces so that the autoreply transport does not
3097 PH/05 If group was specified without a user on a router, and no group or user
3098 was specified on a transport, the group from the router was ignored.
3100 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3101 for visible compile-time settings that can be used to change these
3102 numbers, for those that want even more. Backwards compatibility with old
3103 spool files has been maintained. However, going back to a previous Exim
3104 release will lost any variables that are in spool files.
3106 PH/07 Two small changes when running in the test harness: increase delay when
3107 passing a TCP/IP connection to a new process, in case the original
3108 process has to generate a bounce, and remove special handling of
3109 127.0.0.2 (sic), which is no longer necessary.
3111 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3112 be the same on different OS.
3114 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3117 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3118 whether --show-vars was specified or not
3120 JJ/02 exipick: Added support for new ACL variable spool format introduced
3123 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3124 syntactically invalid From: or Reply-to: line, and a filter used this to
3125 generate an autoreply, and therefore failed to obtain an address for the
3126 autoreply, Exim could try to deliver to a non-existent relative file
3127 name, causing unrelated and misleading errors. What now happens is that
3128 it logs this as a hard delivery error, but does not attempt to create a
3131 PH/11 The exinext utility has a -C option for testing purposes, but although
3132 the given file was scanned by exinext itself; it wasn't being passed on
3133 when Exim was called.
3135 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3136 an end-of-file indication when reading a command response.
3138 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3139 compiled. In many other places in Exim, IPv6 addresses are always
3140 recognized, so I have changed this. It also means that IPv4 domain
3141 literals of the form [IPV4:n.n.n.n] are now always recognized.
3143 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3144 used if the router is not running as root, for example, when verifying at
3145 ACL time, or when using -bh. The debugging output from this situation was
3146 non-existent - all you got was a failure to exec. I have made two
3149 (a) Failures to set uid/gid, the current directory, or a process leader
3150 in a subprocess such as that created by queryprogram now generate
3151 suitable debugging output when -d is set.
3153 (b) The queryprogram router detects when it is not running as root,
3154 outputs suitable debugging information if -d is set, and then runs
3155 the subprocess without attempting to change uid/gid.
3157 PH/15 Minor change to Makefile for building test_host (undocumented testing
3160 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3161 additional section of a DNS packet that returns MX or SRV records.
3162 Instead, it always explicitly searches for A/AAAA records. This avoids
3163 major problems that occur when a DNS server includes only records of one
3164 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3165 fixed another bug: if SRV records were looked up and the corresponding
3166 address records were *not* found in the additional section, the port
3167 values from the SRV records were lost.
3169 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3170 using the correct key (the original address) when searching the retry
3171 rules in order to find which one to use for generating the retry hint.
3173 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3174 adding the default one. Similarly, if it contains a Reply-To: header, the
3175 errors_reply_to option, if set, is not used.
3177 PH/19 When calculating a retry time, Exim used to measure the "time since
3178 failure" by looking at the "first failed" field in the retry record. Now
3179 it does not use this if it is later than than the arrival time of the
3180 message. Instead it uses the arrival time. This makes for better
3181 behaviour in cases where some deliveries succeed, thus re-setting the
3182 "first failed" field. An example is a quota failure for a huge message
3183 when small messages continue to be delivered. Without this change, the
3184 "time since failure" will always be short, possible causing more frequent
3185 delivery attempts for the huge message than are intended.
3186 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3188 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3189 $1, $2, $3) because the numerical variables can be reset during some
3190 expansion items (e.g. "match"), thereby losing the authentication data.
3192 PH/21 Make -bV show the size of off_t variables so that the test suite can
3193 decide whether to run tests for quotas > 2G.
3195 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3196 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3197 filecount value is greater than 2G or if a quota value is greater than 2G
3198 on a system where the size of off_t is not greater than 4, a panic error
3201 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3202 never match. The debug and -bh output now contains an explicit error
3203 message indicating a malformed IPv4 address or mask.
3205 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3206 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3207 PH/23 above applies.
3209 PH/25 Do not write to syslog when running in the test harness. The only
3210 occasion when this arises is a failure to open the main or panic logs
3211 (for which there is an explicit test).
3213 PH/26 Added the /no_tell option to "control=freeze".
3215 PH/27 If a host name lookup failed very early in a connection, for example, if
3216 the IP address matched host_lookup and the reverse lookup yielded a name
3217 that did not have a forward lookup, an error message of the form "no IP
3218 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3219 could be logged. Now it outputs the IP address instead of "NULL".
3221 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3222 allows the sender and the authenticated sender to be set when
3223 submitting a message from within Exim. Since child_open_exim() is
3224 documented for local_scan(), the new function should be too.
3226 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3227 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3228 results in an empty string is now treated as unset.
3230 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3232 PH/31 Added sender_verify_fail logging option.
3234 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3235 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3236 tidied the source and removed it altogether.
3238 PH/33 When a queue run was abandoned because the load average was too high, a
3239 log line was always written; now it is written only if the queue_run log
3240 selector is set. In addition, the log line for abandonment now contains
3241 information about the queue run such as the pid. This is always present
3242 in "start" and "stop" lines but was omitted from the "abandon" line.
3244 PH/34 Omit spaces between a header name and the colon in the error message that
3245 is given when verify = headers_syntax fails (if there are lots of them,
3246 the message gets confusing).
3248 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3249 names, as there are now some PTR records that contain slashes. This check
3250 is only to protect against broken name servers that fall over on strange
3251 characters, so the fact that it applies to all lookups doesn't matter.
3253 PH/36 Now that the new test suite is complete, we can remove some of the
3254 special code in Exim that was needed for the old test suite. For example,
3255 sorting DNS records because real resolvers return them in an arbitrary
3256 order. The new test suite's fake resolver always returns records in the
3259 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3260 bounces) except when queue_only is set, to avoid logging races between
3261 the different processes.
3263 PH/38 Panic-die if .include specifies a non-absolute path.
3265 PH/39 A tweak to the "H" retry rule from its user.
3267 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3268 a label. They prevented compilation on older perls.
3270 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3271 a warning to be raised on newish perls.
3273 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3274 on queue. Changes to match documented behaviour of showing count of
3275 messages matching specified criteria.
3277 PH/40 Changed the default ident timeout from 30s to 5s.
3279 PH/41 Added support for the use of login_cap features, on those BSD systems
3280 that have them, for controlling the resources used by pipe deliveries.
3282 PH/42 The content-scanning code uses fopen() to create files in which to put
3283 message data. Previously it was not paying any attention to the mode of
3284 the files. Exim runs with umask(0) because the rest of the code creates
3285 files with open(), and sets the required mode explicitly. Thus, these
3286 files were ending up world-writeable. This was not a big issue, because,
3287 being within the spool directory, they were not world-accessible. I have
3288 created a function called modefopen, which takes an additional mode
3289 argument. It sets umask(777), creates the file, chmods it to the required
3290 mode, then resets the umask. All the relevant calls to fopen() in the
3291 content scanning code have been changed to use this function.
3293 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3294 to 24 hours. This avoids potential overflow problems when processing G
3295 and H retry rules. I suspect nobody ever tinkers with this value.
3297 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3299 PH/45 When the plaintext authenticator is running as a client, the server's
3300 challenges are checked to ensure they are valid base64 strings. By
3301 default, the authentication attempt is cancelled if an invalid string is
3302 received. Setting client_ignore_invalid_base64 true ignores these errors.
3303 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3304 they are received. Thus, the responses can be made to depend on the
3305 challenges. If an invalid string is ignored, an empty string is placed in
3308 PH/46 Messages that are created by the autoreply transport now contains a
3309 References: header, in accordance with RFCs 2822 and 3834.
3311 PH/47 Added authenticated_sender_force to the smtp transport.
3313 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3315 PH/49 Installed latest patch from the Sieve maintainer.
3317 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3318 was also set, the appendfile transport was unnecessarily scanning a
3319 directory of message files (e.g. for maildir delivery) to find the count
3320 of files (along with the size), even though it did not need this
3321 information. It now does the scan only if it needs to find either the
3322 size of the count of files.
3324 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3326 PH/52 Two bugs concerned with error handling when the smtp transport is
3329 (i) Exim was not creating retry information for temporary errors given
3330 for individual recipients after the DATA command when the smtp transport
3331 was used in LMTP mode. This meant that they could be retried too
3332 frequently, and not timed out correctly.
3334 (ii) Exim was setting the flag that allows error details to be returned
3335 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3336 recipients that were returned after the DATA command.
3338 PH/53 This is related to PH/52, but is more general: for any failing address,
3339 when detailed error information was permitted to be returned to the
3340 sender, but the error was temporary, then after the final timeout, only
3341 "retry timeout exceeded" was returned. Now it returns the full error as
3342 well as "retry timeout exceeded".
3344 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3345 do this, and (what is worse) MTAs that accept it.
3347 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3348 will now be deprecated.
3350 PH/56 New os.c-cygwin from the Cygwin maintainer.
3352 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3353 formats (previously only available in exim formats via -bpr, -bpru,
3354 and -bpra. Now also available in native and exiqgrep formats)
3356 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3357 with very large, slow to parse queues
3359 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3361 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3363 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3364 responses to authentication challenges, though it was showing the
3365 challenges; (ii) I've removed the CR characters from the debug output for
3368 PH/58 Allow for the insertion of a newline as well as a space when a string
3369 is turned into more than one encoded-word during RFC 2047 encoding. The
3370 Sieve code now uses this.
3372 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3373 data_4xx, lost_connection, tls_required.
3375 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3376 message was being sent as an SMTP response.
3378 PH/61 Add -l and -k options to exicyclog.
3380 PH/62 When verifying, if an address was redirected to one new address, so that
3381 verification continued, and the new address failed or deferred after
3382 having set something in $address_data, the value of $address_data was not
3383 passed back to the ACL. This was different to the case when no
3384 redirection occurred. The value is now passed back in both cases.
3386 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3387 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3388 use login_cap.h, so on its own it isn't the distinguishing feature. The
3389 new name refers directly to the setclassresources() function.
3391 PH/65 Added configuration files for NetBSD3.
3393 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3395 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3396 is preferred over IPv4.
3398 PH/68 The bounce_return_message and bounce_return_body options were not being
3399 honoured for bounces generated during the reception of non-SMTP messages.
3400 In particular, this applied to messages rejected by the ACL. This bug has
3401 been fixed. However, if bounce_return_message is true and bounce_return_
3402 body is false, the headers that are returned for a non-SMTP message
3403 include only those that have been read before the error was detected.
3404 (In the case of an ACL rejection, they have all been read.)
3406 PH/69 The HTML version of the specification is now built in a directory called
3407 spec_html instead of spec.html, because the latter looks like a path with
3408 a MIME-type, and this confuses some software.
3410 PH/70 Catch two compiler warnings in sieve.c.
3412 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3413 function verify_get_ident() calls ip_connect() to connect a socket, but
3414 if the "connect()" function timed out, ip_connect() used to close the
3415 socket. However, verify_get_ident() also closes the socket later, and in
3416 between Exim writes to the log, which may get opened at this point. When
3417 the socket was closed in ip_connect(), the log could get the same file
3418 descriptor number as the socket. This naturally causes chaos. The fix is
3419 not to close the socket in ip_connect(); the socket should be closed by
3420 the function that creates it. There was only one place in the code where
3421 this was missing, in the iplookup router, which I don't think anybody now
3422 uses, but I've fixed it anyway.
3424 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3425 well as to direct DNS lookups. Otherwise the handling of names in host
3426 lists is inconsistent and therefore confusing.
3432 PH/01 Two changes to the default runtime configuration:
3434 (1) Move the checks for relay_from_hosts and authenticated clients from
3435 after to before the (commented out) DNS black list checks.
3437 (2) Add control=submission to the relay_from_hosts and authenticated
3438 clients checks, on the grounds that messages accepted by these
3439 statements are most likely to be submissions.
3441 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3443 (1) Generate an error if the third argument for the ${prvs expansion is
3446 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3449 (3) Reset the variables that are obtained from the first argument of
3450 ${prvscheck and used in the second argument before leaving the code,
3451 because their memory is reclaimed, so using them afterwards may do
3454 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3455 one (it's much easier than Tom thought :-).
3457 (5) Because of (4), we can now allow for the use of $prvscheck_result
3458 inside the third argument.
3460 PH/03 For some reason, the default setting of PATH when running a command from
3461 a pipe transport was just "/usr/bin". I have changed it to
3464 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3465 anything to be listed in the output from -bV.
3467 PH/05 When a filter generated an autoreply, the entire To: header line was
3468 quoted in the delivery log line, like this:
3470 => >A.N.Other <ano@some.domain> <original@ddress> ...
3472 This has been changed so that it extracts the operative address. There
3473 may be more than one such address. If so, they are comma-separated, like
3476 => >ano@some.domain,ona@other.domain <original@ddress> ...
3478 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3479 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3480 not being looked up in the rDNS to get a host name, Exim was showing the
3481 IP address twice in Received: lines, even though the IP addresses were
3482 identical. For example:
3484 Received: from [1.2.3.4] (helo=[1.2.3.4])
3486 However, if the real host name was known, it was omitting the HELO data
3487 if it matched the actual IP address. This has been tidied up so that it
3488 doesn't show the same IP address twice.
3490 PH/07 When both +timestamp and +memory debugging was on, the value given by
3491 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3492 called by the debug printing, thereby overwriting the timestamp buffer.
3493 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3495 PH/08 When the original message was included in an autoreply transport, it
3496 always said "this is a copy of the message, including all the headers",
3497 even if body_only or headers_only was set. It now gives an appropriate
3500 PH/09 Applied a patch from the Sieve maintainer which:
3502 o fixes some comments
3503 o adds the (disabled) notify extension core
3504 o adds some debug output for the result of if/elsif tests
3505 o points to the current vacation draft in the documentation
3506 and documents the missing references header update
3510 o fixes a bug in processing the envelope test (when testing
3511 multiple envelope elements, the last element determined the
3514 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3515 Electronic Mail") by including:
3517 Auto-submitted: auto-generated
3519 in the messages that it generates (bounce messages and others, such as
3520 warnings). In the case of bounce messages for non-SMTP messages, there was
3521 also a typo: it was using "Auto_submitted" (underscore instead of
3522 hyphen). Since every message generated by Exim is necessarily in response
3523 to another message, thes have all been changed to:
3525 Auto-Submitted: auto-replied
3527 in accordance with these statements in the RFC:
3529 The auto-replied keyword:
3531 - SHOULD be used on messages sent in direct response to another
3532 message by an automatic process,
3534 - MUST NOT be used on manually-generated messages,
3536 - MAY be used on Delivery Status Notifications (DSNs) and Message
3537 Disposition Notifications (MDNs),
3539 - MUST NOT be used on messages generated by automatic or periodic
3540 processes, except for messages which are automatic responses to
3543 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3544 to the default Received: header definition.
3546 PH/12 Added log selector acl_warn_skipped (default on).
3548 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3549 variables because (a) they are in the wrong storage pool and (b) even if
3550 they were copied, it wouldn't work properly because of the caching.
3552 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3553 checking when decoding. Apparently there are clients that generate
3554 overlong encoded strings. Why am I not surprised?
3556 PH/15 If the first argument of "${if match_address" was not empty, but did not
3557 contain an "@" character, Exim crashed. Now it writes a panic log message
3558 and treats the condition as false.
3560 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3562 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3563 extension "envelope-auth". The code is finished and in agreement with
3564 other implementations, but there is no documentation so far and in fact,
3565 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3566 not changing the active code.
3568 Print executed "if" and "elsif" statements when debugging is used. This
3569 helps a great deal to understand what a filter does.
3571 Document more things not specified clearly in RFC3028. I had all this
3572 sorted out, when out of a sudden new issues came to my mind. Oops."
3574 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3577 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3578 canonical form (as documented). However, after a host name lookup from
3579 the IP address, check_host() was doing a simple string comparison with
3580 addresses acquired from the DNS when checking that the found name did
3581 have the original IP as one of its addresses. Since any found IPv6
3582 addresses are likely to be in abbreviated form, the comparison could
3583 fail. Luckily, there already exists a function for doing the comparison
3584 by converting both addresses to binary, so now that is used instead of
3585 the text comparison.
3587 PH/20 There was another similar case to PH/19, when a complete host name was
3588 given in a host list; looking up its IP address could give an abbreviated
3589 form, whereas the current host's name might or might not be abbreviated.
3590 The same fix has been applied.
3596 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3597 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3600 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3601 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3603 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3605 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3606 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3607 Also, for "Auto-Submitted", treat anything other than "no" as
3608 non-personal, in accordance with RFC 3834. (Previously it treated
3609 anything starting "auto-" as non-personal.)
3611 TF/01 The control=submission/name=... option had a problem with syntax
3612 errors if the name included a slash character. The /name= option
3613 now slurps the rest of the string, so it can include any characters
3614 but it must come last in the list of options (after /sender_retain
3617 PH/05 Some modifications to the interface to the fake nameserver for the new
3625 TK/01 Added the "success_on_redirect" address verification option. See
3626 NewStuff for rationale and an example.
3628 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3630 PH/02 Patch to exigrep to allow it to work on syslog lines.
3632 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3633 fread() to skip over the body file's header line, because in Cygwin the
3634 header line is locked and is inaccessible.
3636 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3637 co-exist for some time) to make it clear that it is the Exim ID that is
3638 referenced, not the Message-ID: header line.
3640 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3641 string_format() function, because snprintf() does not exist on all
3644 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3645 ${stat: expansion item.
3647 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3648 protocol synchronization error", to keep the pedants happy.
3650 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3651 well as for IRIX systems, when gcc is being used. See the host.c source
3654 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3656 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3659 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3660 transport and to the smtp transport in LMTP mode.
3662 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3664 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3665 run a filter in a subprocess. This could lead to confusion in subsequent
3666 lookups in the parent process. There should also be a search_tidyup() at
3667 the end of the subprocess.
3669 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3670 only if the host matched helo_try_verify_hosts, which caused the
3671 verification to occur when the EHLO/HELO command was issued. The ACL just
3672 tested the remembered result. Now, if a previous verification attempt has
3673 not happened, "verify = helo" does it there and then.
3675 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3677 TK/03 Fix log output including CR from clamd.
3679 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3680 exist provoked a memory error which could cause a segfault.
3682 PH/15 Installed PCRE 6.2
3684 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3686 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3687 of the problem. Specifically, suggested +O2 rather than +O1 for the
3690 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3692 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3693 record, the port was not correctly logged when the outgoing_port log
3694 selector was set (it logged the transort's default port).
3696 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3697 fallback_hosts, and "hosts" in the smtp transport.
3699 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3700 host errors such as "Connection refused".
3702 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3703 authentication with radiusclient 0.4.9:
3705 - Error returned from rc_read_config was caught wrongly
3706 - Username/password not passed on to radius server due to wrong length.
3708 The presumption is that some radiusclient API changes for 4.51/PH/17
3709 were not taken care of correctly. The code is still untested by me (my
3710 Linux distribution still has 0.3.2 of radiusclient), but it was
3711 contributed by a Radius user.
3713 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3714 expanding the "port" option of the smtp transport.
3716 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3717 while reading a MIME header. Thanks to Tom Hughes for a patch.
3719 PH/24 Include config.h inside local_scan.h so that configuration settings are
3722 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3723 that in an ACL for RCPT (for example), you can examine exactly what was
3726 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3727 commands, but it was not correctly comparing the address with the actual
3728 client host address. Thus, it would show the EHLO address in Received:
3729 header lines when this was not necessary.
3731 PH/27 Added the % operator to ${eval:}.
3733 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3734 it should be ignoring failures (because with -C, for example, it has lost
3735 privilege). It wasn't ignoring creation failures other than "already
3738 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3741 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3742 because an input line was too long, either on its own, or by virtue of
3743 too many continuations, the temporary file was not being removed, and the
3744 return code was incorrect.
3746 PH/31 Missing "BOOL" in function definition in filtertest.c.
3748 PH/32 Applied Sieve patches from the maintainer.
3750 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3752 PH/33 Added "verify = not_blind".
3754 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3755 Local/Makefile (with some defaults set). These are used in built scripts
3756 such as exicyclog, but they have never been used in the exim_install
3757 script (though there are many overriding facilities there). I have
3758 arranged that the exim_install script now takes note of these two
3761 PH/35 Installed configuration files for Dragonfly.
3763 PH/36 When a locally submitted message by a trusted user did not contain a
3764 From: header, and the sender address was obtained from -f or from an SMTP
3765 MAIL command, and the trusted user did not use -F to supply a sender
3766 name, $originator_name was incorrectly used when constructing a From:
3767 header. Furthermore, $originator_name was used for submission mode
3768 messages from external hosts without From: headers in a similar way,
3769 which is clearly wrong.
3771 PH/37 Added control=suppress_local_fixups.
3773 PH/38 When log_selector = +received_sender was set, and the addition of the
3774 sender made the log line's construction buffer exactly full, or one byte
3775 less than full, an overflow happened when the terminating "\n" was
3778 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3779 when the result of a list match is failure because a DNS lookup failed.
3781 PH/40 RM_COMMAND is now used in the building process.
3783 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3784 the "build-* directories that it finds.
3786 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3787 address in a domain literal was a prefix of an interface address.
3789 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3790 when verifying a sender address, unless rewrite_headers is false.
3792 PH/44 Wrote a long comment about why errors_to addresses are verified as
3793 recipients, not senders.
3795 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3796 the ratelimit ACL was added.
3798 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3800 PH/46 Added extra information about PostgreSQL errors to the error string.
3802 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3803 avoiding the need to install special zones in a real server. This is
3804 backwards compatible; if it can't find the fake resolver, it drops back.
3805 Thus, both old and new test suites can be run.
3807 TF/02 Added util/ratelimit.pl
3809 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3810 clock is set back in time.
3812 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3813 Candler <B.Candler@pobox.com>.
3815 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3816 OK for addresses that are the result of redirections.
3818 PH/48 A number of further additions for the benefit of the new test suite,
3819 including a fake gethostbyname() that interfaces to the fake DNS resolver
3822 TF/06 The fix for widen_domains has also been applied to qualify_single and
3823 search_parents which are the other dnslookup options that can cause
3826 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3829 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3831 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3832 service" if it encountered an MX record with an empty target hostname.
3833 The message is now "an MX or SRV record indicated no SMTP service".
3835 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3836 if the DNS of the sending site is misconfigured. This is quite a
3837 common situation. This change restores the behaviour of treating a
3838 helo verification defer as a failure.
3840 PH/51 If self=fail was set on a router, the bounce message did not include the
3841 actual error message.
3847 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3849 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3850 command itself ended in error, the underlying message about the transport
3851 filter timeout was being overwritten with the pipe command error. Now the
3852 underlying error message should be appended to the second error message.
3854 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3856 PH/02 Reduce the amount of output that "make" produces by default. Full output
3857 can still be requested.
3859 PH/03 The warning log line about a condition test deferring for a "warn" verb
3860 was being output only once per connection, rather than after each
3861 occurrence (because it was using the same function as for successful
3862 "warn" verbs). This seems wrong, so I have changed it.
3864 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3865 it should not have, which might have caused a crash in the right
3866 circumstances, but probably never did.
3868 PH/04 Installed a modified version of Tony Finch's patch to make submission
3869 mode fix the return path as well as the Sender: header line, and to
3870 add a /name= option so that you can make the user's friendly name appear
3873 TF/03 Added the control = fakedefer ACL modifier.
3875 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3876 Mark Lowes for thorough testing.
3878 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3880 TK/03 Merged latest SRS patch from Miles Wilton.
3882 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3883 with the definition in sysexits.h (which is #included earlier).
3884 Fortunately, Exim does not actually use EX_OK. The code used to try to
3885 preserve the sysexits.h value, by assuming that macro definitions were
3886 scanned for macro replacements. I have been disabused of this notion,
3887 so now the code just undefines EX_OK before #including unistd.h.
3889 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3890 in the smtp transport. When a block could not be written in a single
3891 write() function, the timeout was being re-applied to each part-write.
3892 This seems wrong - if the receiver was accepting one byte at a time it
3893 would take for ever. The timeout is now adjusted when this happens. It
3894 doesn't have to be particularly precise.
3896 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3897 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3899 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3900 without a domain if the check to <postmaster@domain> fails.
3902 SC/01 Eximstats: added -xls and the ability to specify output files
3903 (patch written by Frank Heydlauf).
3905 SC/02 Eximstats: use FileHandles for outputting results.
3907 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3909 SC/04 Eximstats: fixed display of large numbers with -nvr option
3911 SC/05 Eximstats: fixed merging of reports with empty tables.
3913 SC/06 Eximstats: added the -include_original_destination flag
3915 SC/07 Eximstats: removed tabs and trailing whitespace.
3917 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3919 TK/06 MBOX spool code: Add real "From " MBOX separator line
3920 so the .eml file is really in mbox format (even though
3921 most programs do not really care). Patch from Alex Miller.
3923 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3924 The latter is generated from $received_to and is only set if the
3925 message has one envelope recipient. SA can use these headers,
3926 obviously out-of-the-box. Patch from Alex Miller.
3928 PH/08 The ${def test on a variable was returning false if the variable's
3929 value was "0", contrary to what the specification has always said!
3930 The result should be true unless the variable is empty.
3932 PH/09 The syntax error of a character other than { following "${if
3933 def:variable_name" (after optional whitespace) was not being diagnosed.
3934 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3935 accidental colon was present, for example, could give incorrect results.
3937 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3938 result is used (not including appendfile, where other changes are about
3941 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3942 This involved changing a lot of size variables from int to off_t. It
3943 should work with maildirs and everything.
3945 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3946 spamd dying while we are connected to it.
3948 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3951 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3952 function for BDB 4.3.
3954 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3956 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3957 the sources for PCRE have been re-arranged and more files are now
3960 PH/15 The code I had for printing potentially long long variables in PH/11
3961 above was not the best (it lost precision). The length of off_t variables
3962 is now inspected at build time, and an appropriate printing format (%ld
3963 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3964 to be "long long int" or "long int". This is needed for the internal
3965 formatting function string_vformat().
3967 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3968 the configuration file to be ":syslog", then the script "guesses" where
3969 the logs files are, rather than using the compiled in default. In our
3970 case the guess is not the same as the compiled default, so the script
3971 suddenly stopped working when I started to use syslog. The patch checks
3972 to see if log_file_path is "". If so, it attempts to read it from exim
3973 with no configuration file to get the compiled in version, before it
3974 falls back to the previous guessing code."
3976 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3977 implementing BATV in an Exim configuration. See NewStuff for the gory
3980 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3981 Makefile that are specific to HP-UX.
3983 PH/18 If the "use_postmaster" option was set for a recipient callout together
3984 with the "random" option, the postmaster address was used as the MAIL
3985 FROM address for the random test, but not for the subsequent recipient
3986 test. It is now used for both.
3988 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3989 patch removes a few documentation additions to RFC 3028, because the
3990 latest draft now contains them. It adds the new en;ascii-case comparator
3991 and a new error check for 8bit text in MIME parts. Comparator and
3992 require names are now matched exactly. I enabled the subaddress
3993 extension, but it is not well tested yet (read: it works for me)."
3995 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3996 rework some of the code of TK/09 above to avoid the hardwired use of
3997 "%lld" and "long long". Replaced the call to snprintf() with a call to
4000 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4001 records point to non-existent hosts", "retry timeout exceeded", and
4002 "retry time not reached for any host after a long failure period".
4004 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4005 experimental DomainKeys support:
4007 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4008 (2) On an error such as an illegally used "control", the wrong name for
4009 the control was given.
4011 These problems did NOT occur unless DomainKeys support was compiled.
4013 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4015 PH/24 Added ${if match_ip condition.
4017 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4018 databases so that it will be absolutely obvious if a crash occurs in the
4019 DB library. This is a regular occurrence (often caused by mis-matched
4022 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4023 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4024 picked up on a user's system that detects such things. There doesn't seem
4025 to be a gcc warning option for this - only an attribute that has to be
4026 put on the function's prototype. It seems that in Fedora Core 4 they have
4027 set this on a number of new functions. No doubt there will be more in due
4030 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4031 specify a transport. However, if an address that was verified by such a
4032 router was the subject of a callout, Exim crashed because it tried to
4033 read the rcpt_include_affixes from the non-existent transport. Now it
4034 just assumes that the setting of that option is false. This bug was
4035 introduced by 4.51/PH/31.
4037 PH/28 Changed -d+all to exclude +memory, because that information is very
4038 rarely of interest, but it makes the output a lot bigger. People tend to
4039 do -d+all out of habit.
4041 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4042 code in os-type was giving problems when libc.so lives in lib64, like on
4045 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4046 aren't the modern standard, and it seems that some systems' include files
4047 don't always have them. Exim was already checking for some of the newer
4048 ones like T_AAAA, and defining it itself. I've added checks for all the
4049 record types that Exim uses.
4051 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4052 not automatically generating a new one, as it is supposed to. This
4053 prevented TLS from working. If the file did exist, but contained invalid
4054 data, a new version was generated, as expected. It was only the case of a
4055 non-existent file that was broken.
4057 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4058 with a change in libdomainkeys > 0.64.
4060 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4061 from DNS. If the selector record carries the flag, it now has
4062 precedence over the domain-wide flag.
4064 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4066 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4067 the use of a port name that isn't defined in /etc/services) occurred, the
4068 message was deferred as in a normal delivery, and thus remained on the
4069 spool, instead of being failed because of the mua_wrapper setting. This
4070 is now fixed, and I tidied up some of the mua_wrapper messages at the
4073 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4074 the messages in a hash of arrays rather than using individual hashes.
4075 This is a bit cleaner and results in dramatic memory savings, albeit
4076 at a slight CPU cost.
4078 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4079 as requested by Marc Sherman.
4081 SC/10 Eximstats: added histograms for user specified patterns as requested
4084 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4086 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4087 fopen() in the content-scanning modules that did not already have it.
4093 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4094 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4096 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4098 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4100 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4101 to test Sieve filters that use "vacation".
4103 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4104 that changes the way the GnuTLS parameters are stored in the cache file.
4105 The new format can be generated externally. For backward compatibility,
4106 if the data in the cache doesn't make sense, Exim assumes it has read an
4107 old-format file, and it generates new data and writes a new file. This
4108 means that you can't go back to an older release without removing the
4111 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4112 work if there are any delivery delays because "one_time" forces the
4113 parent to be marked "delivered", so its unseen clone is never tried
4114 again. For this reason, Exim now forbids the simultaneous setting of
4117 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4118 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4119 entry if you want to know the details. The fix, however, neglected to
4120 consider the case when local delivery batching is involved. The test for
4121 "previously delivered" was not happening when checking to see if an
4122 address could be batched with a previous (undelivered) one; under
4123 certain circumstances this could lead to multiple deliveries to the same
4126 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4127 in its include files, and this causes problems building Exim.
4129 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4130 header_syntax) but Exim was just ignoring anything given after a slash.
4131 In particular, this caused confusion with an attempt to use "verify =
4132 reverse_host_lookup/defer_ok". An error is now given when options are
4133 supplied for verify items that do not have them. (Maybe reverse_host_
4134 lookup should have a defer_ok option, but that's a different point.)
4136 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4137 defined by RFC 821) to 2048, because there were problems with some AUTH
4138 commands, and RFC 1869 says the size should be increased for extended
4139 SMTP commands that take arguments.
4141 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4144 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4145 "unknown" error; now it says that the functionality isn't in the binary.
4147 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4148 an address' error message when a string expansion fails (syntax or
4149 whatever). Otherwise the password may appear in the log. Following change
4150 PH/42 below, there is no longer a chance of it appearing in a bounce
4153 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4155 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4156 of its addresses was ever tried. (Bugzilla bug #2.)
4158 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4159 the result incorrectly in the debug output. (It correctly added a newline
4160 to what was transported.)
4162 TF/01 Added $received_time.
4164 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4165 commented out examples of how to interface to a virus scanner and to
4166 SpamAssassin. Also added commented examples of av_scanner and
4167 spamd_address settings.
4169 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4170 and controls are allowed in which ACLs. There were a couple of minor
4171 errors. Some of the entries in the conditions table (which is a table of
4172 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4173 negation of where the condition IS allowed.
4175 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4177 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4178 header file does not have a version number, so I've had to invent a new
4179 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4180 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4181 radiusclient), but it was contributed by a Radius user.
4183 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4184 files or directories, for OpenSSL.
4186 PH/19 When an Exim process that is running as root has to create an Exim log
4187 file, it does so in a subprocess that runs as exim:exim so as to get the
4188 ownership right at creation (otherwise, other Exim processes might see
4189 the file with the wrong ownership). There was no test for failure of this
4190 fork() call, which would lead to the process getting stuck as it waited
4191 for a non-existent subprocess. Forks do occasionally fail when resources
4192 run out. I reviewed all the other calls to fork(); they all seem to check
4195 PH/20 When checking for unexpected SMTP input at connect time (before writing
4196 the banner), Exim was not dealing correctly with a non-positive return
4197 from the read() function. If the client had disconnected by this time,
4198 the result was a log entry for a synchronization error with an empty
4199 string after "input=" when read() returned zero. If read() returned -1
4200 (an event I could not check), uninitialized data bytes were printed.
4201 There were reports of junk text (parts of files, etc) appearing after
4204 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4206 PH/22 Added support for macro redefinition, and (re)definition in between
4207 driver and ACL definitions.
4209 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4210 forgetting to use the resulting value; it was using the unexpanded value.
4212 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4213 hadn't been configured. The fix is from Juergen Kreileder, who
4214 understands it better than I do:
4216 "Here's what I see happening with three configured cyrus_sasl
4217 authenticators configured (plain, login, cram-md5):
4219 On startup auth_cyrus_sasl_init() gets called for each of these.
4220 This means three calls to sasl_listmech() without a specified mech_list.
4221 => SASL tests which mechs of all available mechs actually work
4222 => three warnings about OTP not working
4223 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4225 With the patch, sasl_listmech() also gets called three times. But now
4226 SASL's mech_list option is set to the server_mech specified in the the
4227 authenticator. Or in other words, the answer from sasl_listmech()
4228 gets limited to just the mech you're testing for (which is different
4230 => the return list contains just 'plain' or 'login', 'cram-md5' or
4231 nothing depending on the value of ob->server_mech.
4233 I've just tested the patch: Authentication still works fine,
4234 unavailable mechs specified in the exim configuration are still
4235 caught, and the auth.log warnings about OTP are gone."
4237 PH/25 When debugging is enabled, the contents of the command line are added
4238 to the debugging output, even when log_selector=+arguments is not
4241 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4242 answer is "GNU", and only if the return is "GNU/something" is the answer
4245 PH/27 $acl_verify_message is now set immediately after the failure of a
4246 verification in an ACL, and so is available in subsequent modifiers. In
4247 particular, the message can be preserved by coding like this:
4249 warn !verify = sender
4250 set acl_m0 = $acl_verify_message
4252 Previously, $acl_verify_message was set only while expanding "message"
4253 and "log_message" when a very denied access.
4255 PH/28 Modified OS/os.c-Linux with
4257 -#ifndef OS_LOAD_AVERAGE
4258 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4260 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4261 nomenclature these days.)
4263 PH/29 Installed patch from the Sieve maintainer that adds the options
4264 sieve_useraddress and sieve_subaddress to the redirect router.
4266 PH/30 In these circumstances:
4267 . Two addresses routed to the same list of hosts;
4268 . First host does not offer TLS;
4269 . First host accepts first address;
4270 . First host gives temporary error to second address;
4271 . Second host offers TLS and a TLS session is established;
4272 . Second host accepts second address.
4273 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4274 and peerdn, if requested) that were in fact used only for the second
4277 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4278 attention to any local part prefix or suffix that was matched by the
4279 router that accepted the address. It now behaves in the same way as it
4280 does for delivery: the affixes are removed from the local part unless
4281 rcpt_include_affixes is set on the transport.
4283 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4284 timeout during the DATA phase of an incoming message.
4286 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4287 applied a patch sanctioned by the Sieve maintainer.
4289 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4290 the uid or gid is negative. A case of a negative gid caused this to be
4291 noticed. The fix allows for either to be negative.
4293 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4294 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4297 PH/36 The change PH/12 above was broken. Fixed it.
4299 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4300 the grounds that routing the same address twice would always produce the
4301 same answer. This might have been true once, but it is certainly no
4302 longer true now. Routing a child address may depend on the previous
4303 routing that produced that child. Some complicated redirection strategies
4304 went wrong when messages had multiple recipients, and made Exim's
4305 behaviour dependent on the order in which the addresses were given.
4307 I have moved the duplicate checking until after the routing is complete.
4308 Exim scans the addresses that are assigned to local and remote
4309 transports, and removes any duplicates. This means that more work will be
4310 done, as duplicates will always all be routed, but duplicates are
4311 presumably rare, so I don't expect this is of any significance.
4313 For deliveries to pipes, files, and autoreplies, the duplicate checking
4314 still happens during the routing process, since they are not going to be
4317 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4318 It corrects a timeout issue with spamd. This is Ian's comment: "The
4319 background is that sometimes spamd either never reads data from a
4320 connection it has accepted, or it never writes response data. The exiscan
4321 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4322 blindly assumes that writes won't block so it may never time out."
4324 PH/39 Allow G after quota size as well as K and M.
4326 PH/40 The value set for $authenticated_id in an authenticator may not contain
4327 binary zeroes or newlines because the value is written to log lines and
4328 to spool files. There was no check on this. Now the value is run through
4329 the string_printing() function so that such characters are converted to
4330 printable escape sequences.
4332 PH/41 $message_linecount is a new variable that contains the total number of
4333 lines in the message. Compare $body_linecount, which is the count for the
4336 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4337 bounce and delay warning messages, except in certain special cases, which
4340 (a) An SMTP error message from a remote host;
4341 (b) A message specified in a :fail: redirection;
4342 (c) A message specified in a "fail" command in a system filter;
4343 (d) A message specified in a FAIL return from the queryprogram router;
4344 (e) A message specified by the cannot_route_message router option.
4346 In these cases only, Exim does include the error details in bounce and
4347 warning messages. There are also a few cases where bland messages such
4348 as "unrouteable address" or "local delivery error" are given.
4350 PH/43 $value is now also set for the "else" part of a ${run expansion.
4352 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4353 being worked on, but at least Exim now implements the latest version to
4356 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4357 process to complete was treated as a delivery failure, a timeout while
4358 writing the message to the pipe was logged, but erroneously treated as a
4359 successful delivery. Such timeouts include transport filter timeouts. For
4360 consistency with the overall process timeout, these timeouts are now
4361 treated as errors, giving rise to delivery failures by default. However,
4362 there is now a new Boolean option for the pipe transport called
4363 timeout_defer, which, if set TRUE, converts the failures into defers for
4364 both kinds of timeout. A transport filter timeout is now identified in
4367 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4368 systems where "make" and "gmake" are different, calling "gmake" at top
4369 level broke things. I've arranged for the value of $(MAKE) to be passed
4370 from the Makefile to this script so that it can call the same version of
4374 A note about Exim versions 4.44 and 4.50
4375 ----------------------------------------
4377 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4378 changes of various kinds. As a consequence, a big documentation update was
4379 needed. This delayed the release for rather longer than seemed good, especially
4380 in the light of a couple of (minor) security issues. Therefore, the changes
4381 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4382 release. So 4.44 and 4.50 are in effect two different branches that both start
4385 I have left the 4.50 change log unchanged; it contains all the changes since
4386 4.43. The change log for 4.44 is below; many of its items are identical to
4387 those for 4.50. This seems to be the most sensible way to preserve the
4388 historical information.
4394 1. Minor wording change to the doc/README.SIEVE file.
4396 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4397 computation of the current number of files was incorrect.
4399 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4400 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4403 4. Give more explanation in the error message when the command for a transport
4404 filter fails to execute.
4406 5. There are several places where Exim runs a non-Exim command in a
4407 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4408 was being done only for the command run by the queryprogram router. It is
4409 now done for all such subprocesses. The other cases are: ${run, transport
4410 filters, and the commands run by the lmtp and pipe transports.
4412 6. Added CONFIGURE_GROUP build-time option.
4414 7. Some older OS have a limit of 256 on the maximum number of file
4415 descriptors. Exim was using setrlimit() to set 1000 as a large value
4416 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4417 systems. I've change it so that if it can't get 1000, it tries for 256.
4419 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4420 was an oversight, and furthermore, ever since the addition of extra
4421 controls (e.g. 4.43/32), the checks on when to allow different forms of
4422 "control" were broken. There should now be diagnostics for all cases when a
4423 control that does not make sense is encountered.
4425 9. Added the /retain_sender option to "control=submission".
4427 10. $recipients is now available in the predata ACL (oversight).
4429 11. Tidy the search cache before the fork to do a delivery from a message
4430 received from the command line. Otherwise the child will trigger a lookup
4431 failure and thereby defer the delivery if it tries to use (for example) a
4432 cached ldap connection that the parent has called unbind on.
4434 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4435 of $address_data from the recipient verification was clobbered by the
4436 sender verification.
4438 13. The value of address_data from a sender verification is now available in
4439 $sender_address_data in subsequent conditions in the ACL statement.
4441 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4443 15. Added a new option "connect=<time>" to callout options, to set a different
4446 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4447 was its contents. (It was OK if the option was not defined at all.)
4449 17. A "Completed" log line is now written for messages that are removed from
4450 the spool by the -Mrm option.
4452 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4453 information about exactly what failed.
4455 19. Added -dd to debug only the daemon process.
4457 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4458 handles timeouts, both on the server side and network timeouts. Renamed the
4459 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4461 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4462 It is now set to "smtps".
4464 22. $host_address is now set to the target address during the checking of
4465 ignore_target_hosts.
4467 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4468 being passed; this would have caused $sender_host_name to have been used if
4469 matching the list had actually called for a host name (not very likely,
4470 since this list is usually IP addresses). A host name is now passed as
4473 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4474 code that specifies a non-restarting handler (typically sigaction() in
4475 modern systems) in an attempt to fix a rare and obscure crash bug.
4477 25. Narrowed the window for a race in the daemon that could cause it to ignore
4478 SIGCHLD signals. This is not a major problem, because they are used only to
4479 wake it up if nothing else does.
4481 26. A malformed maildirsize file could cause Exim to calculate negative values
4482 for the mailbox size or file count. Odd effects could occur as a result.
4483 The maildirsize information is now recalculated if the size or filecount
4486 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4487 support for a long time. Removed HAVE_SYS_VFS_H.
4489 28. Installed the latest version of exipick from John Jetmore.
4491 29. In an address list, if the pattern was not a regular expression, an empty
4492 subject address (from a bounce message) matched only if the pattern was an
4493 empty string. Non-empty patterns were not even tested. This was the wrong
4494 because it is perfectly reasonable to use an empty address as part of a
4495 database query. An empty address is now tested by patterns that are
4496 lookups. However, all the other forms of pattern expect the subject to
4497 contain a local part and a domain, and therefore, for them, an empty
4498 address still always fails if the pattern is not itself empty.
4500 30. Exim went into a mad DNS loop when attempting to do a callout where the
4501 host was specified on an smtp transport, and looking it up yielded more
4502 than one IP address.
4504 31. Re-factored the code for checking spool and log partition space into a
4505 function that finds that data and another that does the check. The former
4506 is then used to implement four new variables: $spool_space, $log_space,
4507 $spool_inodes, and $log_inodes.
4509 32. The RFC2047 encoding function was originally intended for short strings
4510 such as real names; it was not keeping to the 75-character limit for
4511 encoded words that the RFC imposes. It now respects the limit, and
4512 generates multiple encoded words if necessary. To be on the safe side, I
4513 have increased the buffer size for the ${rfc2047: expansion operator from
4516 33. It is now permitted to omit both strings after an "if" condition; if the
4517 condition is true, the result is "true". As before, when the second string
4518 is omitted, a false condition yields an empty string. This makes it less
4519 cumbersome to write custom ACL and router conditions.
4521 34. Failure to deliver a bounce message always caused it to be frozen, even if
4522 there was an errors_to setting on the router. The errors_to setting is now
4525 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4526 canonical form (fully expanded) before being placed in
4527 $sender_host_address.
4529 36. The table in the code that translates DNS record types into text (T_A to
4530 "A" for instance) was missing entries for NS and CNAME. It is just possible
4531 that this could have caused confusion if both these types were looked up
4532 for the same domain, because the text type is used as part of Exim's
4533 per-process caching. But the chance of anyone hitting this buglet seems
4536 37. The dnsdb lookup has been extended in a number of ways.
4538 (1) There is a new type, "zns", which walks up the domain tree until it
4539 finds some nameserver records. It should be used with care.
4541 (2) There is a new type, "mxh", which is like "mx" except that it returns
4542 just the host names, not the priorities.
4544 (3) It is now possible to give a list of domains (or IP addresses) to be
4545 looked up. The behaviour when one of the lookups defers can be
4546 controlled by a keyword.
4548 (4) It is now possible to specify the separator character for use when
4549 multiple records are returned.
4551 38. The dnslists ACL condition has been extended: it is now possible to supply
4552 a list of IP addresses and/or domains to be looked up in a particular DNS
4555 39. Added log_selector=+queue_time_overall.
4557 40. When running the queue in the test harness, wait just a tad after forking a
4558 delivery process, to get repeatability of debugging output.
4560 41. Include certificate and key file names in error message when GnuTLS fails
4561 to set them up, because the GnuTLS error message doesn't include the name
4562 of the failing file when there is a problem reading it.
4564 42. Allow both -bf and -bF in the same test run.
4566 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4568 44. The "Exiscan patch" is now merged into the mainline Exim source.
4570 45. Sometimes the final signoff response after QUIT could fail to get
4571 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4572 < 0 before doing a fflush(). This bug looks as though it goes back to the
4573 introduction of TLS in release 3.20, but "sometimes" must have been rare
4574 because the tests only now provoked it.
4576 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4577 (this can affect the format of dates).
4579 47. exim_tidydb, when checking for the continued existence of a message for
4580 which it has found a message-specific retry record, was not finding
4581 messages that were in split spool directories. Consequently, it was
4582 deleting retry records that should have stayed in existence.
4584 48. Steve fixed some bugs in eximstats.
4586 49. The SPA authentication driver was not abandoning authentication and moving
4587 on to the next authenticator when an expansion was forced to fail,
4588 contradicting the general specification for all authenticators. Instead it
4589 was generating a temporary error. It now behaves as specified.
4591 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4592 (the order specifies the preference for clients). The order is now AES256,
4593 AES128, 3DES, ARCFOUR128.
4595 51. Small patch to Sieve code - explicitly set From: when generating an
4598 52. Exim crashed if a remote delivery caused a very long error message to be
4599 recorded - for instance if somebody sent an entire SpamAssassin report back
4600 as a large number of 550 error lines. This bug was coincidentally fixed by
4601 increasing the size of one of Exim's internal buffers (big_buffer) that
4602 happened as part of the Exiscan merge. However, to be on the safe side, I
4603 have made the code more robust (and fixed the comments that describe what
4606 53. Now that there can be additional text after "Completed" in log lines (if
4607 the queue_time_overall log selector is set), a one-byte patch to exigrep
4608 was needed to allow it to recognize "Completed" as not the last thing in
4611 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4612 patch that reportedly fixes this has been added. I am not expert enough to
4613 create a test for it. This is what the patch creator wrote:
4615 "I found a little strange behaviour of ldap code when working with
4616 Windows 2003 AD Domain, where users was placed in more than one
4617 Organization Units. When I tried to give exim partial DN, the exit code
4618 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4619 But simultaneously result of request was absolutely normal ldap result,
4620 so I produce this patch..."
4622 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4623 REFERENCE, so I have modified the code to exclude the patch when that macro
4626 55. Some experimental protocols are using DNS PTR records for new purposes. The
4627 keys for these records are domain names, not reversed IP addresses. The
4628 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4629 leaves it alone. Component reversal etc. now happens only for IP addresses.
4632 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4634 57. Double the size of the debug message buffer (to 2048) so that more of very
4635 long debug lines gets shown.
4637 58. The exicyclog utility now does better if the number of log files to keep
4638 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4640 59. Two changes related to the smtp_active_hostname option:
4642 (1) $smtp_active_hostname is now available as a variable.
4643 (2) The default for smtp_banner uses $smtp_active_hostname instead
4644 of $primary_hostname.
4646 60. The host_aton() function is supposed to be passed a string that is known
4647 to be a valid IP address. However, in the case of IPv6 addresses, it was
4648 not checking this. This is a hostage to fortune. Exim now panics and dies
4649 if the condition is not met. A case was found where this could be provoked
4650 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4651 components; fortuitously, this particular loophole had already been fixed
4652 by change 4.50/55 above.
4654 If there are any other similar loopholes, the new check in host_aton()
4655 itself should stop them being exploited. The report I received stated that
4656 data on the command line could provoke the exploit when Exim was running as
4657 exim, but did not say which command line option was involved. All I could
4658 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4659 running as the user.
4662 61. There was a buffer overflow vulnerability in the SPA authentication code
4663 (which came originally from the Samba project). I have added a test to the
4664 spa_base64_to_bits() function which I hope fixes it.
4667 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4668 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4670 63. The daemon start-up calls getloadavg() while still root for those OS that
4671 need the first call to be done as root, but it missed one case: when
4672 deliver_queue_load_max is set with deliver_drop_privilege. This is
4673 necessary for the benefit of the queue runner, because there is no re-exec
4674 when deliver_drop_privilege is set.
4676 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4677 This has been fixed.
4679 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4680 was in use, was not putting the data itself into the right store pool;
4681 consequently, it could be overwritten for a subsequent message in the same
4682 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4685 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4687 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4688 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4689 the function were treating the return as a boolean value, which happened to
4690 work because 0=false and not-0=true, but is not correct code.
4692 68. The host_aton() function was not handling scoped IPv6 addresses (those
4693 with, for example, "%eth0" on the end) correctly.
4695 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4696 negated items (that is, ~something) in unsigned ints. Some compilers
4697 apparently mutter when there is no cast.
4699 70. If an address verification called from an ACL failed, and did not produce a
4700 user-specific message (i.e. there was only a "system" message), nothing was
4701 put in $acl_verify_message. In this situation, it now puts the system
4704 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4705 session; change 4.31/43 added the unwanted input to the log line - except
4706 that it did not do this in the start of session case. It now does.
4708 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4709 This is wrong and can cause the other end to generate a synchronization
4710 error if it is another Exim or anything else that does the synchronization
4711 check. A QUIT command is no longer sent after a timeout.
4713 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4714 during host lookups.
4716 74. The defer_ok option of callout verification was not working if it was used
4717 when verifying addresses in header lines, that is, for this case:
4719 verify = header_sender/callout=defer_ok
4721 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4722 those file descriptors could be used for SMTP connections. If anything
4723 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4724 could be sent to the SMTP client, causing chaos. The daemon now opens
4725 stdin, stdout, and stderr to /dev/null when it puts itself into the
4728 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4729 log by default. The user can override this with suitable Perl magic.
4731 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4732 the log message when discard triggers, was not working for the DATA ACL or
4733 for the non-SMTP ACL.
4735 78. Error message wording change in sieve.c.
4737 79. If smtp_accept_max_per_host was set, the number of connections could be
4738 restricted to fewer than expected, because the daemon was trying to set up
4739 a new connection before checking whether the processes handling previous
4740 connections had finished. The check for completed processes is now done
4741 earlier. On busy systems, this bug wouldn't be noticed because something
4742 else would have woken the daemon, and it would have reaped the completed
4745 80. If a message was submitted locally by a user whose login name contained one
4746 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4747 It caused a spool format error. I have fixed the spool reading code. A
4748 related problem was that the "from" clause in the Received: line became
4749 illegal because of the space(s). It is now covered by ${quote_local_part.
4751 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4754 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4756 83. Patch from Sieve maintainer to fix unterminated string problem in
4757 "vacation" handling.
4759 84. Some minor changes to the Linux configuration files to help with other
4760 OS variants using glibc.
4762 85. One more patch for Sieve to update vacation handling to latest spec.
4765 ----------------------------------------------------
4766 See the note above about the 4.44 and 4.50 releases.
4767 ----------------------------------------------------
4773 1. Change 4.43/35 introduced a bug that caused file counts to be
4774 incorrectly computed when quota_filecount was set in an appendfile
4777 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4778 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4781 3. Give more explanation in the error message when the command for a transport
4782 filter fails to execute.
4784 4. There are several places where Exim runs a non-Exim command in a
4785 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4786 was being done only for the command run by the queryprogram router. It is
4787 now done for all such subprocesses. The other cases are: ${run, transport
4788 filters, and the commands run by the lmtp and pipe transports.
4790 5. Some older OS have a limit of 256 on the maximum number of file
4791 descriptors. Exim was using setrlimit() to set 1000 as a large value
4792 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4793 systems. I've change it so that if it can't get 1000, it tries for 256.
4795 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4796 was an oversight, and furthermore, ever since the addition of extra
4797 controls (e.g. 4.43/32), the checks on when to allow different forms of
4798 "control" were broken. There should now be diagnostics for all cases when a
4799 control that does not make sense is encountered.
4801 7. $recipients is now available in the predata ACL (oversight).
4803 8. Tidy the search cache before the fork to do a delivery from a message
4804 received from the command line. Otherwise the child will trigger a lookup
4805 failure and thereby defer the delivery if it tries to use (for example) a
4806 cached ldap connection that the parent has called unbind on.
4808 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4809 of $address_data from the recipient verification was clobbered by the
4810 sender verification.
4812 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4813 was its contents. (It was OK if the option was not defined at all.)
4815 11. A "Completed" log line is now written for messages that are removed from
4816 the spool by the -Mrm option.
4818 12. $host_address is now set to the target address during the checking of
4819 ignore_target_hosts.
4821 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4822 being passed; this would have caused $sender_host_name to have been used if
4823 matching the list had actually called for a host name (not very likely,
4824 since this list is usually IP addresses). A host name is now passed as
4827 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4828 code that specifies a non-restarting handler (typically sigaction() in
4829 modern systems) in an attempt to fix a rare and obscure crash bug.
4831 15. Narrowed the window for a race in the daemon that could cause it to ignore
4832 SIGCHLD signals. This is not a major problem, because they are used only to
4833 wake it up if nothing else does.
4835 16. A malformed maildirsize file could cause Exim to calculate negative values
4836 for the mailbox size or file count. Odd effects could occur as a result.
4837 The maildirsize information is now recalculated if the size or filecount
4840 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4841 support for a long time. Removed HAVE_SYS_VFS_H.
4843 18. Updated exipick to current release from John Jetmore.
4845 19. Allow an empty sender to be matched against a lookup in an address list.
4846 Previously the only cases considered were a regular expression, or an
4849 20. Exim went into a mad DNS lookup loop when doing a callout where the
4850 host was specified on the transport, if the DNS lookup yielded more than
4853 21. The RFC2047 encoding function was originally intended for short strings
4854 such as real names; it was not keeping to the 75-character limit for
4855 encoded words that the RFC imposes. It now respects the limit, and
4856 generates multiple encoded words if necessary. To be on the safe side, I
4857 have increased the buffer size for the ${rfc2047: expansion operator from
4860 22. Failure to deliver a bounce message always caused it to be frozen, even if
4861 there was an errors_to setting on the router. The errors_to setting is now
4864 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4865 canonical form (fully expanded) before being placed in
4866 $sender_host_address.
4868 24. Updated eximstats to version 1.33
4870 25. Include certificate and key file names in error message when GnuTLS fails
4871 to set them up, because the GnuTLS error message doesn't include the name
4872 of the failing file when there is a problem reading it.
4874 26. Expand error message when OpenSSL has problems setting up cert/key files.
4877 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4878 (this can affect the format of dates).
4880 28. exim_tidydb, when checking for the continued existence of a message for
4881 which it has found a message-specific retry record, was not finding
4882 messages that were in split spool directories. Consequently, it was
4883 deleting retry records that should have stayed in existence.
4885 29. eximstats updated to version 1.35
4886 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4887 1.35 - bugfix such that pie charts by volume are generated correctly
4889 30. The SPA authentication driver was not abandoning authentication and moving
4890 on to the next authenticator when an expansion was forced to fail,
4891 contradicting the general specification for all authenticators. Instead it
4892 was generating a temporary error. It now behaves as specified.
4894 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4895 (the order specifies the preference for clients). The order is now AES256,
4896 AES128, 3DES, ARCFOUR128.
4898 31. Small patch to Sieve code - explicitly set From: when generating an
4901 32. Exim crashed if a remote delivery caused a very long error message to be
4902 recorded - for instance if somebody sent an entire SpamAssassin report back
4903 as a large number of 550 error lines. This bug was coincidentally fixed by
4904 increasing the size of one of Exim's internal buffers (big_buffer) that
4905 happened as part of the Exiscan merge. However, to be on the safe side, I
4906 have made the code more robust (and fixed the comments that describe what
4909 33. Some experimental protocols are using DNS PTR records for new purposes. The
4910 keys for these records are domain names, not reversed IP addresses. The
4911 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4912 leaves it alone. Component reversal etc. now happens only for IP addresses.
4915 34. The host_aton() function is supposed to be passed a string that is known
4916 to be a valid IP address. However, in the case of IPv6 addresses, it was
4917 not checking this. This is a hostage to fortune. Exim now panics and dies
4918 if the condition is not met. A case was found where this could be provoked
4919 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4920 components; fortuitously, this particular loophole had already been fixed
4921 by change 4.50/55 or 4.44/33 above.
4923 If there are any other similar loopholes, the new check in host_aton()
4924 itself should stop them being exploited. The report I received stated that
4925 data on the command line could provoke the exploit when Exim was running as
4926 exim, but did not say which command line option was involved. All I could
4927 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4928 running as the user.
4931 35. There was a buffer overflow vulnerability in the SPA authentication code
4932 (which came originally from the Samba project). I have added a test to the
4933 spa_base64_to_bits() function which I hope fixes it.
4936 36. The daemon start-up calls getloadavg() while still root for those OS that
4937 need the first call to be done as root, but it missed one case: when
4938 deliver_queue_load_max is set with deliver_drop_privilege. This is
4939 necessary for the benefit of the queue runner, because there is no re-exec
4940 when deliver_drop_privilege is set.
4942 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4943 was in use, was not putting the data itself into the right store pool;
4944 consequently, it could be overwritten for a subsequent message in the same
4945 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4948 38. Sometimes the final signoff response after QUIT could fail to get
4949 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4950 < 0 before doing a fflush(). This bug looks as though it goes back to the
4951 introduction of TLS in release 3.20, but "sometimes" must have been rare
4952 because the tests only now provoked it.
4958 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4959 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4960 Now it returns an integer. A number of calls were still expecting a T/F
4961 return. Fortuitously, in all cases, the tests worked in OK situations,
4962 which is the norm. However, things would have gone wrong on any write
4963 failures on the smtp file descriptor. This function is used when sending
4964 messages over SMTP and also when doing verify callouts.
4966 2. When Exim is called to do synchronous delivery of a locally submitted
4967 message (the -odf or -odi options), it no longer closes stderr before doing
4970 3. Implemented the mua_wrapper option.
4972 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4974 5. Implemented the functions header_remove(), header_testname(),
4975 header_add_at_position(), and receive_remove_recipient(), and exported them
4978 6. If an ACL "warn" statement specified the addition of headers, Exim already
4979 inserted X-ACL-Warn: at the start if there was no header name. However, it
4980 was not making this test for the second and subsequent header lines if
4981 there were newlines in the string. This meant that an invalid header could
4982 be inserted if Exim was badly configured.
4984 7. Allow an ACL "warn" statement to add header lines at the start or after all
4985 the Received: headers, as well as at the end.
4987 8. Added the rcpt_4xx retry error code.
4989 9. Added postmaster_mailfrom=xxx to callout verification option.
4991 10. Added mailfrom=xxxx to the callout verification option, for verify=
4994 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4995 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4997 12. Inserted some casts to stop certain compilers warning when using pointer
4998 differences as field lengths or precisions in printf-type calls (mostly
4999 affecting debugging statements).
5001 13. Added optional readline() support for -be (dynamically loaded).
5003 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5004 same clock tick as a message's arrival, so that its received time was the
5005 same as the "first fail" time on the retry record, and that message
5006 remained on the queue past the ultimate address timeout, every queue runner
5007 would try a delivery (because it was past the ultimate address timeout) but
5008 after another failure, the ultimate address timeout, which should have then
5009 bounced the address, did not kick in. This was a "< instead of <=" error;
5010 in most cases the first failure would have been in the next clock tick
5011 after the received time, and all would be well.
5013 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5014 being recognized when the domain list was tested by the match_domain
5015 condition in an expansion string.
5017 16. Added the ${str2b64: operator.
5019 17. Exim was always calling setrlimit() to set a large limit for the number of
5020 processes, without checking whether the existing limit was already
5021 adequate. (It did check for the limit on file descriptors.) Furthermore,
5022 errors from getrlimit() and setrlimit() were being ignored. Now they are
5023 logged to the main and panic logs, but Exim does carry on, to try to do its
5024 job under whatever limits there are.
5026 18. Imported PCRE 5.0.
5028 19. Trivial typo in log message " temporarily refused connection" (the leading
5031 20. If the log selector return_path_on_delivery was set and an address was
5032 redirected to /dev/null, the delivery process crashed because it assumed
5033 that a return path would always be set for a "successful" delivery. In this
5034 case, the whole delivery is bypassed as an optimization, and therefore no
5037 21. Internal re-arrangement: the function for sending a challenge and reading
5038 a response while authentication was assuming a zero-terminated challenge
5039 string. It's now changed to take a pointer and a length, to allow for
5040 binary data in such strings.
5042 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5044 23. Exim was not respecting finduser_retries when seeking the login of the
5045 uid under which it was called; it was always trying 10 times. (The default
5046 setting of finduser_retries is zero.) Also, it was sleeping after the final
5047 failure, which is pointless.
5049 24. Implemented tls_on_connect_ports.
5051 25. Implemented acl_smtp_predata.
5053 26. If the domain in control=submission is set empty, Exim assumes that the
5054 authenticated id is a complete email address when it generates From: or
5055 Sender: header lines.
5057 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5058 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5059 chown and chgrp in /bin and hostname in /usr/bin.
5061 28. Exim was keeping the "process log" file open after each use, just as it
5062 does for the main log. This opens the possibility of it remaining open for
5063 long periods when the USR1 signal hits a daemon. Occasional processlog
5064 errors were reported, that could have been caused by this. Anyway, it seems
5065 much more sensible not to leave this file open at all, so that is what now
5068 29. The long-running daemon process does not normally write to the log once it
5069 has entered its main loop, and it closes the log before doing so. This is
5070 so that log files can straightforwardly be renamed and moved. However,
5071 there are a couple of unusual error situations where the daemon does write
5072 log entries, and I had neglected to close the log afterwards.
5074 30. The text of an SMTP error response that was received during a remote
5075 delivery was being truncated at 512 bytes. This is too short for some of
5076 the long messages that one sometimes sees. I've increased the limit to
5079 31. It is now possible to make retry rules that apply only when a message has a
5080 specific sender, in particular, an empty sender.
5082 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5083 it possible to be selective about when SMTP synchronization is enforced.
5085 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5087 32. Implemented hosts_connection_nolog.
5089 33. Added an ACL for QUIT.
5091 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5094 35. Added mailbox_size and mailbox_filecount to appendfile.
5096 36. Added control = no_multiline_responses to ACLs.
5098 37. There was a bug in the logic of the code that waits for the clock to tick
5099 in the case where the clock went backwards by a substantial amount such
5100 that the microsecond fraction of "now" was more than the microsecond
5101 fraction of "then" (but the whole seconds number was less).
5103 38. Added support for the libradius Radius client library this is found on
5104 FreeBSD (previously only the radiusclient library was supported).
5110 1. When certain lookups returned multiple values in the form name=value, the
5111 quoting of the values was not always being done properly. Specifically:
5112 (a) If the value started with a double quote, but contained no whitespace,
5114 (b) If the value contained whitespace other than a space character (i.e.
5115 tabs or newlines or carriage returns) it was not quoted.
5116 This fix has been applied to the mysql and pgsql lookups by writing a
5117 separate quoting function and calling it from the lookup code. The fix
5118 should probably also be applied to nisplus, ibase and oracle lookups, but
5119 since I cannot test any of those, I have not disturbed their existing code.
5121 2. A hit in the callout cache for a specific address caused a log line with no
5122 reason for rejecting RCPT. Now it says "Previous (cached) callout
5123 verification failure".
5125 3. There was an off-by-one bug in the queryprogram router. An over-long
5126 return line was truncated at 256 instead of 255 characters, thereby
5127 overflowing its buffer with the terminating zero. As well as fixing this, I
5128 have increased the buffer size to 1024 (and made a note to document this).
5130 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5131 when Exim is waiting for an SMTP response from a remote server, Exim
5132 restarts its select() call on the socket, thereby resetting its timeout.
5133 This is not a problem when such interrupts are rare. Somebody set up a cron
5134 job to run exiwhat every 2 minutes, which is less than the normal select()
5135 timeout (5 or 10 minutes). This meant that the select() timeout never
5136 kicked in because it was always reset. I have fixed this by comparing the
5137 time when an interrupt arrives with the time at the start of the first call
5138 to select(). If more time than the timeout has elapsed, the interrupt is
5139 treated as a timeout.
5141 5. Some internal re-factoring in preparation for the addition of Sieve
5142 extensions (by MH). In particular, the "personal" test is moved to a
5143 separate function, and given an option for scanning Cc: and Bcc: (which is
5144 not set for Exim filters).
5146 6. When Exim created an email address using the login of the caller as the
5147 local part (e.g. when creating a From: or Sender: header line), it was not
5148 quoting the local part when it contained special characters such as @.
5150 7. Installed new OpenBSD configuration files.
5152 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5153 try to make them clearer.
5155 9. Callout options, other than the timeout value, were being ignored when
5156 verifying sender addresses in header lines. For example, when using
5158 verify = header_sender/callout=no_cache
5160 the cache was (incorrectly) being used.
5162 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5163 only the Cygwin environment.
5165 11. When return_path_on_delivery was set as a log selector, if different remote
5166 addresses in the same message used different return paths and parallel
5167 remote delivery occurred, the wrong values would sometimes be logged.
5168 (Whenever a remote delivery process finished, the return path value from
5169 the most recently started remote delivery process was logged.)
5171 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5172 lines when AUTH and/or TLS are in use. This is the "received protocol"
5173 field. Exim used to use "asmtp" for authenticated SMTP, without any
5174 indication (in the protocol name) for TLS use. Now it follows the RFC and
5175 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5176 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5177 names appear in log lines as well as in Received: header lines.
5179 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5180 extensions, and comparison tests, and to fix some bugs.
5182 14. Changes to the "personal" filter test:
5184 (1) The test was buggy in that it was just doing the equivalent of
5185 "contains" tests on header lines. For example, if a user's address was
5186 anne@some.where, the "personal" test would incorrectly be true for
5188 To: susanne@some.where
5190 This test is now done by extracting each address from the header in turn,
5191 and checking the entire address. Other tests that are part of "personal"
5192 are now done using regular expressions (for example, to check local parts
5193 of addresses in From: header lines).
5195 (2) The list of non-personal local parts in From: addresses has been
5196 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5197 taken from the Sieve specification recommendations.
5199 (3) If the message contains any header line starting with "List-" it is
5200 treated as non-personal.
5202 (4) The test for "circular" in the Subject: header line has been removed
5203 because it now seems ill-conceived.
5205 15. Minor typos in src/EDITME comments corrected.
5207 16. Installed latest exipick from John Jetmore.
5209 17. If headers_add on a router specified a text string that was too long for
5210 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5211 of string_sprintf() is now avoided.
5213 18. $message_body_size was not set (it was always zero) when running the DATA
5214 ACL and the local_scan() function.
5216 19. For the "mail" command in an Exim filter, no default was being set for
5217 the once_repeat time, causing a random time value to be used if "once" was
5218 specified. (If the value happened to be <= 0, no repeat happened.) The
5219 default is now 0s, meaning "never repeat". The "vacation" command was OK
5220 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5221 (I found it when inspecting the code).
5223 20. There is now an overall timeout for performing a callout verification. It
5224 defaults to 4 times the callout timeout, which applies to individual SMTP
5225 commands during the callout. The overall timeout applies when there is more
5226 than one host that can be tried. The timeout is checked before trying the
5227 next host. This prevents very long delays if there are a large number of
5228 hosts and all are timing out (e.g. when the network connections are timing
5229 out). The value of the overall timeout can be changed by specifying an
5230 additional sub-option for "callout", called "maxwait". For example:
5232 verify = sender/callout=5s,maxwait=20s
5234 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5235 to the end before writing, but this should make it even safer).
5237 22. Exim was forgetting that it had advertised PIPELINING for the second and
5238 subsequent messages on an SMTP connection. It was also not resetting its
5239 memory on STARTTLS and an internal HELO.
5241 23. When Exim logs an SMTP synchronization error within a session, it now
5242 records whether PIPELINING has been advertised or not.
5244 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5245 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5246 rather than long int.
5248 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5250 26. Added the never_mail option to autoreply.
5256 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5257 crash if the getsockname() call failed; this can happen if a connection is
5258 closed very soon after it is established. The problem was simply in the
5259 order in which certain operations were done, causing Exim to try to write
5260 to the SMTP stream before it had set up the file descriptor. The bug has
5261 been fixed by making things happen in the correct order.
5267 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5268 before the connection was closed, thus losing the rejection response.
5270 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5271 some early Solaris releases, but causes trouble in current releases where
5272 socklen_t is defined.
5274 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5277 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5280 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5281 that releasing the top of it at the end releases what was used for sub-
5282 expansions (unless the block got too big). However, discard this block if
5283 the first thing is a variable or header, so that we can use its block when
5284 it is dynamic (useful for very large $message_headers, for example).
5286 6. Lookups now cache *every* query, not just the most recent. A new, separate
5287 store pool is used for this. It can be recovered when all lookup caches are
5288 flushed. Lookups now release memory at the end of their result strings.
5289 This has involved some general refactoring of the lookup sources.
5291 7. Some code has been added to the store_xxx() functions to reduce the amount
5292 of flapping under certain conditions.
5294 8. log_incoming_interface used to affect only the <= reception log lines. Now
5295 it causes the local interface and port to be added to several more SMTP log
5296 lines, for example "SMTP connection from", and rejection lines.
5298 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5300 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5302 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5303 could be overwritten at the end of the current message (or the start of a
5304 new message if it was set in a HELO ACL). The value is now preserved for
5305 the duration of the SMTP connection.
5307 12. If a transport had a headers_rewrite setting, and a matching header line
5308 contained an unqualified address, that address was qualified, even if it
5309 did not match any rewriting rules. The underlying bug was that the values
5310 of the flags that permit the existence of unqualified sender and recipient
5311 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5312 non-local messages, and by -bnq for local messages) were not being
5313 preserved with the message after it was received.
5315 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5316 "next input=" as part of the text comprising the host identity instead of
5317 the correct text. The code was using the same buffer for two different
5318 strings. However, depending on which order the printing function evaluated
5319 its arguments, the bug did not always show up. Under Linux, for example, my
5320 test suite worked just fine.
5322 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5323 doesn't work with some older versions of Perl. It has been changed to "my",
5324 which in any case is probably the better facility to use.
5326 15. A really picky compiler found some instances of statements for creating
5327 error messages that either had too many or two few arguments for the format
5330 16. The size of the buffer for calls to the DNS resolver has been increased
5331 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5332 for addresses that have a lot of PTR records. This alleviates a problem; it
5333 does not fully solve it.
5335 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5336 buffer now truncates the list and logs the incident, which is the same
5337 action as happens when Exim is looking up a host name and its aliases.
5338 Previously in this situation something unpredictable would happen;
5339 sometimes it was "internal error: store_reset failed".
5341 18. If a server dropped the connection unexpectedly when an Exim client was
5342 using GnuTLS and trying to read a response, the client delivery process
5343 crashed while trying to generate an error log message.
5345 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5346 string, for example:
5348 warn message = H1: something\nH2: something
5350 the text was added as a single header line from Exim's point of view
5351 though it ended up OK in the delivered message. However, searching for the
5352 second and subsequent header lines using $h_h2: did not work. This has been
5353 fixed. Similarly, if a system filter added multiple headers in this way,
5354 the routers could not see them.
5356 20. Expanded the error message when iplsearch is called with an invalid key to
5357 suggest using net-iplsearch in a host list.
5359 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5360 ACLs are no longer actually imposed (and a message to that effect is
5363 22. If a "gecos" field in a passwd entry contained escaped characters, in
5364 particular, if it contained a \" sequence, Exim got it wrong when building
5365 a From: or a Sender: header from that name. A second bug also caused
5366 incorrect handling when an unquoted " was present following a character
5367 that needed quoting.
5369 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5370 was not being matched caselessly.
5372 24. Arranged for all hyphens in the exim.8 source to be escaped with
5375 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5376 properly. Recipient callout cache records were still being keyed to include
5377 the sender, even when use_sender was set false. This led to far more
5378 callouts that were necessary. The sender is no longer included in the key
5379 when use_sender is false.
5381 26. Added "control = submission" modifier to ACLs.
5383 27. Added the ${base62d: operator to decode base 62 numbers.
5385 28. dnsdb lookups can now access SRV records.
5387 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5388 the configuration file.
5390 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5391 (-v) mode. This makes the output for a verbose queue run more intelligible.
5393 31. Added a use_postmaster feature to recipient callouts.
5395 32. Added the $body_zerocount variable, containing the number of binary zero
5396 bytes in the message body.
5398 33. The time of last modification of the "new" subdirectory is now used as the
5399 "mailbox time last read" when there is a quota error for a maildir
5402 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5404 35. Added +ignore_unknown as a special item in host lists.
5406 36. Code for decoding IPv6 addresses in host lists is now included, even if
5407 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5408 address was recognized as an IP address, but was then not correctly decoded
5409 into binary, causing unexpected and incorrect effects when compared with
5416 1. Very minor rewording of debugging text in manualroute to say "list of
5417 hosts" instead of "hostlist".
5419 2. If verify=header_syntax was set, and a header line with an unqualified
5420 address (no domain) and a large number of spaces between the end of the
5421 name and the colon was received, the reception process suffered a buffer
5422 overflow, and (when I tested it) crashed. This was caused by some obsolete
5423 code that should have been removed. The fix is to remove it!
5425 3. When running in the test harness, delay a bit after writing a bounce
5426 message to get a bit more predictability in the log output.
5428 4. Added a call to search_tidyup() just before forking a reception process. In
5429 theory, someone could use a lookup in the expansion of smtp_accept_max_
5430 per_host which, without the tidyup, could leave open a database connection.
5432 5. Added the variables $recipient_data and $sender_data which get set from a
5433 lookup success in an ACL "recipients" or "senders" condition, or a router
5434 "senders" option, similar to $domain_data and $local_part_data.
5436 6. Moved the writing of debug_print from before to after the "senders" test
5439 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5440 problems for message scanning, either using a data ACL, or using
5441 local_scan() because the Received: header was not generated till after they
5442 were called (in order to set the time as the time of reception completion).
5443 I have revised the way this works. The header is now generated after the
5444 body is received, but before the ACL or local_scan() are called. After they
5445 are run, the timestamp in the header is updated.
5451 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5452 before starting a queue runner without re-exec. This happened only when
5453 deliver_drop_privilege was set or when the Exim user was set to root. The
5454 effect of the bug was that timeouts during subsequent deliveries caused
5455 crashes instead of being properly handled. The handler is now left at its
5456 default (and expected) setting.
5458 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5459 message, again when deliver_drop_privilege is set or Exim is run as root.
5460 The bug described in (1) was not present in this case, but the tidying up
5461 of the other signals was missing. I have made the two cases consistent.
5463 3. The ignore_target_hosts setting on a manualroute router was being ignored
5464 for hosts that were looked up using the /MX notation.
5466 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5469 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5470 operated on the sender address. After changing the $sender_address to <>
5471 for the sender address verify, Exim was re-instated it as the original
5472 (before rewriting) address, but remembering that it had rewritten it, so it
5473 wasn't rewriting it again. This bug also had the effect of breaking the
5474 sender address verification caching when the sender address was rewritten.
5476 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5477 This has been changed so that if the ip literal address matches
5478 ignore_target_hosts, the router declines.
5480 7. Added expansion conditions match_domain, match_address, and match_local_
5481 part (NOT match_host).
5483 8. The placeholder for the Received: header didn't have a length field set.
5485 9. Added code to Exim itself and to exim_lock to test for a specific race
5486 condition that could lead to file corruption when using MBX delivery. The
5487 issue is with the lockfile that is created in /tmp. If this file is removed
5488 after a process has opened it but before that process has acquired a lock,
5489 there is the potential for a second process to recreate the file and also
5490 acquire a lock. This could lead to two Exim processes writing to the file
5491 at the same time. The added code performs the same test as UW imapd; it
5492 checks after acquiring the lock that its file descriptor still refers to
5493 the same named file.
5495 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5496 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5499 11. Added the smtp_active_hostname option. If used, this will typically be made
5500 to depend on the incoming interface address. Because $interface_address is
5501 not set up until the daemon has forked a reception process, error responses
5502 that can happen earlier (such as "too many connections") no longer contain
5505 12. If an expansion in a condition on a "warn" statement fails because a lookup
5506 defers, the "warn" statement is abandoned, and the next ACL statement is
5507 processed. Previously this caused the whole ACL to be aborted.
5509 13. Added the iplsearch lookup type.
5511 14. Added ident_timeout as a log selector.
5513 15. Added tls_certificate_verified as a log selector.
5515 16. Added a global option tls_require_ciphers (compare the smtp transport
5516 option of the same name). This controls incoming TLS connections.
5518 17. I finally figured out how to make tls_require_ciphers do a similar thing
5519 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5520 before starting the TLS session.
5522 18. Tabs are now shown as \t in -bP output.
5524 19. If the log selector return_path_on_delivery was set, Exim crashed when
5525 bouncing a message because it had too many Received: header lines.
5527 20. If two routers both had headers_remove settings, and the first one included
5528 a superfluous trailing colon, the final name in the first list and the
5529 first name in the second list were incorrectly joined into one item (with a
5530 colon in the middle).
5536 1. Added -C and -D options to the exinext utility, mainly to make it easier
5537 to include in the automated testing, but these could be helpful when
5538 multiple configurations are in use.
5540 2. The exinext utility was not formatting the output nicely when there was
5541 an alternate port involved in the retry record key, nor when there was a
5542 message id as well (for retries that were specific to a specific message
5543 and a specific host). It was also confused by IPv6 addresses, because of
5544 the additional colons they contain. I have fixed the IPv4 problem, and
5545 patched it up to do a reasonable job for IPv6.
5547 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5548 delivery, the log line now contains "pipelined" if PIPELINING was used.
5550 4. An SMTP transport process used to panic and die if the bind() call to set
5551 an explicit outgoing interface failed. This has been changed; it is now
5552 treated in the same way as a connect() failure.
5554 5. A reference to $sender_host_name in the part of a conditional expansion
5555 that was being skipped was still causing a DNS lookup. This no longer
5558 6. The def: expansion condition was not recognizing references to header lines
5559 that used bh_ and bheader_.
5561 7. Added the _cache feature to named lists.
5563 8. The code for checking quota_filecount in the appendfile transport was
5564 allowing one more file than it should have been.
5566 9. For compatibility with Sendmail, the command line option
5574 and sets the incoming protocol and host name (for trusted callers). The
5575 host name and its colon can be omitted when only the protocol is to be set.
5576 Note the Exim already has two private options, -pd and -ps, that refer to
5577 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5578 "s", but I don't think that's a major issue.
5580 10. A number of refactoring changes to the code, none of which should affect
5583 (a) The number of logging options was getting close to filling up the
5584 32-bit word that was used as a bit map. I have split them into two classes:
5585 those that are passed in the argument to log_write(), and those that are
5586 only ever tested independently outside of that function. These are now in
5587 separate 32-bit words, so there is plenty of room for expansion again.
5588 There is no change in the user interface or the logging behaviour.
5590 (b) When building, for example, log lines, the code previously used a
5591 macro that called string_cat() twice, in order to add two strings. This is
5592 not really sufficiently general. Furthermore, there was one instance where
5593 it was actually wrong because one of the argument was used twice, and in
5594 one call a function was used. (As it happened, calling the function twice
5595 did not affect the overall behaviour.) The macro has been replaced by a
5596 function that can join an arbitrary number of extra strings onto a growing
5599 (c) The code for expansion conditions now uses a table and a binary chop
5600 instead of a serial search (which was left over from when there were very
5601 few conditions). Also, it now recognizes conditions like "pam" even when
5602 the relevant support is not compiled in: a suitably worded error message is
5603 given if an attempt is made to use such a condition.
5605 11. Added ${time_interval:xxxxx}.
5607 12. A bug was causing one of the ddress fields not to be passed back correctly
5608 from remote delivery subprocesses. The field in question was not being
5609 subsequently used, so this caused to problems in practice.
5611 13. Added new log selectors queue_time and deliver_time.
5613 14. Might have fixed a bug in maildirsizefile handling that threw up
5614 "unexpected character" debug warnings, and recalculated the data
5615 unnecessarily. In any case, I expanded the warning message to give more
5618 15. Added the message "Restricted characters in address" to the statements in
5619 the default ACL that block characters like @ and % in local parts.
5621 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5622 Three changes have been made:
5624 (a) There was a serious bug; a negative response to MAIL caused the whole
5625 recipient domain to be cached as invalid, thereby blocking all messages
5626 to all local parts at the same domain, from all senders. This bug has
5627 been fixed. The domain is no longer cached after a negative response to
5628 MAIL if the sender used is not empty.
5630 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5633 (c) A new callout option, "use_sender" has been added for people who want
5634 the modified behaviour.
5640 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5643 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5644 indeed breaks things for older releases.
5646 3. Added additional logging to the case where there is a problem reading data
5647 from a filter that is running in a subprocess using a pipe, in order to
5648 try to track down a specific problem.
5650 4. Testing facility fudge: when running in the test harness and attempting
5651 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5652 getting "No route to host". Convert this to a timeout.
5654 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5657 6. Some OS don't have socklen_t but use size_t instead. This affects the
5658 fifth argument of getsockopt() amongst other things. This is now
5659 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5660 can be set for individual OS. I have set it for SunOS5, OSF1, and
5661 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5662 some earlier ones do not.
5664 7. Change 4.30/15 was not doing the test caselessly.
5666 8. The standard form for an IPv6 address literal was being rejected by address
5667 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5668 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5669 this, as well as the form without the "IPv6" on the front (but only when
5670 address literals are enabled, of course).
5672 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5674 10. Exim crashed if a message with an empty sender address specified by -f
5675 encountered a router with an errors_to setting. This could be provoked only
5676 by a command such as
5680 where an empty string was supplied; "<>" did not hit this bug.
5682 11. Installed PCRE release 4.5.
5684 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5685 remained set. It is now erased.
5687 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5688 times from message ids (which are base 36 rather than the normal 62).
5690 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5691 were being counted as actual protocol errors, and logged if the log
5692 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5693 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5694 and DATA following a set of rejected RCPTs do not count as protocol errors.
5695 In other words, Exim assumes they were pipelined, though this may not
5696 actually be the case. Of course, in all cases the client gets an
5697 appropriate error code.
5699 15. If a lookup fails in an ACL condition, a message about the failure may
5700 be available; it is used if testing the ACL cannot continue, because most
5701 such messages specify what the cause of the deferral is. However, some
5702 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5703 that caused an old message to be retained and used if a later statement
5704 caused a defer, replacing the real cause of the deferral.
5706 16. If an IP address had so many PTR records that the DNS lookup buffer
5707 was not large enough to hold them, Exim could crash while trying to process
5708 the truncated data. It now detects and logs this case.
5710 17. Further to 4.21/58, another change has been made: if (and only if) the
5711 first line of a message (the first header line) ends with CRLF, a bare LF
5712 in a subsequent header line has a space inserted after it, so as not to
5713 terminate the header.
5715 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5716 unnecessarily, used atoi() instead of strtol(), and didn't check the
5717 termination when getting file sizes from file names by regex.
5719 19. Completely re-implemented the support for maildirsize files, in the light
5720 of a number of problems with the previous contributed implementation
5721 (4.30/29). In particular:
5723 . If the quota is zero, the maildirsize file is maintained, but no quota is
5726 . If the maildir directory does not exist, it is created before any attempt
5727 to write a maildirsize file.
5729 . The quota value in the file is just a cache; if the quota is changed in
5730 the transport, the new value overrides.
5732 . A regular expression is available for excluding directories from the
5735 20. The autoreply transport checks the characters in options that define the
5736 message's headers; it allows continued headers, but it was checking with
5737 isspace() after an embedded newline instead of explicitly looking for a
5740 21. If all the "regular" hosts to which an address was routed had passed their
5741 expiry times, and had not reached their retry times, the address was
5742 bounced, even if fallback hosts were defined. Now Exim should go on to try
5745 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5746 equivalent code in the SMTP transport. Some hosts send humungous responses
5747 to HELO/EHLO, more than 1024 it seems.
5749 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5750 gives compiler warnings in some environments. I've now done it "properly",
5753 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5754 (because of problems with the built-in one) was declared to return uschar *
5755 instead of char *, causing compiler failure.
5757 25. Fixed a file descriptor leak when processing alias/forward files.
5759 26. Fixed a minor format string issue in dbfn.c.
5761 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5763 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5764 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5767 29. When a sender address is verified, it is cached, to save repeating the test
5768 when there is more than one recipient in a message. However, when the
5769 verification involves a callout, it is possible for different callout
5770 options to be set for different recipients. It is too complicated to keep
5771 track of this in the cache, so now Exim always runs a verification when a
5772 callout is required, relying on the callout cache for the optimization.
5773 The overhead is duplication of the address routing, but this should not be
5776 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5777 to be verified with callout=postmaster, and the main callout worked but the
5778 postmaster check failed, the verification correctly failed. However, if a
5779 subsequent RCPT command asked for sender verification *without* the
5780 postmaster check, incorrect caching caused this verification also to fail,
5783 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5784 it was not caching the DNS options (qualify_single, search_parents) that
5785 were used when the lookup failed. A subsequent lookup with different
5786 options therefore always gave the same answer, though there were cases
5787 where it should not have. (Example: a "domains = !$mx_any" option on a
5788 dnslookup router: the "domains" option is always processed without any
5789 widening, but the router might have qualify_single set.) Now Exim uses the
5790 cached value only when the same options are set.
5792 32. Added John Jetmore's "exipick" utility to the distribution.
5794 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5795 than a timeout (e.g. a certificate is required, and is not provided), an
5796 Exim server now closes the connection immediately. Previously it waited for
5797 the client to close - but if the client is SSL, it seems that they each
5798 wait for each other, leading to a delay before one of them times out.
5800 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5801 maintained 0.8.x compatibility because I don't think many are using it, and
5802 it is clearly obsolete.
5804 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5807 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5808 client certificate was expired. A simple patch fixes this, though I don't
5809 understand the full logic of why the verify callback is called multiple
5812 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5813 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5814 which causes problems with some clients (such as the Certicom SSL Plus
5815 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5816 disables the coutermeasure allowing Eudora to connect."
5818 38. Exim was not checking that a write() to a log file succeeded. This could
5819 lead to Bad Things if a log got too big, in particular if it hit a file
5820 size limit. Exim now panics and dies if it cannot write to a log file, just
5821 as it does if it cannot open a log file.
5823 39. Modified OS/Makefile-Linux so that it now contains
5825 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5827 The two -D definitions ensure that Exim is compiled with large file
5828 support, which makes it possible to handle log files that are bigger than
5831 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5832 instance) a domain was checked against a named list that involved a lookup,
5833 causing $domain_data to be set, then another domain was checked against the
5834 same list, then the first domain was re-checked, the value of $domain_data
5835 after the final check could be wrong. In particular, if the second check
5836 failed, it could be set empty. This bug probably also applied to
5839 41. The strip_trailing_dot option was not being applied to the address given
5840 with the -f command-line option.
5842 42. The code for reading a message's header from the spool was incrementing
5843 $received_count, but never initializing it. This meant that the value was
5844 incorrect (doubled) while delivering a message in the same process in which
5845 it was received. In the most common configuration of Exim, this never
5846 happens - a fresh exec is done - but it can happen when
5847 deliver_drop_privilege is set.
5849 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5850 it now includes up to 150 characters of the unexpected data in the log
5853 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5854 and building data strings. The size of both of these buffers was 10 000
5855 bytes - far larger than anybody would *ever* want, thought I. Needless to
5856 say, somebody hit the limit. I have increased the maximum line length to
5857 20 000 and the maximum data length of concatenated lines to 100 000. I have
5858 also fixed two bugs, because there was no checking on these buffers. Tsk,
5859 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5860 buffer is too small.
5862 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5863 lsearch lookups. Now it does.
5865 46. When parsing a route_list item in a manualroute router, a fixed-length
5866 buffer was used for the list of hosts. I made this 1024 bytes long,
5867 thinking that nobody would ever have a list of hosts that long. Wrong.
5868 Somebody had a whole pile of complicated expansion conditions, and the
5869 string was silently truncated, leading to an expansion error. It turns out
5870 that it is easier to change to an unlimited length (owing to other changes
5871 that have happened since this code was originally written) than to build
5872 structure for giving a limitation error. The length of the item that
5873 expands into the list of hosts is now unlimited.
5875 47. The lsearch lookup could not handle data where the length of text line was
5876 more than 4095 characters. Such lines were truncated, leading to shortened
5877 data being returned. It should now handle lines of any length.
5879 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5880 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5883 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5884 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5885 no longer happens in comment lines. A list of replacements is now placed
5886 at the head of all of the source files, except those whose only change is
5887 to replace PERL_COMMAND in the very first #! line.
5889 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5890 messages on the queue, with a bottom-up merge sort, using code contributed
5891 by Michael Haardt. This should make operations like -bp somewhat faster on
5892 large queues. It won't affect queue runners, except when queue_run_in_order
5895 51. Installed eximstats 1.31 in the distribution.
5897 52. Added support for SRV lookups to the dnslookup router.
5899 53. If an ACL referred to $message_body or $message_body_end, the value was not
5900 reset for any messages that followed in the same SMTP session.
5902 54. The store-handling optimization for building very long strings was not
5903 differentiating between the different store pools. I don't think this
5904 actually made any difference in practice, but I've tidied it.
5906 55. While running the routers to verify a sender address, $sender_address
5907 was still set to the sender address. This is wrong, because when routing to
5908 send a bounce to the sender, it would be empty. Therefore, I have changed
5909 it so that, while verifying a sender address, $sender_address is set to <>.
5910 (There is no change to what happens when verifying a recipient address.)
5912 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5913 target A or AAAA records (if not already returned) without resetting the
5914 qualify_single or search_parents options of the DNS resolver. These are
5915 inappropriate in this case because the targets of MX and SRV records must
5916 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5917 target that, when qualified, matched something in the local domain. These
5918 two options are now turned off when doing these lookups.
5920 57. It seems that at least some releases of Reiserfs (which does not have the
5921 concept of a fixed number of inodes) returns zero and not -1 for the
5922 number of available inodes. This interacted badly with check_spool_inodes,
5923 which assumed that -1 was the "no such thing" setting. What I have done is
5924 to check that the total number of inodes is greater than zero before doing
5925 the test of how many are available.
5927 58. When a "warn" ACL statement has a log_message modifier, the message is
5928 remembered, and not repeated. This is to avoid a lot of repetition when a
5929 message has many recipients that cause the same warning to be written.
5930 However, Exim was preserving the list of already written lines for an
5931 entire SMTP session, which doesn't seem right. The memory is now reset if a
5932 new message is started.
5934 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5935 debugging output unless log_rewrite was also set.
5937 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5938 of (int)(handle) when we know that handle contains (void *)(-1).
5940 61. The Exim daemon panic-logs an error return when it closes the incoming
5941 connection. However "connection reset by peer" seems to be common, and
5942 isn't really an error worthy of noting specially, so that particular error
5945 62. When Exim is trying to find all the local interfaces, it used to panic and
5946 die if the ioctl to get the interface flags failed. However, it seems that
5947 on at least one OS (Solaris 9) it is possible to have an interface that is
5948 included in the list of interfaces, but for which you get a failure error
5949 for this call. This happens when the interface is not "plumbed" into a
5950 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5951 failure of the "get flags" call assumes that the interface is down.
5953 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5954 makes life easier for people who are doing arithmetic on fields extracted
5955 from dates, where you often get leading zeros that should not be
5956 interpreted as octal.
5958 64. Added qualify_domain to the redirect router, to override the global
5961 65. If a pathologically long header line contained very many addresses (the
5962 report of this problem mentioned 10 000) and each of them was rewritten,
5963 Exim could use up a very large amount of memory. (It kept on making new
5964 copies of the header line as it rewrote, and never released the old ones.)
5965 At the expense of a bit more processing, the header rewriting function has
5966 been changed so that it no longer eats memory in this way.
5968 66. The generation of the Received: header has been moved from the time that a
5969 message starts to be received, to the time that it finishes. The timestamp
5970 in the Received: header should now be very close to that of the <= log
5971 line. There are two side-effects of this change:
5973 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5974 logged header lines no longer include the local Received: line, because
5975 it has not yet been created. The same applies to a copy of the message
5976 that is returned to a non-SMTP sender when a message is rejected.
5978 (b) When a filter file is tested using -bf, no additional Received: header
5979 is added to the test message. After some thought, I decided that this
5982 This change does not affect the value of $received_for. It is still set
5983 after address rewriting, but before local_scan() is called.
5985 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5987 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5988 gave an unhelpful panic error message, and a defer error. I have managed to
5989 change this behaviour so that it now rejects any supplied certificate,
5990 which seems right, as the list of acceptable certificates is empty.
5992 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5993 gave an unhelpful defer error. I have not managed to make this reject any
5994 supplied certificates, but the error message it gives is "no certificate
5995 supplied", which is not helpful.
5997 70. exigrep's output now also includes lines that are not associated with any
5998 message, but which match the given pattern. Implemented by a patch from
5999 Martin Sluka, which also tidied up the Perl a bit.
6001 71. Recipient callout verification, like sender verification, was using <> in
6002 the MAIL FROM command. This isn't really the right thing, since the actual
6003 sender may affect whether the remote host accepts the recipient or not. I
6004 have changed it to use the actual sender in the callout; this means that
6005 the cache record is now keyed on a recipient/sender pair, not just the
6006 recipient address. There doesn't seem to be a real danger of callout loops,
6007 since a callout by the remote host to check the sender would use <>.
6008 [SEE ABOVE: changed after hitting problems.]
6010 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6011 temporary errors. However, in the case of such a code being given after
6012 the end of a data transmission (i.e. after ".") Exim was failing to write
6013 a retry record for the message. (Yes, there was some broken host that was
6014 actually sending 8xx at this point.)
6016 73. An unknown lookup type in a host list could cause Exim to panic-die when
6017 the list was checked. (An example that provoked this was putting <; in the
6018 middle of a list instead of at the start.) If this happened during a DATA
6019 ACL check, a -D file could be left lying around. This kind of configuration
6020 error no longer causes Exim to die; instead it causes a defer error. The
6021 incident is still logged to the main and panic logs.
6023 74. Buglet left over from Exim 3 conversion. The message "too many messages
6024 in one connection" was written to the rejectlog but not the mainlog, except
6025 when address rewriting (yes!) was being logged.
6027 75. Added write_rejectlog option.
6029 76. When a system filter was run not as root (that is, when system_filter_user
6030 was set), the values of the $n variables were not being returned to the
6031 main process; thus, they were not subsequently available in the $sn
6034 77. Added +return_path_on_delivery log selector.
6036 78. A connection timeout was being treated differently from recipients deferred
6037 when testing hosts_max_try with a message that was older than the host's
6038 retry timeout. (The host should not be counted, thus allowing all hosts to
6039 be tried at least once before bouncing.) This may have been the cause of an
6040 occasionally reported bug whereby a message would remain on the queue
6041 longer than the retry timeout, but would be bounced if a delivery was
6042 forced. I say "may" because I never totally pinned down the problem;
6043 setting up timeout/retry tests is difficult. See also the next item.
6045 79. The ultimate address timeout was not being applied to errors that involved
6046 a combination of host plus message (for example, a timeout on a MAIL
6047 command). When an address resolved to a number of possible hosts, and they
6048 were not all tried for each delivery (e.g. because of hosts_max_try), a
6049 message could remain on the queue longer than the retry timeout.
6051 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6054 81. Fixed an obscure SMTP outgoing bug which required at least the following
6055 conditions: (a) there was another message waiting for the same server;
6056 (b) the server returned 5xx to all RCPT commands in the first message so
6057 that the message was not completed; (c) the server dropped the connection
6058 or gave a negative response to the RSET that Exim sends to abort the
6059 transaction. The observed case was a dropped connection after DATA that had
6060 been sent in pipelining mode. That is, the server had advertised PIPELINING
6061 but was not implementing it correctly. The effect of the bug was incorrect
6062 behaviour, such as trying another host, and this could lead to a crash.
6068 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6069 and daemon.c were passed as pointers to ints; they should have been
6070 pointers to socklen_t variables (which are typically unsigned ints).
6072 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6075 3. Fixed a really odd bug that affected only the testing scheme; patching a
6076 certain fixed string in the binary changed the value of another string that
6077 happened to be identical to the end of the original first string.
6079 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6080 name", it returns that address as the IP address. On some operating
6081 systems (e.g. Solaris), it also passes back the IP address string as the
6082 "host name". However, on others (e.g. Linux), it passes back an empty
6083 string. Exim wasn't checking for this, and was changing the host name to an
6084 empty string, assuming it had been canonicalized.
6086 5. Although rare, it is permitted to have more than one PTR record for a given
6087 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6088 all the names associated with an address, because they do in Solaris.
6089 However, it seems that they do not in Linux for data that comes from the
6090 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6091 I found this out when I moved to a new Linux workstation and tried to run
6092 the Exim test suite.
6094 To get round this problem I have changed the code so that it now does its
6095 own call to the DNS to look up PTR records when searching for a host name.
6096 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6097 addresses that are only in /etc/hosts are still found.
6099 This behaviour is, however, controlled by an option called host_lookup_
6100 order, which defaults to "bydns:byaddr". If people want to use the other
6101 order, or indeed, just use one or the other means of lookup, they can
6102 specify it in this variable.
6104 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6105 some operating systems, this comes back from gethostbyaddr() as an empty
6106 string, and this is what Exim used to test for. However, it seems that in
6107 other systems, "." is yielded. Exim now tests for this case too.
6109 7. The values of check_spool_space and check_log_space are now held internally
6110 as a number of kilobytes instead of an absolute number of bytes. If a
6111 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6112 kilobyte. This means that much larger values can be stored.
6114 8. Exim monitor: an attempt to get the action menu when not actually pointing
6115 at a message produces an empty menu entitled "No message selected". This
6116 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6117 no entries in it ("Shell widget menu has zero width and/or height"). So I
6118 have added a single, blank menu entry in this case.
6120 9. Added ${quote_local_part.
6122 10. MIME decoding is now applied to the contents of Subject: header lines when
6125 11. Now that a reference to $sender_host_address automatically causes a reverse
6126 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6127 host lookup before query-style lookups in lists that might use this
6128 variable. This has therefore been abolished, and the "net-" prefix is no
6129 longer necessary for query-style lookups.
6131 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6132 have been a typo for LFLAGS, so it has been changed.
6134 13. The install script calls Exim with "-C /dev/null" in order to find the
6135 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6136 to be output. However, since Exim outputs its version number before the
6137 error, it didn't break the script. It just looked ugly. I fixed this by
6138 always allowing "-C /dev/null" if the caller is root.
6140 14. Ignore overlarge ACL variable number when reading spool file - insurance
6141 against a later release with more variables having written the file.
6143 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6144 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6145 this, as well as the form without the "IPv6" on the front.
6147 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6148 OS/Makefile-Darwin file.
6150 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6151 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6153 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6154 variable. However, some OS use ints for inodes. I've added cast to long int
6155 to get rid of the compiler warning.
6157 19. I had forgotten to lock out "/../" in configuration file names when
6158 ALT_CONFIG_PREFIX was set.
6160 20. Routers used for verification do not need to specify transports. However,
6161 if such a router generated a host list, and callout was configured, Exim
6162 crashed, because it could not find a port number from the (non-existent)
6163 transport. It now assumes port 25 in this circumstance.
6165 21. Added the -t option to exigrep.
6167 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6168 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6169 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6170 them is set, LOOKUP_LSEARCH is forced.
6172 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6174 24. Added sender and host information to the "rejected by local_scan()" log
6175 line; previously there was no indication of these.
6177 25. Added .include_if_exists.
6179 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6180 Linux. It turns out that not all file systems support this. Apparently some
6181 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6182 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6183 not supported on the file descriptor, is now ignored when Exim is trying to
6184 sync a directory. This applies only to Linux.
6186 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6188 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6189 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6192 29. Added support for maildirsize files from supplied patch (modified a bit).
6194 30. The use of :fail: followed by an empty string could lead Exim to respond to
6195 sender verification failures with (e.g.):
6197 550 Verification failed for <xxx>
6198 550 Sender verify failed
6200 where the first response line was missing the '-' that indicates it is not
6201 the final line of the response.
6203 31. The loop for finding the name of the user that called Exim had a hardwired
6204 limit of 10; it now uses the value of finduser_retries, which is used for
6205 all other user lookups.
6207 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6210 33. Exim was neglecting to zero errno before one call of strtol() when
6211 expanding a string and expecting an integer value. On some systems this
6212 resulted in spurious "integer overflow" errors. Also, it was casting the
6213 result into an int without checking.
6215 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6216 did not work. The code looks as if it has *never* worked, though it appears
6217 to have been documented since at least release 1.62. I have made it work.
6219 35. The "timeout_DNS" error in retry rules, also documented since at least
6220 1.62, also never worked. As it isn't clear exactly what this means, and
6221 clearly it isn't a major issue, I have abolished the feature by treating it
6222 as "timeout", and writing a warning to the main and panic logs.
6224 36. The display of retry rules for -brt wasn't always showing the error code
6227 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6228 timeout_connect_A, timeout_connect_MX.
6230 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6231 to the empty sender.
6233 39. The daemon was not analysing the content of -oX till after it had closed
6234 stderr and disconnected from the controlling terminal. This meant that any
6235 syntax errors were only noted on the panic log, and the return code from
6236 the command was 0. By re-arranging the code a little, I've made the
6237 decoding happen first, so such errors now appear on stderr, and the return
6238 code is 1. However, the actual setting up of the sockets still happens in
6239 the disconnected process, so errors there are still only recorded on the
6242 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6243 connections (as happens on some IP stacks) was logged at start up time as
6244 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6245 it from "IPv6 and IPv4", which means that two separate sockets are being
6248 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6249 says whether AF_INET or AF_INET6 was passed as an argument.
6251 42. Exiwhat output was messed up when time zones were included in log
6254 43. Exiwhat now gives more information about the daemon's listening ports,
6255 and whether -tls-on-connect was used.
6257 44. The "port" option of the smtp transport is now expanded.
6259 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6260 silently ignored. Now an error message is written to the main and panic
6263 46. There's a new ACL modifier called "logwrite" which writes to a log file
6264 as soon as it is encountered.
6266 47. Added $local_user_uid and $local_user_gid at routing time.
6268 48. Exim crashed when trying to verify a sender address that was being
6271 49. Exim was recognizing only a space character after ".include". It now also
6272 recognizes a tab character.
6274 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6275 extracting the relevant information from the specification. The man page no
6276 longer contains scrambled data for the -d option, and I've added a section
6277 at the front about calling Exim under different names.
6279 51. Added "extra_headers" argument to the "mail" command in filter files.
6281 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6284 53. Installed eximstats 1.29.
6286 54. Added transport_filter_timeout as a generic transport option.
6288 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6289 Cc: header lines. This was required by RFC 822, but it not required by RFC
6292 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6293 incoming messages that did not have them. Now it does so only if the
6294 message originates locally, that is, if there is no associated remote host
6295 address. When Resent- header lines are present, this applies to the Resent-
6296 lines rather than the non-Resent- lines.
6298 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6299 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6301 58. Messages for configuration errors now include the name of the main
6302 configuration file - useful now that there may be more than one file in a
6303 list (.included file names were always shown).
6305 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6306 for those rare installations that do not start the daemon as root or run it
6307 setuid root. I've cut out the call to initgroups() if the daemon is not
6310 60. The Exim user and group can now be bound into the binary as text strings
6311 that are looked up at the start of Exim's processing.
6313 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6315 62. Added $mailstore_basename variable.
6317 63. Installed patch to sieve.c from Michael Haardt.
6319 64. When Exim failed to open the panic log after failing to open the main log,
6320 the original message it was trying to log was written to stderr and debug
6321 output, but if they were not available (the usual case in production), it
6322 was lost. Now it is written to syslog before the two lines that record the
6323 failures to open the logs.
6325 65. Users' Exim filters run in subprocesses under the user's uid. It is
6326 possible for a "deliver" command or an alias in a "personal" command to
6327 provoke an address rewrite. If logging of address rewriting is configured,
6328 this fails because the process is not running as root or exim. There may be
6329 a better way of dealing with this, but for the moment (because 4.30 needs
6330 to be released), I have disabled address rewrite logging when running a
6331 filter in a non-root, non-exim process.
6337 1. The buildconfig auxiliary program wasn't quoting the value set for
6338 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6339 not defined. This bug was masked in 4.22 by the effect that was fixed in
6342 2. Some messages that were rejected after a message id was allocated were
6343 shown as "incomplete" by exigrep. It no longer does this for messages that
6344 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6346 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6347 have allow_domain_literals set, the ID did not get logged in the <= line.
6348 Domain literals are now always recognized in Message-ID: header lines.
6350 4. The first argument for a ${extract expansion item is the key name or field
6351 number. Leading and trailing spaces in this item were not being ignored,
6352 causing some misleading effects.
6354 5. When deliver_drop_privilege was set, single queue runner processes started
6355 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6356 same command in the process it spins off) were not dropping privilege.
6358 6. When the daemon running as "exim" started a queue runner, it always
6359 re-executed Exim in the spun-off process. This is a waste of effort when
6360 deliver_drop_privilege is set. The new process now just calls the
6361 queue-runner function directly.
6367 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6370 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6371 ignored. Though the use of -oP was forcing the writing of a pid file, it
6372 was always written to the default place.
6374 3. If the message "no IP address found for host xxxx" is generated during
6375 incoming verification, it is now followed by identification of the incoming
6376 connection (so you can more easily find what provoked it).
6378 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6380 5. Added some features to "harden" Exim a bit more against certain attacks:
6382 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6383 be put in Local/Makefile. This is like the never_users runtime option,
6384 but it cannot be overridden. The default setting is "root".
6386 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6387 prefix string with which any file named in a -C command line option
6390 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6391 is retained for -C and -D only if the caller of Exim is root. Without
6392 it, the exim user may also use -C and -D and retain privilege.
6394 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6395 command line option is disabled.
6397 6. Macro names set by the -D option must start with an upper case letter, just
6398 like macro names defined in the configuration file.
6400 7. Added "dereference=" facility to LDAP.
6402 8. Two instances of the typo "uknown" in the source files are fixed.
6404 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6405 the Configure-Makefile script screwed up while processing it.
6407 10. Incorporated PCRE 4.4.
6409 11. The SMTP synchronization check was not operating right at the start of an
6410 SMTP session. For example, it could not catch a HELO sent before the client
6411 waited for the greeting. There is now a check for outstanding input at the
6412 point when the greeting is written. Because of the duplex, asynchronous
6413 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6414 way, but not yet received, when the check is performed.
6416 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6417 on TCP/IP sockets, because this apparently causes some broken clients to
6420 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6421 unchanged) from the Cygwin maintainer.
6423 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6424 was supported instead of testing for mbx. Effectively a typo.
6426 15. The spa authenticator server code was not checking that the input it
6427 received was valid base64.
6429 16. The debug output line for the "set" modifier in ACLs was not showing the
6430 name of the variable that was being set.
6432 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6434 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6435 lookup on its own. Something else was needed to trigger the lookup. For
6436 example, a match in host_lookup or the need for a host name in a host list.
6437 Now, if $sender_host_name is referenced and the host name has not yet been
6438 looked up, a lookup is performed. If the lookup fails, the variable remains
6439 empty, and $host_lookup_failed is set to "1".
6441 19. Added "eqi" as a case-independent comparison operator.
6443 20. The saslauthd authentication condition could segfault if neither service
6444 nor realm was specified.
6446 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6447 error message that was logged was misleading, and incoming SMTP
6448 connections were dropped. The message is now more accurate, and temporary
6449 errors are given to SMTP connections.
6451 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6452 (except RSET) with a 421 error, until QUIT is received. However, it was
6453 failing to send a response to QUIT.
6455 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6456 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6457 ACL is now run after the helo_try_verify_hosts code.
6459 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6460 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6461 case-independent, so other case variants are also recognized). Apparently
6462 some systems use these upper case variants.
6464 25. If more than two messages were waiting for the same host, and a transport
6465 filter was specified for the transport, Exim sent two messages over the
6466 same TCP/IP connection, and then failed with "socket operation on non-
6467 socket" when it tried to send the third.
6469 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6471 27. The extern definition of crypt16() in expand.c was not being excluded when
6472 the OS had its own crypt16() function.
6474 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6475 as a preferred synonym for return_size_limit, both as an option and as an
6478 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6480 30. Changed the default configuration ACL to relax the local part checking rule
6481 for addresses that are not in any local domains. For these addresses,
6482 slashes and pipe symbols are allowed within local parts, but the sequence
6483 /../ is explicitly forbidden.
6485 31. SPA server authentication was not clearing the challenge buffer before
6488 32. log_message in a "warn" ACL statement was writing to the reject log as
6489 well as to the main log, which contradicts the documentation and doesn't
6490 seem right (because no rejection is happening). So I have stopped it.
6492 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6493 However, I am unable to do any testing of this.
6495 34. Fixed an infelicity in the appendfile transport. When checking directories
6496 for a mailbox, to see if any needed to be created, it was accidentally
6497 using path names with one or more superfluous leading slashes; tracing
6498 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6500 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6501 contents are added to the log line that is written for every discarded
6502 recipient. (Previously a log_message setting was ignored.)
6504 36. The ${quote: operator now quotes the string if it is empty.
6506 37. The install script runs exim in order to find its version number. If for
6507 some reason other than non-existence or emptiness, which it checks, it
6508 could not run './exim', it was installing it with an empty version number,
6509 i.e. as "exim-". This error state is now caught, and the installation is
6512 38. An argument was missing from the function that creates an error message
6513 when Exim fails to connect to the socket for saslauthd authentication.
6514 This could cause Exim to crash, or give a corrupted message.
6516 39. Added isip, isip4, and isip6 to ${if conditions.
6518 40. The ACL variables $acl_xx are now saved with the message, and can be
6519 accessed later in routers, transports, and filters.
6521 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6522 strings in the file are not string-expanded.
6524 42. If a MAIL command specified a SIZE value that was too large to fit into an
6525 int variable, the check against message_size_limit failed. Such values are
6526 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6527 day this will have to be increased, but I don't think I want to be around
6528 when emails are that large.
6535 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6536 iconv() is not standard in FreeBSD.
6538 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6539 IPv6 enabled. The observed symptom was a segmentation fault on return from
6540 the function os_common_find_running_interfaces() in src/os.c.
6542 3. In the check_special_case() function in daemon.c I had used "errno" as an
6543 argument name, which causes warnings on some systems. This was basically a
6544 typo, since it was named "eno" in the comments!
6546 4. The code that waits for the clock to tick (at a resolution of some fraction
6547 of a second) so as to ensure message-id uniqueness was always waiting for
6548 at least one whole tick, when it could have waited for less. [This is
6549 almost certainly not relevant at current processor speeds, where it is
6550 unlikely to ever wait at all. But we try to future-proof.]
6552 5. The function that sleeps for a time interval that includes fractions of a
6553 second contained a race. It did not block SIGALRM between setting the
6554 timer, and suspending (a couple of lines later). If the interval was short
6555 and the sigsuspend() was delayed until after it had expired, the suspension
6556 never ended. On busy systems this could lead to processes getting stuck for
6559 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6560 process, before it forks any delivery processes. The open lookup caching
6561 mechanism meant that the open file or database connection was passed into
6562 the delivery process. The problem was that delivery processes always tidy
6563 up cached lookup data. This could cause a problem for the next delivery
6564 process started by the queue runner, because the external queue runner
6565 process does not know about the closure. So the next delivery process
6566 still has data in the lookup cache. In the case of a file lookup, there was
6567 no problem because closing a file descriptor in a subprocess doesn't affect
6568 the parent. However, if the lookup was caching a connection to a database,
6569 the connection was closed, and the second delivery process was likely to
6570 see errors such as "PGSQL: query failed: server closed the connection
6571 unexpectedly". The problem has been fixed by closing all cached lookups
6572 in a queue runner before running a delivery process.
6574 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6575 seem to have the "const" qualifier which it has on other OS. I've
6578 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6579 *of the same type* (client or server) with the same public name that an
6580 error should be diagnosed.
6582 9. When Exim looked up a host name for an IP address, but failed to find the
6583 original IP address when looking up the host name (a safety check), it
6584 output the message "<ip address> does not match any IP for NULL", which was
6585 confusing, to say the least. The bug was that the host name should have
6586 appeared instead of "NULL".
6588 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6589 user that is built into the binary, and the -C or -D options is used, root
6590 privilege is dropped before the configuration file is read. In addition,
6591 logging is switched to stderr instead of the normal log files. If the
6592 configuration then re-defines the Exim user, the unprivileged environment
6593 is probably not what is expected, so Exim logs a panic warning message (but
6596 However, if deliver_drop_privilege is set, the unprivileged state may well
6597 be exactly what is intended, so the warning has been cut out in that case,
6598 and Exim is allowed to try to write to its normal log files.
6604 1. smtp_return_error_details was not giving details for temporary sender
6605 or receiver verification errors.
6607 2. Diagnose a configuration error if two authenticators have the same public
6610 3. Exim used not to create the message log file for a message until the first
6611 delivery attempt. This could be confusing when incoming messages were held
6612 for policy or load reasons. The message log file is now created at the time
6613 the message is received, and an initial "Received" line is written to it.
6615 4. The automatically generated man page for command line options had a minor
6616 bug that caused no ill effects; however, a more serious problem was that
6617 the procedure for building the man page automatically didn't always
6618 operate. Consequently, release 4.20 contains an out-of-date version. This
6619 shouldn't happen again.
6621 5. When building Exim with embedded Perl support, the script that builds the
6622 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6623 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6625 6. The freeze_tell option was not being used for messages that were frozen on
6626 arrival, either by an ACL or by local_scan().
6628 7. Added the smtp_incomplete_transaction log selector.
6630 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6631 was accepting AUTH without a new EHLO.
6633 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6634 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6637 10. Logging of TCP/IP connections (when configured) now happens in the main
6638 daemon process instead of the child process, so that the TCP/IP connection
6639 count is more accurate (but it can never be perfect).
6641 11. The use of "drop" in a nested ACL was not being handled correctly in the
6642 outer ACL. Now, if condition failure induced by the nested "drop" causes
6643 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6644 or "require"), the connection is dropped.
6646 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6647 that yield "discard" can now be used with an "accept" or a "discard" verb,
6648 but an error is generated for any others (because I can't see a useful way
6649 to define what should happen).
6651 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6652 are now processed in the same way as lines in the Exim configuration file.
6653 In particular, continuation lines are supported.
6655 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6657 15. Added -ti meaning -t -i.
6659 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6660 domains, and warn by logging if others are found.
6662 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6663 returned from GIFCONF when Exim is trying to find the list of interfaces on
6664 a host. The code in os.c has been modified to copy each ifreq to an aligned
6665 structure in all cases.
6667 Also, in some cases, the returned ifreq's were being copied to a 'struct
6668 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6669 means the last couple of bytes of an IPv6 address could be chopped if the
6670 ifreq contained only a normal sockaddr (14 bytes storage).
6672 18. Named domain lists were not supported in the hosts_treat_as_local option.
6673 An entry such as +xxxx was not recognized, and was treated as a literal
6676 19. Ensure that header lines added by a DATA ACL are included in the reject log
6677 if the ACL subsequently rejects the message.
6679 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6680 MD5 (which is deprecated).
6682 21. When testing a filter file using -bf, Exim was writing a message when it
6683 took the sender from a "From " line in the message, but it was not doing so
6684 when it took $return_path from a Return-Path: header line. It now does.
6686 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6687 with a valid header line field name (a series of printing characters
6688 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6690 23. Changed "disc" in the source to "disk" to conform to the documentation and
6691 the book and for uniformity.
6693 24. Ignore Sendmail's -Ooption=value command line item.
6695 25. When execve() failed while trying to run a command in a pipe transport,
6696 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6697 could be confused with a return value of 69 from the command itself. This
6698 has been changed to 127, the value the shell returns if it is asked to run
6699 a non-existent command. The wording for the related log line suggests a
6700 non-existent command as the problem.
6702 26. If received_header_text expands to an empty string, do not add a Received:
6703 header line to the message. (Well, it adds a token one on the spool, but
6704 marks it "old" so that it doesn't get used or transmitted.)
6706 27. Installed eximstats 1.28 (addition of -nt option).
6708 28. There was no check for failure on the call to getsockname() in the daemon
6709 code. This can fail if there is a shortage of resources on the system, with
6710 ENOMEM, for example. A temporary error is now given on failure.
6712 29. Contrary to the C standard, it seems that in some environments, the
6713 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6714 program. Exim now does this explicitly; it affects the formatting of
6715 timestamps using strftime().
6717 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6718 complaints. I've now initialized all the variables, to avoid this.
6720 32. Header lines added by a system filter were not being "seen" during
6721 transport-time rewrites.
6723 33. The info_callback() function passed to OpenSSL is set up with type void
6724 (*)(SSL *, int, int), as described somewhere. However, when calling the
6725 function (actually a macro) that sets it up, the type void(*)() is
6726 expected. I've put in a cast to prevent warnings from picky compilers.
6728 34. If a DNS black list lookup found a CNAME record, but there were no A
6729 records associated with the domain it pointed at, Exim crashed.
6731 35. If a DNS black list lookup returned more than one A record, Exim ignored
6732 all but the first. It now scans all returned addresses if a particular IP
6733 value is being sought. In this situation, the contents of the
6734 $dnslist_value variable are a list of all the addresses, separated by a
6737 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6738 to accept a host name and all its aliases if the forward lookup for any of
6739 them yielded the IP address of the incoming connection. Now it accepts only
6740 those names whose forward lookup yields the correct IP address. Any other
6741 names are discarded. This closes a loophole whereby a rogue DNS
6742 administrator could create reverse DNS records to break through a
6743 wildcarded host restriction in an ACL.
6745 37. If a user filter or a system filter that ran in a subprocess used any of
6746 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6747 the wrong values were passed to the pipe command ($thisaddress had the
6748 value of $0, $0 had the value of $1, etc). This bug was introduced by
6749 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6751 38. Improved the line breaking for long SMTP error messages from ACLs.
6752 Previously, if there was no break point between 40 and 75 characters, Exim
6753 left the rest of the message alone. Two changes have been made: (a) I've
6754 reduced the minimum length to 35 characters; (b) if it can't find a break
6755 point between 35 and 75 characters, it looks ahead and uses the first one
6756 that it finds. This may give the occasional overlong line, but at least the
6757 remaining text gets split now.
6759 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6760 file descriptors might be low, and that setting 1000 would always raise it.
6761 It turns out that in some environments, the limit is already over 1000 and
6762 that lowering it causes trouble. So now Exim takes care not to decrease it.
6764 40. When delivering a message, the value of $return_path is set to $sender_
6765 address at the start of routing (routers may change the value). By an
6766 oversight, this default was not being set up when an address was tested by
6767 -bt or -bv, which affected the outcome if any router or filter referred to
6770 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6771 log only when "message" or "log_message" are set. However, if one of the
6772 conditions was an address verification, or a call to a nested ACL, the
6773 messages generated by the underlying test were being passed through. This
6774 no longer happens. The underlying message is available in $acl_verify_
6775 message for both "message" and "log_message" expansions, so it can be
6776 passed through if needed.
6778 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6779 new expansion $bh_ to give the encoded byte string without charset
6780 translation. Translation happens only if iconv() is available; HAVE_ICONV
6781 indicates this at build time. HEADERS_CHARSET gives the charset to
6782 translate to; headers_charset can change it in the configuration, and
6783 "headers charset" can change it in an individual filter file.
6785 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6786 that creates RFC 2047 encoded "words" labels them as that charset instead
6787 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6788 expansion operator; (ii) when Exim creates a From: line for a local
6789 message; (iii) when a header line is rewritten to include a "phrase" part.
6791 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6792 buggy, causing it to skip the first lines of messages whose message ID
6793 ended in 'D'. This would not have bitten before Exim release 4.14, because
6794 message IDs were unlikely to end in 'D' before then. The effect was to have
6795 incorrect size information for certain domains.
6797 45. #include "config.h" was missing at the start of the crypt16.c module. This
6798 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6801 46. If there was a timeout during a "random" callout check, Exim treated it as
6802 a failure of the random address, and carried on sending RSET and the real
6803 address. If the delay was just some slowness somewhere, the response to the
6804 original RCPT would be taken as a response to RSET and so on, causing
6805 mayhem of various kinds.
6807 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6808 when I implemented it. It didn't allow for the fact that some option values
6809 may legitimately be negative (e.g. size_addition), and it didn't even do
6810 the right test for positive values.
6812 48. Domain names in DNS records are case-independent. Exim always looks them up
6813 in lower case. Some resolvers return domain names in exactly the case they
6814 appear in the zone file, that is, they may contain uppercase letters. Not
6815 all resolvers do this - some return always lower case. Exim was treating a
6816 change of case by a resolver as a change of domain, similar to a widening
6817 of a domain abbreviation. This triggered its re-routing code and so it was
6818 trying to route what was effectively the same domain again. This normally
6819 caused routing to fail (because the router wouldn't handle the domain
6820 twice). Now Exim checks for this case specially, and just changes the
6821 casing of the domain that it ultimately uses when it transmits the message
6824 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6827 50. If a filter generated a file delivery with a non-absolute name (possible if
6828 no home directory exists for the router), the forbid_file option was not
6831 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6832 the existing equality matching.
6834 52. Exim was using ints instead of ino_t variables in some places where it was
6835 dealing with inode numbers.
6837 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6838 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6839 called TMPDIR, and if it finds it is different, it changes its value.
6841 54. The smtp_printf() function is now made available to local_scan() so
6842 additional output lines can be written before returning. There is also an
6843 smtp_fflush() function to enable the detection of a dropped connection.
6844 The variables smtp_input and smtp_batched_input are exported to
6847 55. Changed the default runtime configuration: the message "Unknown user"
6848 has been removed from the ACL, and instead placed on the localuser router,
6849 using the cannot_route_message feature. This means that any verification
6850 failures that generate their own messages won't get overridden. Similarly,
6851 the "Unrouteable address" message that was in the ACL for unverifiable
6852 relay addresses has also been removed.
6854 56. Added hosts_avoid_esmtp to the smtp transport.
6856 57. The exicyclog script was not checking for the esoteric option
6857 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6858 will work only if exicyclog is run under the appropriate euid.
6860 58. Following a discussion on the list, the rules by which Exim recognises line
6861 endings on incoming messages have been changed. The -dropcr and drop_cr
6862 options are now no-ops, retained only for backwards compatibility. The
6863 following line terminators are recognized: LF CRLF CR. However, special
6864 processing applies to CR:
6866 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6867 nor a local message in the state where . is a terminator.
6869 (ii) If a bare CR is encountered in a header line, an extra space is added
6870 after the line terminator so as not to end the header. The reasoning
6871 behind this is that bare CRs in header lines are most likely either
6872 to be mistakes, or people trying to play silly games.
6874 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6875 was being incorrectly given as 18 bytes larger than it should have been.
6876 This is a VOB (very old bug).
6878 60. This may never have affected anything current, but just in case it has:
6879 When the local host is found other than at the start of a list of hosts,
6880 the local host, those with the same MX, and any that follow, are discarded.
6881 When the list in question was part of a longer list of hosts, the following
6882 hosts (not currently being processed) were also being discarded. This no
6883 longer happens. I'm not sure if this situation could ever has previously
6886 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6889 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6890 header. This is something that is recommended in a new Internet Draft, and
6891 is something that is documented as being done by Sendmail. There are two
6892 possible values. For messages generated by the autoreply transport, Exim
6895 Auto-Submitted: auto-replied
6897 whereas for all other generated messages (e.g. bounces) it adds
6899 Auto-Submitted: auto-generated
6901 63. The "personal" condition in filters now includes a test for the
6902 Auto-Submitted: header. If it contains the string "auto-" the message it
6903 not considered personal.
6905 64. Added rcpt_include_affixes as a generic transport option.
6907 65. Added queue_only_override (default true).
6909 66. Added the syslog_duplication option.
6911 67. If what should have been the first header line of a message consisted of
6912 a space followed by a colon, Exim was mis-interpreting it as a header line.
6913 It isn't of course - it is syntactically invalid and should therefore be
6914 treated as the start of the message body. The misbehaviour could have
6915 caused a number of strange effects, including loss of data in subsequent
6916 header lines, and spool format errors.
6918 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6919 client host had authenticated. This control can now be exercised by an ACL
6920 for more flexibility.
6922 69. By default, callouts do not happen when testing with -bh. There is now a
6923 variant, -bhc, which does actually run the callout code, including
6924 consulting and updating the callout cache.
6926 70. Added support for saslauthd authentication, courtesy of Alexander
6929 71. If statvfs() failed on the spool or log directories while checking their
6930 size for availability, Exim confusingly gave the error "space shortage".
6931 Furthermore, in debugging mode it crashed with a floating point exception.
6932 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6933 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6934 serious problem, Exim now writes to the main and panic logs when this
6935 happens, with details of the failure. It then refuses to accept the
6936 incoming message, giving the message "spool directory problem" or "log
6937 directory problem" with a 421 code for SMTP messages.
6939 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6940 0, 1, and 2 exist, because some OS complain for execs without them (see
6941 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6942 descriptors. However, the code omitted to check that the open succeeded,
6943 causing mysterious errors if for some reason the permissions on /dev/null
6944 got screwed. Now Exim writes a message to the main and panic logs, and
6945 bombs out if it can't open /dev/null.
6947 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6948 interact so that it is all more flexible. It is supposed to remain
6949 backwards compatible. Also added extra_local_interfaces.
6951 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6952 to bomb out with an assertion failure - to the client this appears as a
6953 connection drop. This problem occurs in the part of the code that was taken
6954 from the Samba project. Fortunately, the assertion is in a very simple
6955 function, so I have fixed this by reproducing the function inline in the
6956 one place where it is called, and arranging for authentication to fail
6957 instead of killing the process with assert().
6959 75. The SPA client code was not working when the server requested OEM rather
6960 than Unicode encoding.
6962 76. Added code to make require_files with a specific uid setting more usable in
6963 the case where statting the file as root fails - usually a non-root-mounted
6964 NFS file system. When this happens and the failure is EACCES, Exim now
6965 forks a subprocess and does the per-uid checking as the relevant uid.
6967 77. Added process_log_path.
6969 78. If log_file_path was not explicitly set, a setting of check_log_space or
6970 check_log_inodes was ignored.
6972 79. If a space check for the spool or log partitions fails, the incident is now
6973 logged. Of course, in the latter case the data may get lost...
6975 80. Added the %p formatting code to string_format() so that it can be used to
6976 print addresses in debug_print(). Adjusted all the address printing in the
6977 debugging in store.c to use %p rather than %d.
6979 81. There was a concern that a line of code in smtp_in.c could overflow a
6980 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6981 initially expressed, the concern was not well-founded, because trailing
6982 spaces are removed early. However, if the trailing spaces were followed by
6983 a NULL, they did not get removed, so the overflow was possible. Two fixes
6986 (a) I re-wrote the offending code in a cleaner fashion.
6987 (b) If an incoming SMTP command contains a NULL character, it is rejected
6990 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6991 runs initgroups(), so that if the Exim user is in any additional groups,
6992 they will be used during message reception.
6998 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob