1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
8 Since Exim version 4.90
11 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
12 Previously only the last row was returned.
14 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
15 we assumed that tags in the header were well-formed, and parsed the
16 element content after inspecting only the first char of the tag.
17 Assumptions at that stage could crash the receive process on malformed
20 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
21 While running the DKIM ACL we operate on the Permanent memory pool so that
22 variables created with "set" persist to the DATA ACL. Also (at any time)
23 DNS lookups that fail create cache records using the Permanent pool. But
24 expansions release any allocations made on the current pool - so a dnsdb
25 lookup expansion done in the DKIM ACL releases the memory used for the
26 DNS negative-cache, and bad things result. Solution is to switch to the
27 Main pool for expansions.
28 While we're in that code, add checks on the DNS cache during store_reset,
29 active in the testsuite.
30 Problem spotted, and debugging aided, by Wolfgang Breyha.
32 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
33 When none of the hosts presented to a transport match an already-open
34 connection, close it and proceed with the list. Previously we would
35 queue the message. Spotted by Lena with Yahoo, probably involving
38 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
39 Previously a spurious "250 OK id=" response was appended to the proper
42 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
43 rows affected is given instead).
45 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
46 SMTP connection. Previously, when one had more receipients than the
47 first, an abortive onward connection was made. Move to full support for
48 multiple onward connections in sequence, handling cutthrough connection
49 for all multi-message initiating connections.
51 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
52 routers. Previously, a multi-recipient message would fail to match the
53 onward-connection opened for the first recipient, and cause its closure.
59 JH/01 Rework error string handling in TLS interface so that the caller in
60 more cases is responsible for logging. This permits library-sourced
61 string to be attached to addresses during delivery, and collapses
62 pairs of long lines into single ones.
64 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
65 during configuration. Wildcards are allowed and expanded.
67 JH/02 Rework error string handling in DKIM to pass more info back to callers.
68 This permits better logging.
70 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
71 do not close it down and have the child transport start it up again on
72 the passed-on TCP connection. Instead, proxy the child (and any
73 subsequent ones) for TLS via a unix-domain socket channel. Logging is
74 affected: the continued delivery log lines do not have any DNSSEC, TLS
75 Certificate or OCSP information. TLS cipher information is still logged.
77 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
78 identical IP addresses on different listening ports. Will also affect
81 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
82 add noisy ifdef guards to special-case this sillyness.
83 Patch from Bernd Kuhls.
85 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
86 than 255 are no longer allowed.
88 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
89 Disable the session-cache too, which might reduce our load. Since we
90 currrectly use a new context for every connection, both as server and
91 client, there is no benefit for these.
92 GnuTLS appears to not support tickets server-side by default (we don't
93 call gnutls_session_ticket_enable_server()) but client side is enabled
94 by default on recent versions (3.1.3 +) unless the PFS priority string
97 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
98 <https://reproducible-builds.org/specs/source-date-epoch/>.
100 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
101 the check for any unsuccessful recipients did not notice the limit, and
102 erroneously found still-pending ones.
104 JH/08 Pipeline CHUNKING command and data together, on kernels that support
105 MSG_MORE. Only in-clear (not on TLS connections).
107 JH/09 Avoid using a temporary file during transport using dkim. Unless a
108 transport-filter is involved we can buffer the headers in memory for
109 creating the signature, and read the spool data file once for the
110 signature and again for transmission.
112 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
113 4.77 as the kernel support then wasn't solid, having issues in 64bit
114 mode. Now, it's been long enough. Add support for FreeBSD also.
116 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
117 case where the routing stage had gathered several addresses to send to
118 a host before calling the transport for the first, we previously failed
119 to close down TLS in the old transport process before passing the TCP
120 connection to the new process. The new one sent a STARTTLS command
121 which naturally failed, giving a failed delivery and bloating the retry
122 database. Investigation and fix prototype from Wolfgang Breyha.
124 JH/12 Fix check on SMTP command input synchronisation. Previously there were
125 false-negatives in the check that the sender had not preempted a response
126 or prompt from Exim (running as a server), due to that code's lack of
127 awareness of the SMTP input buffering.
129 PP/04 Add commandline_checks_require_admin option.
130 Exim drops privileges sanely, various checks such as -be aren't a
131 security problem, as long as you trust local users with access to their
132 own account. When invoked by services which pass untrusted data to
133 Exim, this might be an issue. Set this option in main configuration
134 AND make fixes to the calling application, such as using `--` to stop
137 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
138 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
139 responses to those, into a single TLS record each way (this usually means
140 a single packet). As a side issue, smtp_enforce_sync now works on TLS
143 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
144 affects you only if you're dancing at the edge of the param size limits.
145 If you are, and this message makes sense to you, then: raise the
146 configured limit or use OpenSSL 1.1. Nothing we can do for older
149 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
150 cmdline element to get the documented default one. Previously it was
153 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
154 are unsuable for non-message-specific reasons - eg. connection timeout,
155 banner-time rejection.
157 JH/16 Fix logging of delivery remote port, when specified by router, under
160 PP/06 Repair manualroute's ability to take options in any order, even if one
161 is the name of a transport.
164 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
166 JH/17 Change the list-building routines interface to use the expanding-string
167 triplet model, for better allocation and copying behaviour.
169 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
170 Previously it was constructed the first time a possibly-matching string
171 was met in the configuration file input during startup; now it is done
174 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
175 compatible one, to avoid the (poorly documented) possibility of a config
176 file in the working directory redirecting the DB files, possibly correpting
177 some existing file. CVE-2017-10140 assigned for BDB.
179 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
180 cache-hot. Previously, although the result was properly cached, the
181 initial verify call returned a defer.
183 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
184 the main verify for receipient in uncached-mode.
186 JH/22 Retire historical build files to an "unsupported" subdir. These are
187 defined as "ones for which we have no current evidence of testing".
189 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
190 if present. Previously it was ignored.
192 JH/24 Start using specified-initialisers in C structure init coding. This is
193 a C99 feature (it's 2017, so now considered safe).
195 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
196 if was a fixed-sized field and bitmask ops via macros; it is now more
199 PP/07 GitHub PR 56: Apply MariaDB build fix.
200 Patch provided by Jaroslav Škarvada.
202 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
203 during Coverity cleanups [4.87 JH/47]
204 Diagnosis and fix provided by Michael Fischer v. Mollard.
206 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
207 the right size to place the terminating semicolon on its own folded
208 line, the header hash was calculated to an incorrect value thanks to
209 the (relaxed) space the fold became.
211 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
212 and confused the parent.
214 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
215 which could crash as a result. This could lead to undeliverable messages.
217 JH/28 Logging: "next input sent too soon" now shows where input was truncated
220 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
221 matters on fast-turnover and PID-randomising systems, which were getting
222 out-of-order delivery.
224 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
225 a possibly-overlapping copy. The symptom was that "Remote host closed
226 connection in response to HELO" was logged instead of the actual 4xx
229 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
230 Previously only that bufferd was discarded, resulting in SYMTP command
233 JH/32 DKIM: when a message has multiple signatures matching an identity given
234 in dkim_verify_signers, run the dkim acl once for each. Previously only
235 one run was done. Bug 2189.
237 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
238 "panic the current process" to "deliberately defer". The panic log is
239 still written with the problem list name; the mail and reject logs now
240 get a temp-reject line for the message that was being handled, saying
241 something like "domains check lookup or other defer". The SMTP 451
242 message is still "Temporary local problem".
244 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
245 A crafted sequence of BDAT commands could result in in-use memory beeing
246 freed. CVE-2017-16943.
248 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
249 from SMTP input. Previously it was always done; now only done for DATA
250 and not BDAT commands. CVE-2017-16944.
252 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
253 to the message (such as an overlong header line). Previously this was
254 not done and we did not exit BDAT mode. Followon from the previous item
255 though a different problem.
261 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
262 than -2003 did; needs libidn2 in addition to libidn.
264 JH/02 The path option on a pipe transport is now expanded before use.
266 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
267 Patch provided by "Björn", documentation fix added too.
269 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
270 missing a wire-to-host endian conversion.
272 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
273 close after a BDAT command line could be taken as a following command,
274 giving a synch failure. Fix by only checking for synch immediately
275 before acknowledging the chunk.
277 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
278 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
279 macro. Patches provided by Josh Soref.
281 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
282 Previously we did not; the RFC seems ambiguous and VRFY is not listed
283 by IANA as a service extension. However, John Klensin suggests that we
286 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
287 the dkim code may be unix-mode line endings rather than smtp wire-format
288 CRLF, so prepend a CR to any bare LF.
290 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
291 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
293 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
294 the first were themselves being wrongly included in the feed into dkim
295 processing; with most chunk sizes in use this resulted in an incorrect
296 body hash calculated value.
298 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
299 DKIM signature block, for verification. Although advised against by
300 standards it is specifically not ruled illegal.
302 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
304 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
305 missing a body hash (the bh= tag).
307 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
308 It seems that HAProxy sends the Proxy Protocol information in clear and
309 only then does a TLS startup, so do the same.
311 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
312 TCP connections (such as for Spamd) unless the daemon successfully set
313 Fast Open mode on its listening sockets. This fixes breakage seen on
314 too-old kernels or those not configured for Fast Open, at the cost of
315 requiring both directions being enabled for TFO, and TFO never being used
316 by non-daemon-related Exim processes.
318 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
319 endings, at least on the first header line. Try to canonify any that get
320 past that check, despite the cost.
322 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
323 now limited to an arbitrary five deep, while parsing addresses with the
324 strip_excess_angle_brackets option enabled.
326 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
327 instead leave the unprompted TLS handshake in socket buffer for the
328 TLS library to consume.
330 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
332 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
334 JH/16 Drop variables when they go out of scope. Memory management drops a whole
335 region in one operation, for speed, and this leaves assigned pointers
336 dangling. Add checks run only under the testsuite which checks all
337 variables at a store-reset and panics on a dangling pointer; add code
338 explicitly nulling out all the variables discovered. Fixes one known
339 bug: a transport crash, where a dangling pointer for $sending_ip_address
340 originally assigned in a verify callout, is re-used.
342 PP/06 Drop '.' from @INC in various Perl scripts.
344 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
346 PP/08 Reduce a number of compilation warnings under clang; building with
347 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
348 should be warning-free.
350 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
352 HS/01 Fix portability problems introduced by PP/08 for platforms where
353 realloc(NULL) is not equivalent to malloc() [SunOS et al].
355 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
356 chunk. This allows us to accept broken chunked messages. We need a more
357 general solution here.
359 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
360 already-broken messages in the queue.
362 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
364 JH/19 Fix reference counting bug in routing-generated-address tracking.
370 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
371 supports it and a size is available (ie. the sending peer gave us one).
373 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
374 years of being deprecated). The replacements are the ACLs
375 acl_smtp_mime and acl_not_smtp_mime.
377 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
378 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
379 or even an in-clear connection were permitted. Now, if the host lookup
380 was dnssec and dane was requested then the host is only used if the
381 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
382 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
383 if one fails this test.
384 This means that a poorly-configured remote DNS will make it incommunicado;
385 but it protects against a DNS-interception attack on it.
387 JH/04 Bug 1810: make continued-use of an open smtp transport connection
388 non-noisy when a race steals the message being considered.
390 JH/05 If main configuration option tls_certificate is unset, generate a
391 self-signed certificate for inbound TLS connections.
393 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
394 in rewrites and routers.
396 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
397 and logged a warning sing 4.83; now they are a configuration file error.
399 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
400 (lacking @domain). Apply the same qualification processing as RCPT.
402 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
404 JH/10 Support ${sha256:} applied to a string (as well as the previous
407 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
408 a cutthrough deliver is pending, as we always want to make a connection.
409 This also avoids re-routing the message when later placing the cutthrough
410 connection after a verify cache hit.
411 Do not update it with the verify result either.
413 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
414 when routing results in more than one destination address.
416 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
417 signing (which inhibits the cutthrough capability). Previously only
418 the presence of an option was tested; now an expansion evaluating as
419 empty is permissible (obviously it should depend only on data available
420 when the cutthrough connection is made).
422 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
423 the relevant preceding SMTP command did not note the pipelining mode.
425 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
426 Previously they were not counted.
428 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
429 as one having no matching records. Previously we deferred the message
430 that needed the lookup.
432 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
433 distinguished as "(=".
435 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
436 for missing MX records. Previously it only worked for missing A records.
438 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
440 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
441 after the data-go-ahead and data-ack. Patch from Jason Betts.
443 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
444 even for a "none" policy. Patch from Tony Meyer.
446 JH/22 Fix continued use of a connection for further deliveries. If a port was
447 specified by a router, it must also match for the delivery to be
450 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
451 When one of the recipients of a message was unsuitable for the connection
452 (has no matching addresses), we lost track of needing to mark it
453 deferred. As a result mail would be lost.
455 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
457 JH/25 Decoding ACL controls is now done using a binary search; the source code
458 takes up less space and should be simpler to maintain. Merge the ACL
459 condition decode tables also, with similar effect.
461 JH/26 Fix problem with one_time used on a redirect router which returned the
462 parent address unchanged. A retry would see the parent address marked as
463 delivered, so not attempt the (identical) child. As a result mail would
466 JH/27 Fix a possible security hole, wherein a process operating with the Exim
467 UID can gain a root shell. Credit to http://www.halfdog.net/ for
468 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
471 JH/28 Enable {spool,log} filesystem space and inode checks as default.
472 Main config options check_{log,spool}_{inodes,space} are now
473 100 inodes, 10MB unless set otherwise in the configuration.
475 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
476 Previously it only applied to the main-section connection policy
479 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
481 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
482 by me. Added RFC7919 DH primes as an alternative.
484 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
485 are not in the system include path.
487 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
488 GnuTLS, when a session startup failed (eg because the client disconnected)
489 Exim did stdio operations after fclose. This was exposed by a recent
490 change which nulled out the file handle after the fclose.
492 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
493 signed directly by the cert-signing cert, rather than an intermediate
494 OCSP-signing cert. This is the model used by LetsEncrypt.
496 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
498 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
499 an incoming connection.
501 HS/02 Bug 1802: Do not half-close the connection after sending a request
504 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
505 fallback to "prime256v1".
507 JH/34 SECURITY: Use proper copy of DATA command in error message.
508 Could leak key material. Remotely exploitable. CVE-2016-9963.
514 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
515 and 3.4.4 - once the server is enabled to respond to an OCSP request
516 it does even when not requested, resulting in a stapling non-aware
517 client dropping the TLS connection.
519 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
520 support variable-length bit vectors. No functional change.
522 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
523 The I= interface field on outgoing lines is now after the H= remote
524 host field, same as incoming lines. There is a separate
525 outgoing_interface log selector which allows you to disable the
528 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
529 If not running log_selector +smtp_connection the mainlog would be held
530 open indefinitely after a "too many connections" event, including to a
531 deleted file after a log rotate. Leave the per net connection logging
532 leaving it open for efficiency as that will be quickly detected by the
533 check on the next write.
535 HS/01 Bug 1671: Fix post transport crash.
536 Processing the wait-<transport> messages could crash the delivery
537 process if the message IDs didn't exist for some reason. When
538 using 'split_spool_directory=yes' the construction of the spool
539 file name failed already, exposing the same netto behaviour.
541 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
542 mime_regex ACL conditions.
544 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
545 to DSN fail messages (bounces): remote IP, remote greeting, remote response
546 to HELO, local diagnostic string.
548 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
549 log line to debug. Even when configured with a tls authenticator many
550 client connections are expected to not authenticate in this way, so
551 an authenticate fail is not an error.
553 HS/02 Add the Exim version string to the process info. This way exiwhat
554 gives some more detail about the running daemon.
556 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
557 matter for fast-change records such as DNSBLs.
559 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
560 retry record, even if constant. There may be multiple transports with
561 different interface settings and the retry behaviour needs to be kept
564 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
566 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
568 JH/10 Bug 840: fix log_defer_output option of pipe transport
570 JH/11 Bug 830: use same host for all RCPTS of a message, even under
571 hosts_randomize. This matters a lot when combined with mua_wrapper.
573 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
574 ${quote_pgsql:<string>} operator.
576 JH/13 Bug 1708: avoid misaligned access in cached lookup.
578 JH/14 Change header file name for freeradius-client. Relevant if compiling
579 with Radius support; from the Gentoo tree and checked under Fedora.
581 JH/15 Bug 1712: Introduce $prdr_requested flag variable
583 JH/16 Bug 1714: Permit an empty string as expansion result for transport
584 option transport_filter, meaning no filtering.
586 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
588 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
589 defaults to "*" (all hosts). The variable is now available when not built
590 with TLS, default unset, mainly to enable keeping the testsuite sane.
591 If a server certificate is not supplied (via tls_certificate) an error is
592 logged, and clients will find TLS connections fail on startup. Presumably
593 they will retry in-clear.
594 Packagers of Exim are strongly encouraged to create a server certificate
595 at installation time.
597 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
598 with the $config_file variable.
600 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
601 in transport context, after the attempt, and per-recipient. The latter type
602 is per host attempted. The event data is the error message, and the errno
603 information encodes the lookup type (A vs. MX) used for the (first) host,
604 and the trailing two digits of the smtp 4xx response.
606 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
607 to write to mainlog (or rejectlog, paniclog) in the window between file
608 creation and permissions/ownership being changed. Particularly affects
609 installations where exicyclog is run as root, rather than exim user;
610 result is that the running daemon panics and dies.
612 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
614 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
615 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
616 "pri" and "weight". Note that the previous implicit priority given by the
617 list order is no longer honoured.
619 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
622 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
623 by defining SUPPORT_SOCKS.
625 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
626 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
627 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
628 variables are renamed to proxy_{local,external}_{address,port}.
630 JH/27 Move Internationalisation support from Experimental to mainline, enabled
631 for a build by defining SUPPORT_I18N
633 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
634 of the query string, and make ${quote_redis:} do that quoting.
636 JH/29 Move Events support from Experimental to mainline, enabled by default
637 and removable for a build by defining DISABLE_EVENT.
639 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
641 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
642 cached by the daemon.
644 JH/32 Move Redis support from Experimental to mainline, enabled for a build
645 by defining LOOKUP_REDIS. The libhiredis library is required.
647 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
648 keys are given for lookup.
650 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
651 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
652 only supported when built with TLS support. The PolarSSL SHA routines
653 are still used when the TLS library is too old for convenient support.
655 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
656 openssl_options), for security. OpenSSL forces this from version 1.1.0
657 server-side so match that on older versions.
659 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
660 allocation for $value could be released as the expansion processing
661 concluded, but leaving the global pointer active for it.
663 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
664 and to use the domains and local_parts ACL conditions.
666 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
667 incorrectly not doubled on cutthrough transmission, hence seen as a
668 body-termination at the receiving system - resulting in truncated mails.
669 Commonly the sender saw a TCP-level error, and retransmitted the message
670 via the normal store-and-forward channel. This could result in duplicates
671 received - but deduplicating mailstores were liable to retain only the
672 initial truncated version.
674 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
676 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
678 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
679 we're in there, support oversigning also; bug 1309.
681 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
683 HS/04 Add support for keep_environment and add_environment options.
685 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
686 either intentional arithmetic overflow during PRNG, or testing config-
689 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
690 delivery resulted in actual delivery. Cancel cutthrough before DATA
693 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
694 rejects a recipient: pass the reject to the originator.
696 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
697 Many were false-positives and ignorable, but it's worth fixing the
700 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
701 for the new environment-manipulation done at startup. Move the routines
702 from being local to tls.c to being global via the os.c file.
704 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
705 an extract embedded as result-arg for a map, the first arg for extract
706 is unavailable so we cannot tell if this is a numbered or keyed
707 extraction. Accept either.
713 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
716 JH/02 The smtp transport option "multi_domain" is now expanded.
718 JH/03 The smtp transport now requests PRDR by default, if the server offers
721 JH/04 Certificate name checking on server certificates, when exim is a client,
722 is now done by default. The transport option tls_verify_cert_hostnames
723 can be used to disable this per-host. The build option
724 EXPERIMENTAL_CERTNAMES is withdrawn.
726 JH/05 The value of the tls_verify_certificates smtp transport and main options
727 default to the word "system" to access the system default CA bundle.
728 For GnuTLS, only version 3.0.20 or later.
730 JH/06 Verification of the server certificate for a TLS connection is now tried
731 (but not required) by default. The verification status is now logged by
732 default, for both outbound TLS and client-certificate supplying inbound
735 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
738 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
739 Status Notification (bounce) messages are now MIME format per RFC 3464.
740 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
741 under the control of the dsn_advertise_hosts option, and routers may
742 have a dsn_lasthop option.
744 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
745 default, modifiable by a malware= option. The list separator for
746 the options can now be changed in the usual way. Bug 68.
748 JH/10 The smtp_receive_timeout main option is now expanded before use.
750 JH/11 The incoming_interface log option now also enables logging of the
751 local interface on delivery outgoing connections.
753 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
754 if the interface and destination host and port all match.
756 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
759 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
760 Patch from Andrew Lewis.
762 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
763 now supports optional time-restrictions, weighting, and priority
764 modifiers per server. Patch originally by <rommer@active.by>.
766 JH/16 The spamd_address main option now supports a mixed list of local
767 and remote servers. Remote servers can be IPv6 addresses, and
768 specify a port-range.
770 JH/17 Bug 68: The spamd_address main option now supports an optional
771 timeout value per server.
773 JH/18 Bug 1581: Router and transport options headers_add/remove can
774 now have the list separator specified.
776 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
779 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
782 JH/21 Support for the A6 type of dns record is withdrawn.
784 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
785 rather than the verbs used.
787 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
788 from 255 to 1024 chars.
790 JH/24 Verification callouts now attempt to use TLS by default.
792 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
793 are generic router options now. The defaults didn't change.
795 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
796 Original patch from Alexander Shikoff, worked over by JH.
798 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
799 files if ZCAT_COMMAND is not executable.
801 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
803 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
805 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
806 Normally benign, it bites when the pair was led to by a CNAME;
807 modern usage is to not canonicalize the domain to a CNAME target
808 (and we were inconsistent anyway for A-only vs AAAA+A).
810 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
812 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
813 when evaluating $sender_host_dnssec.
815 JH/31 Check the HELO verification lookup for DNSSEC, adding new
816 $sender_helo_dnssec variable.
818 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
820 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
822 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
824 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
825 documented as working, but never had. Support all but $spam_report.
827 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
828 added for tls authenticator.
830 HS/03 Add perl_taintmode main config option
836 TL/01 When running the test suite, the README says that variables such as
837 no_msglog_check are global and can be placed anywhere in a specific
838 test's script, however it was observed that placement needed to be near
839 the beginning for it to behave that way. Changed the runtest perl
840 script to read through the entire script once to detect and set these
841 variables, reset to the beginning of the script, and then run through
842 the script parsing/test process like normal.
844 TL/02 The BSD's have an arc4random API. One of the functions to induce
845 adding randomness was arc4random_stir(), but it has been removed in
846 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
847 function when detected.
849 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
850 cause callback expansion.
852 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
853 syntax errors in an expansion can be treated as a string instead of
854 logging or causing an error, due to the internal use of bool_lax
855 instead of bool when processing it.
857 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
858 server certificates when making smtp deliveries.
860 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
862 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
864 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
866 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
867 Merged patch from Sebastian Wiedenroth.
869 JH/05 Fix results-pipe from transport process. Several recipients, combined
870 with certificate use, exposed issues where response data items split
871 over buffer boundaries were not parsed properly. This eventually
872 resulted in duplicates being sent. This issue only became common enough
873 to notice due to the introduction of connection certificate information,
874 the item size being so much larger. Found and fixed by Wolfgang Breyha.
876 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
877 size buffer was used, resulting in syntax errors when an expansion
880 JH/07 Add support for directories of certificates when compiled with a GnuTLS
881 version 3.3.6 or later.
883 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
884 is EXPERIMENTAL_EVENT, the main-configuration and transport options
885 both become "event_action", the variables become $event_name, $event_data
886 and $event_defer_errno. There is a new variable $verify_mode, usable in
887 routers, transports and related events. The tls:cert event is now also
888 raised for inbound connections, if the main configuration event_action
891 TL/06 In test suite, disable OCSP for old versions of openssl which contained
892 early OCSP support, but no stapling (appears to be less than 1.0.0).
894 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
895 server certificate names available under the smtp transport option
896 "tls_verify_cert_hostname" now do not permit multi-component wildcard
899 JH/10 Time-related extraction expansions from certificates now use the main
900 option "timezone" setting for output formatting, and are consistent
901 between OpenSSL and GnuTLS compilations. Bug 1541.
903 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
904 encoded parameter in the incoming message. Bug 1558.
906 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
907 include certificate info, eximon was claiming there were spoolfile
910 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
912 JH/14 Log delivery-related information more consistently, using the sequence
913 "H=<name> [<ip>]" wherever possible.
915 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
916 are problematic for Debian distribution, omit them from the release
919 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
921 JH/16 Fix string representation of time values on 64bit time_t architectures.
924 JH/17 Fix a null-indirection in certextract expansions when a nondefault
925 output list separator was used.
930 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
931 checkers that were complaining about end of non-void function with no
934 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
935 This was a regression introduced in 4.83 by another bugfix.
937 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
939 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
940 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
946 TF/01 Correctly close the server side of TLS when forking for delivery.
948 When a message was received over SMTP with TLS, Exim failed to clear up
949 the incoming connection properly after forking off the child process to
950 deliver the message. In some situations the subsequent outgoing
951 delivery connection happened to have the same fd number as the incoming
952 connection previously had. Exim would try to use TLS and fail, logging
953 a "Bad file descriptor" error.
955 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
956 utilities have not been installed.
958 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
959 temporary space as the ACL may create new global variables.
961 TL/01 LDAP support uses per connection or global context settings, depending
962 upon the detected version of the libraries at build time.
964 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
965 to extract and use the src ip:port in logging and expansions as if it
966 were a direct connection from the outside internet. PPv2 support was
967 updated based on HAProxy spec change in May 2014.
969 JH/02 Add ${listextract {number}{list}{success}{fail}}.
971 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
972 Properly escape header and check for NULL return.
974 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
977 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
979 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
980 characters in header names, implemented as a verify condition.
981 Contributed by Michael Fischer v. Mollard.
983 TL/05 Rename SPF condition results err_perm and err_temp to standardized
984 results permerror and temperror. Previous values are deprecated but
985 still accepted. In a future release, err_perm and err_temp will be
986 completely removed, which will be a backward incompatibility if the
987 ACL tests for either of these two old results. Patch contributed by
988 user bes-internal on the mailing list.
990 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
992 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
993 selectors, in both main and reject logs.
995 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
998 JH/07 Add malware type "sock" for talking to simple daemon.
1000 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1002 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1003 routers/transports under cutthrough routing.
1005 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1006 numbers. Touch up "bool" conditional to keep the same definition.
1008 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1010 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1012 JH/12 Expand items in router/transport headers_add or headers_remove lists
1013 individually rather than the list as a whole. Bug 1452.
1015 Required for reasonable handling of multiple headers_ options when
1016 they may be empty; requires that headers_remove items with embedded
1017 colons must have them doubled (or the list-separator changed).
1019 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1020 view the policy declared in the DMARC record. Currently, $dmarc_status
1021 is a combined value of both the record presence and the result of the
1024 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1026 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1027 dnslookup router and the smtp transport (applying to the forward
1030 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1031 of ldap servers used for a specific lookup. Patch provided by Heiko
1034 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1035 New variable $lookup_dnssec_authenticated for observability.
1037 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1038 Patch submitted by Lars Timman.
1040 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1042 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1043 Requires trusted mode and valid format message id, aborts otherwise.
1044 Patch contributed by Heiko Schlichting.
1046 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1047 certextract with support for various fields. Bug 1358.
1049 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1050 is requested by default, modifiable by smtp transport option
1053 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1054 operate on certificate variables to give certificate fingerprints
1055 Also new ${sha256:cert_variable}.
1057 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1059 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1060 Christian Aistleitner.
1062 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1064 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1065 file. Patch from Wolfgang Breyha.
1067 JH/25 Expand the coverage of the delivery $host and $host_address to
1068 client authenticators run in verify callout. Bug 1476.
1070 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1071 align with daemon_smtp_ports. Bug 72.
1073 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1074 support and error reporting did not work properly.
1076 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1077 and is readable. Patch from Andrew Colin Kissa.
1079 TL/14 Enhance documentation of ${run expansion and how it parses the
1080 commandline after expansion, particularly in the case when an
1081 unquoted variable expansion results in an empty value.
1083 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1085 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1086 by renaming away from it. Addresses GH issue 15, reported by
1089 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1091 TL/15 SECURITY: prevent double expansion in math comparison functions
1092 (can expand unsanitized data). Not remotely exploitable.
1099 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1101 PP/02 Make -n do something, by making it not do something.
1102 When combined with -bP, the name of an option is not output.
1104 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1107 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1108 $sender_host_name and config options to manage this, and basic check
1111 PP/05 DSCP support for outbound connections and control modifier for inbound.
1113 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1114 (Only plugin which currently uses this is kerberos4, which nobody should
1115 be using, but we should make it available and other future plugins might
1116 conceivably use it, even though it would break NAT; stuff *should* be
1117 using channel bindings instead).
1119 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1120 name; added for Sendmail compatibility; requires admin caller.
1121 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1122 just ignore it); requires trusted caller.
1123 Also parse but ignore: -Ac -Am -X<logfile>
1126 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1128 TL/02 Add +smtp_confirmation as a default logging option.
1130 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1131 Patch by Magnus Holmgren from 2007-02-20.
1133 TL/04 Bugzilla 1281 - Spec typo.
1134 Bugzilla 1283 - Spec typo.
1135 Bugzilla 1290 - Spec grammar fixes.
1137 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1139 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1141 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1142 mailing list by by Dmitry Isaikin.
1144 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1146 JH/02 Support "G" suffix to numbers in ${if comparisons.
1148 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1150 NM/01 Bugzilla 1197 - Spec typo
1151 Bugzilla 1196 - Spec examples corrections
1153 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1155 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1156 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1159 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1160 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1162 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1163 "acl {{name}{arg}...}", and optional args on acl condition
1166 JH/05 Permit multiple router/transport headers_add/remove lines.
1168 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1170 JH/07 Avoid using a waiting database for a single-message-only transport.
1171 Performance patch from Paul Fisher. Bugzilla 1262.
1173 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1176 JH/09 Add $headers_added variable, with content from use of ACL modifier
1177 add_header (but not yet added to the message). Bugzilla 199.
1179 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1180 Pulled from Bugzilla 817 by Wolfgang Breyha.
1182 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1184 (nb: this is the same fix as in Exim 4.80.1)
1186 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1189 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1192 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1194 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1195 advertise SMTP AUTH mechanism to us, instead of a generic
1196 protocol violation error. Also, make Exim more robust to bad
1197 data from the Dovecot auth socket.
1199 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1201 When a queue runner is handling a message, Exim first routes the
1202 recipient addresses, during which it prunes them based on the retry
1203 hints database. After that it attempts to deliver the message to
1204 any remaining recipients. It then updates the hints database using
1207 So if a recipient address works intermittently, it can get repeatedly
1208 deferred at routing time. The retry hints record remains fresh so the
1209 address never reaches the final cutoff time.
1211 This is a fairly common occurrence when a user is bumping up against
1212 their storage quota. Exim had some logic in its local delivery code
1213 to deal with this. However it did not apply to per-recipient defers
1214 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1216 This change adds a proper retry rule check during routing so that the
1217 final cutoff time is checked against the message's age. We only do
1218 this check if there is an address retry record and there is not a
1219 domain retry record; this implies that previous attempts to handle
1220 the address had the retry_use_local_parts option turned on. We use
1221 this as an approximation for the destination being like a local
1222 delivery, as in LMTP.
1224 I suspect this new check makes the old local delivery cutoff check
1225 redundant, but I have not verified this so I left the code in place.
1227 TF/02 Correct gecos expansion when From: is a prefix of the username.
1229 Test 0254 submits a message to Exim with the header
1233 When I ran the test suite under the user fanf2, Exim expanded
1234 the header to contain my full name, whereas it should have added
1235 a Resent-Sender: header. It erroneously treats any prefix of the
1236 username as equal to the username.
1238 This change corrects that bug.
1240 GF/01 DCC debug and logging tidyup
1241 Error conditions log to paniclog rather than rejectlog.
1242 Debug lines prefixed by "DCC: " to remove any ambiguity.
1244 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1246 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1247 Bug spotted by Jeremy Harris; was flawed since initial commit.
1248 Would have resulted in OCSP responses post-SNI triggering an Exim
1249 NULL dereference and crash.
1251 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1253 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1254 Bug detection, analysis and fix by Samuel Thibault.
1255 Bugzilla 1331, Debian bug #698092.
1257 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1259 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1260 Server implementation by Todd Lyons, client by JH.
1261 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1262 config variable "prdr_enable" controls whether the server
1263 advertises the facility. If the client requests PRDR a new
1264 acl_data_smtp_prdr ACL is called once for each recipient, after
1265 the body content is received and before the acl_smtp_data ACL.
1266 The client is controlled by both of: a hosts_try_prdr option
1267 on the smtp transport, and the server advertisement.
1268 Default client logging of deliveries and rejections involving
1269 PRDR are flagged with the string "PRDR".
1271 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1272 fclose(). Diagnosis by Todd Lyons.
1274 PP/17 Update configure.default to handle IPv6 localhost better.
1275 Patch by Alain Williams (plus minor tweaks).
1278 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1279 This is now consistent with GnuTLS, and is now documented: the
1280 previous undocumented portable approach to treating the option as
1281 unset was to force an expansion failure. That still works, and
1282 an empty string is now equivalent.
1284 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1285 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1286 not performing validation itself.
1288 PP/20 Added force_command boolean option to pipe transport.
1289 Patch from Nick Koston, of cPanel Inc.
1291 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1294 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1296 PP/21 Fix eximon continuous updating with timestamped log-files.
1297 Broken in a format-string cleanup in 4.80, missed when I repaired the
1298 other false fix of the same issue.
1299 Report and fix from Heiko Schlichting.
1302 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1303 Report from Prashanth Katuri.
1305 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1306 It's SecureTransport, so affects any MacOS clients which use the
1307 system-integrated TLS libraries, including email clients.
1309 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1310 using a MIME ACL for non-SMTP local injection.
1311 Report and assistance in diagnosis by Warren Baker.
1313 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1315 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1317 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1318 last id that failed so it may be referenced in subsequent ACL's.
1320 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1323 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1324 ldap library initialization, allowing self-signed CA's to be
1325 used. Also properly sets require_cert option later in code by
1326 using NULL (global ldap config) instead of ldap handle (per
1327 session). Bug diagnosis and testing by alxgomz.
1329 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1330 the src/util/ subdirectory.
1332 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1333 renamed to Transport Post Delivery Action by Jeremy Harris, as
1336 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1337 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1338 redis_servers = needs to be configured which will be used by the redis
1339 lookup. Patch from Warren Baker, of The Packet Hub.
1341 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1343 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1344 hostname or reverse DNS when processing a host list. Used suggestions
1345 from multiple comments on this bug.
1347 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1349 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1350 Missed a few lines, added it to make the runtest require no keyboard
1353 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1354 contains upper case chars. Make router use caseful_local_part.
1356 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1357 support when GnuTLS has been built with p11-kit.
1363 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1365 This, or similar/improved, will also be change PP/11 of 4.82.
1371 PP/01 Handle short writes when writing local log-files.
1372 In practice, only affects FreeBSD (8 onwards).
1373 Bugzilla 1053, with thanks to Dmitry Isaikin.
1375 NM/01 Bugzilla 949 - Documentation tweak
1377 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1380 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1382 PP/02 Implemented gsasl authenticator.
1384 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1386 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1387 `pkg-config foo` for cflags/libs.
1389 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1390 with rest of GSASL and with heimdal_gssapi.
1392 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1393 `pkg-config foo` for cflags/libs for the TLS implementation.
1395 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1396 properties get this fed in as external SSF. A number of robustness
1397 and debugging improvements to the cyrus_sasl authenticator.
1399 PP/08 cyrus_sasl server now expands the server_realm option.
1401 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1402 Patch by Jeremy Harris.
1404 PP/10 Added dbmjz lookup type.
1406 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1408 PP/12 MAIL args handles TAB as well as SP, for better interop with
1409 non-compliant senders.
1410 Analysis and variant patch by Todd Lyons.
1412 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1413 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1414 Patch from Dirk Mueller <dmueller@suse.com>
1416 PP/13 tls_peerdn now print-escaped for spool files.
1417 Observed some $tls_peerdn in wild which contained \n, which resulted
1418 in spool file corruption.
1420 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1421 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1422 or write after TLS renegotiation, which otherwise led to messages
1425 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1426 as a tracking header (ie: a signed header comes before the signature).
1427 Patch from Wolfgang Breyha.
1429 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1430 comma-sep list; embedded commas doubled.
1432 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1434 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1436 Report and patch from Dmitry Banschikov.
1438 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1439 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1440 failures. We appear to now support TLS1.1+ with Exim.
1442 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1443 lets Exim select keys and certificates based upon TLS SNI from client.
1444 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1445 before an outbound SMTP session. New log_selector, +tls_sni.
1447 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1448 NULL dereference. Report and patch from Alun Jones.
1450 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1451 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1452 Not seeing resolver debug output on NetBSD, but suspect this is a
1453 resolver implementation change.
1455 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1456 Left warnings. Added "eximon gdb" invocation mode.
1458 PP/21 Defaulting "accept_8bitmime" to true, not false.
1460 PP/22 Added -bw for inetd wait mode support.
1462 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1463 locate the relevant includes and libraries. Made this the default.
1465 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1466 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1468 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1469 This may cause build issues on older platforms.
1471 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1472 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1473 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1474 Added SNI support via GnuTLS too.
1475 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1477 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1479 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1480 Applied second patch from Janne, implementing suggestion to default
1481 multiple-strings-in-record handling to match SPF spec.
1483 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1485 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1486 read-only, out of scope).
1487 Patch from Wolfgang Breyha, report from Stuart Northfield.
1489 PP/29 Fix three issues highlighted by clang analyser static analysis.
1490 Only crash-plausible issue would require the Cambridge-specific
1491 iplookup router and a misconfiguration.
1492 Report from Marcin Mirosław.
1494 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1496 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1497 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1498 As part of this, removing so much warning spew let me fix some minor
1499 real issues in debug logging.
1501 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1502 assignment on my part. Fixed.
1504 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1505 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1506 Janne Snabb (who went above and beyond: thank you).
1508 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1509 string otherwise requires a connection and a bunch more work and it's
1510 relatively easy to get wrong. Should also expose TLS library linkage
1513 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1514 64-bit ${eval} (JH/03).
1516 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1517 GNU libc to support some of the 64-bit stuff, should not lead to
1518 conflicts. Defined before os.h is pulled in, so if a given platform
1519 needs to override this, it can.
1521 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1522 protection layer was required, which is not implemented.
1523 Bugzilla 1254, patch from Wolfgang Breyha.
1525 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1526 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1527 tls_dhparam take prime identifiers. Also unbreak combination of
1528 OpenSSL+DH_params+TLSSNI.
1530 PP/39 Disable SSLv2 by default in OpenSSL support.
1536 PP/01 Solaris build fix for Oracle's LDAP libraries.
1537 Bugzilla 1109, patch from Stephen Usher.
1539 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1541 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1544 TF/02 Fix a couple more cases where we did not log the error message
1545 when unlink() failed. See also change 4.74-TF/03.
1547 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1548 lock up or crash if it happened to be inside a call to libc when it
1549 got a SIGUSR1 from exiwhat.
1551 The SIGUSR1 handler appends the current process status to the process
1552 log which is later printed by exiwhat. It used to use the general
1553 purpose logging code to do this, but several functions it calls are
1554 not safe for signals.
1556 The new output code in the SIGUSR1 handler is specific to the process
1557 log, and simple enough that it's easy to inspect for signal safety.
1558 Removing some special cases also simplifies the general logging code.
1559 Removing the spurious timestamps from the process log simplifies
1562 TF/04 Improved ratelimit ACL condition.
1564 The /noupdate option has been deprecated in favour of /readonly which
1565 has clearer semantics. The /leaky, /strict, and /readonly update modes
1566 are mutually exclusive. The update mode is no longer included in the
1567 database key; it just determines when the database is updated. (This
1568 means that when you upgrade Exim will forget old rate measurements.)
1570 Exim now checks that the per_* options are used with an update mode that
1571 makes sense for the current ACL. For example, when Exim is processing a
1572 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1573 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1574 must specify per_mail/readonly. If you omit the update mode it defaults to
1575 /leaky where that makes sense (as before) or /readonly where required.
1577 The /noupdate option is now undocumented but still supported for
1578 backwards compatibility. It is equivalent to /readonly except that in
1579 ACLs where /readonly is required you may specify /leaky/noupdate or
1580 /strict/noupdate which are treated the same as /readonly.
1582 A useful new feature is the /count= option. This is a generalization
1583 of the per_byte option, so that you can measure the throughput of other
1584 aggregate values. For example, the per_byte option is now equivalent
1585 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1587 The per_rcpt option has been generalized using the /count= mechanism
1588 (though it's more complicated than the per_byte equivalence). When it is
1589 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1590 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1591 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1592 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1593 Note that using per_rcpt with a non-readonly update mode in more than
1594 one ACL will cause the recipients to be double-counted. (The per_mail
1595 and per_byte options don't have this problem.)
1597 The handling of very low rates has changed slightly. If the computed rate
1598 is less than the event's count (usually one) then this event is the first
1599 after a long gap. In this case the rate is set to the same as this event's
1600 count, so that the first message of a spam run is counted properly.
1602 The major new feature is a mechanism for counting the rate of unique
1603 events. The new per_addr option counts the number of different
1604 recipients that someone has sent messages to in the last time period. It
1605 behaves like per_rcpt if all the recipient addresses are different, but
1606 duplicate recipient addresses do not increase the measured rate. Like
1607 the /count= option this is a general mechanism, so the per_addr option
1608 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1609 example, measure the rate that a client uses different sender addresses
1610 with the options per_mail/unique=$sender_address. There are further
1611 details in the main documentation.
1613 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1615 TF/06 Removed a few PCRE remnants.
1617 TF/07 Automatically extract Exim's version number from tags in the git
1618 repository when doing development or release builds.
1620 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1621 Bugzilla 879. Patch from Paul Fisher.
1623 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1624 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1627 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1629 PP/05 Variable $av_failed, true if the AV scanner deferred.
1630 Bugzilla 1078. Patch from John Horne.
1632 PP/06 Stop make process more reliably on build failure.
1633 Bugzilla 1087. Patch from Heiko Schlittermann.
1635 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1636 Bugzilla 1089. Patch from Heiko Schlittermann.
1638 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1639 Bugzilla 1131. Patch from Holger Weiß.
1641 PP/09 Handle IPv6 addresses with SPF.
1642 Bugzilla 860. Patch from Wolfgang Breyha.
1644 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1646 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1649 PP/11 match_* no longer expand right-hand-side by default.
1650 New compile-time build option, EXPAND_LISTMATCH_RHS.
1651 New expansion conditions, "inlist", "inlisti".
1653 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1655 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1657 PP/14 fix log_write() format string regression from TF/03.
1658 Bugzilla 1152. Patch from Dmitry Isaikin.
1664 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1666 PP/02 Harmonised TLS library version reporting; only show if debugging.
1667 Layout now matches that introduced for other libraries in 4.74 PP/03.
1669 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1671 PP/04 New "dns_use_edns0" global option.
1673 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1676 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1677 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1679 TK/01 Updated PolarSSL code to 0.14.2.
1680 Bugzilla 1097. Patch from Andreas Metzler.
1682 PP/07 Catch divide-by-zero in ${eval:...}.
1683 Fixes bugzilla 1102.
1685 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1688 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1689 format-string attack -- SECURITY: remote arbitrary code execution.
1691 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1692 time unintentionally subject to list matching rules, letting the header
1693 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1694 arbitrary string expansion). This allowed for information disclosure.
1696 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1697 INT_MIN/-1 -- value coerced to INT_MAX.
1703 NM/01 Workaround for PCRE version dependency in version reporting
1706 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1707 This fixes portability to compilers other than gcc, notably
1708 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1710 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1711 makefiles for portability to HP-UX and POSIX correctness.
1713 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1714 Also via indented variable definition in the Makefile.
1715 (Debugging by Oliver Heesakkers).
1717 PP/02 Restore caching of spamd results with expanded spamd_address.
1718 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1720 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1721 Improves build reliability. Fix from: Frank Elsner
1723 NM/02 Fix wide character breakage in the rfc2047 coding
1724 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1726 NM/03 Allow underscore in dnslist lookups
1727 Fixes bug 1026. Patch from Graeme Fowler
1729 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1730 Code patches from Adam Ciarcinski of NetBSD.
1732 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1735 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1736 is logged, to avoid truncation. Patch from John Horne.
1738 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1739 Patch from Jakob Hirsch.
1741 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1742 SQL string expansion failure details.
1743 Patch from Andrey Oktyabrski.
1745 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1746 Patch from Simon Arlott.
1748 PP/09 New lookups functionality failed to compile on old gcc which rejects
1749 extern declarations in function scope.
1750 Patch from Oliver Fleischmann
1752 PP/10 Use sig_atomic_t for flags set from signal handlers.
1753 Check getgroups() return and improve debugging.
1754 Fixed developed for diagnosis in bug 927 (which turned out to be
1757 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1758 Patch from Mark Zealey.
1760 PP/12 Bugzilla 1056: Improved spamd server selection.
1761 Patch from Mark Zealey.
1763 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1764 Based on patch from Heiko Schlittermann.
1766 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1767 Patch from Uwe Doering, sign-off by Michael Haardt.
1769 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1770 variable declaration deep within a block. Bug and patch from
1773 PP/15 lookups-Makefile IRIX compatibility coercion.
1775 PP/16 Make DISABLE_DKIM build knob functional.
1777 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1778 Patch by Simon Arlott
1780 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1781 variable argument macros. Our copy now differs from upstream.
1787 TF/01 Failure to get a lock on a hints database can have serious
1788 consequences so log it to the panic log.
1790 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1791 controlled using the smtp_confirmation log selector.
1793 TF/03 Include the error message when we fail to unlink a spool file.
1795 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1796 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1797 for maintaining out-of-tree patches for some time.
1799 PP/01 Bugzilla 139: Documentation and portability issues.
1800 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1801 Handle per-OS dynamic-module compilation flags.
1803 PP/02 Let /dev/null have normal permissions.
1804 The 4.73 fixes were a little too stringent and complained about the
1805 permissions on /dev/null. Exempt it from some checks.
1806 Reported by Andreas M. Kirchwitz.
1808 PP/03 Report version information for many libraries, including
1809 Exim version information for dynamically loaded libraries. Created
1810 version.h, now support a version extension string for distributors
1811 who patch heavily. Dynamic module ABI change.
1813 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1814 privilege escalation vulnerability whereby the Exim run-time user
1815 can cause root to append content of the attacker's choosing to
1818 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1821 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1822 If dropping privileges for untrusted macros, we disabled normal logging
1823 on the basis that it would fail; for the Exim run-time user, this is not
1824 the case, and it resulted in successful deliveries going unlogged.
1825 Fixed. Reported by Andreas Metzler.
1831 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1832 only prepend for the Resent-* case. Fixes regression introduced in
1833 Exim 4.70 by NM/22 for Bugzilla 607.
1835 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1836 increasing numbers of administrators be bitten by this.
1838 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1840 PP/03 Bugzilla 994: added openssl_options main configuration option.
1842 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1844 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1846 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1848 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1849 without a peer certificate, leading to a segfault because of an
1850 assumption that peers always have certificates. Be a little more
1851 paranoid. Problem reported by Martin Tscholak.
1853 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1854 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1855 NB: ClamAV planning to remove STREAM in "middle of 2010".
1856 CL also introduces -bmalware, various -d+acl logging additions and
1857 more caution in buffer sizes.
1859 PP/09 Implemented reverse_ip expansion operator.
1861 PP/10 Bugzilla 937: provide a "debug" ACL control.
1863 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1865 PP/12 Bugzilla 973: Implement --version.
1867 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1869 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1871 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1873 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1874 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1875 ignore trailing whitespace.
1877 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1879 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1882 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1883 Notification from Dr Andrew Aitchison.
1885 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1886 ExtendedDetectionInfo response format.
1887 Notification from John Horne.
1889 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1892 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1893 XSL and documented dependency on system catalogs, with examples of how
1896 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1899 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1900 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1901 configuration file which is writeable by the Exim user or group.
1903 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1904 of configuration files to cover files specified with the -C option if
1905 they are going to be used with root privileges, not just the default
1908 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1909 option (effectively making it always true).
1911 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1912 files to be used while preserving root privileges.
1914 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1915 that rogue child processes cannot use them.
1917 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1918 run-time user, instead of root.
1920 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1921 Exim run-time user without dropping privileges.
1923 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1924 result string, instead of calling string_vformat() twice with the same
1927 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1928 for other users. Others should always drop root privileges if they use
1929 -C on the command line, even for a whitelisted configure file.
1931 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1933 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1939 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1940 $data_path, and $header_path variables; fixed documentation bugs and
1943 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1944 exipick to access non-standard spools, including the "frozen" queue
1947 NM/01 Bugzilla 965: Support mysql stored procedures.
1948 Patch from Alain Williams
1950 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1952 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1953 Patch from Andreas Metzler
1955 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1956 Patch from Kirill Miazine
1958 NM/05 Bugzilla 671: Added umask to procmail example.
1960 JJ/03 installed exipick 20100323.0, fixing doc bug
1962 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1963 directory. Notification and patch from Dan Rosenberg.
1965 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1967 TK/02 Improve log output when DKIM signing operation fails.
1969 MH/01 Treat the transport option dkim_domain as a colon separated
1970 list, not as a single string, and sign the message with each element,
1971 omitting multiple occurences of the same signer.
1973 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1974 Bugzilla 985, 986. Patch by Simon Arlott
1976 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1977 Patch by Simon Arlott
1979 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1980 MBX locking. Notification from Dan Rosenberg.
1986 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1988 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1990 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1992 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1994 NM/04 Bugzilla 915: Flag broken perl installation during build.
2000 TK/01 Added patch by Johannes Berg that expands the main option
2001 "spamd_address" if it starts with a dollar sign.
2003 TK/02 Write list of recipients to X-Envelope-Sender header when building
2004 the mbox-format spool file for content scanning (suggested by Jakob
2007 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2008 (http://www.dcc-servers.net/) support via dccifd. Activated by
2009 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2011 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2012 by Mark Daniel Reidel <mr@df.eu>.
2014 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2015 When building exim an external PCRE library is now needed -
2016 PCRE is a system library on the majority of modern systems.
2017 See entry on PCRE_LIBS in EDITME file.
2019 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2020 conversation. Added nologin parameter to request.
2021 Patch contributed by Kirill Miazine.
2023 TF/01 Do not log submission mode rewrites if they do not change the address.
2025 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2027 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2028 log files in place. Contributed by Roberto Lima.
2030 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2032 TF/03 Bugzilla 615: When checking the local_parts router precondition
2033 after a local_part_suffix or local_part_prefix option, Exim now
2034 does not use the address's named list lookup cache, since this
2035 contains cached lookups for the whole local part.
2037 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2038 Robert Millan. Documentation is in experimental-spec.txt.
2040 TF/04 Bugzilla 668: Fix parallel build (make -j).
2042 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2044 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2045 Patch provided by Jan Srzednicki.
2047 TF/05 Leading white space used to be stripped from $spam_report which
2048 wrecked the formatting. Now it is preserved.
2050 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2051 that they are available at delivery time.
2053 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2055 TF/08 TLS error reporting now respects the incoming_interface and
2056 incoming_port log selectors.
2058 TF/09 Produce a more useful error message if an SMTP transport's hosts
2059 setting expands to an empty string.
2061 NM/06 Bugzilla 744: EXPN did not work under TLS.
2062 Patch provided by Phil Pennock.
2064 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2065 Patch provided by Richard Godbee.
2067 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2068 acl_smtp_notquit, added index entry.
2070 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2071 Patch provided by Eugene Bujak.
2073 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2074 accept(). Patch provided by Maxim Dounin.
2076 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2077 Patch provided by Phil Pennock.
2079 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2081 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2082 Patch provided by Brad "anomie" Jorsch.
2084 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2085 Patch provided by Dean Brooks.
2087 TK/05 Add native DKIM support (does not depend on external libraries).
2089 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2090 Patch provided by Graeme Fowler.
2092 NM/16 Bugzilla 851: Documentation example syntax fix.
2094 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2096 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2099 NM/19 Bugzilla 745: TLS version reporting.
2100 Patch provided by Phil Pennock.
2102 NM/20 Bugzilla 167: bool: condition support.
2103 Patch provided by Phil Pennock.
2105 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2106 clients. Patch provided by Phil Pennock.
2108 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2109 Patch provided by Brad "anomie" Jorsch.
2111 NM/23 Bugzilla 687: Fix misparses in eximstats.
2112 Patch provided by Heiko Schlittermann.
2114 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2115 Patch provided by Heiko Schlittermann.
2117 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2118 plus update to original patch.
2120 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2122 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2123 Patch provided by David Brownlee.
2125 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2127 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2129 NM/30 Bugzilla 888: TLS documentation bugfixes.
2131 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2133 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2134 Unlike the original bugzilla I have changed all shell scripts in src tree.
2136 NM/33 Bugzilla 898: Transport filter timeout fix.
2137 Patch by Todd Rinaldo.
2139 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2140 Patch by Serge Demonchaux.
2142 NM/35 Bugzilla 39: Base64 decode bug fixes.
2143 Patch by Jakob Hirsch.
2145 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2147 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2149 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2151 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2157 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2158 ALT-N's libdkim that I have put here:
2159 http://duncanthrax.net/exim-experimental/
2161 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2162 were called 'true' and it seems that C99 defines that as a reserved
2163 keyword to be used with 'bool' variable types. That means you could
2164 not include C99-style headers which use bools without triggering
2165 build errors in sieve.c.
2167 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2168 as mailq or other aliases. Changed the --help handling significantly
2169 to do whats expected. exim_usage() emits usage/help information.
2171 SC/01 Added the -bylocaldomain option to eximstats.
2173 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2175 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2177 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2183 PH/01 Another patch from the Sieve maintainer.
2185 PH/02 When an IPv6 address is converted to a string for single-key lookup
2186 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2187 dots are used instead of colons so that keys in lsearch files need not
2188 contain colons. This was done some time before quoting was made available
2189 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2190 (notated using the quote facility) so as to distinguish them from IPv4
2191 keys. This meant that lookups for IP addresses in host lists did not work
2192 for iplsearch lookups.
2194 This has been fixed by arranging for IPv6 addresses to be expressed with
2195 colons if the lookup type is iplsearch. This is not incompatible, because
2196 previously such lookups could never work.
2198 The situation is now rather anomalous, since one *can* have colons in
2199 ordinary lsearch keys. However, making the change in all cases is
2200 incompatible and would probably break a number of configurations.
2202 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2205 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2206 conversion specification without a maximum field width, thereby enabling
2207 a rogue spamd server to cause a buffer overflow. While nobody in their
2208 right mind would setup Exim to query an untrusted spamd server, an
2209 attacker that gains access to a server running spamd could potentially
2210 exploit this vulnerability to run arbitrary code as the Exim user.
2212 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2213 $primary_hostname instead of what libspf2 thinks the hosts name is.
2215 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2216 a directory entry by the name of the lookup key. Previously, if a
2217 symlink pointed to a non-existing file or a file in a directory that
2218 Exim lacked permissions to read, a lookup for a key matching that
2219 symlink would fail. Now it is enough that a matching directory entry
2220 exists, symlink or not. (Bugzilla 503.)
2222 PH/03 The body_linecount and body_zerocount variables are now exported in the
2225 PH/04 Added the $dnslist_matched variable.
2227 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2228 This means they are set thereafter only if the connection becomes
2231 PH/06 Added the client_condition to authenticators so that some can be skipped
2232 by clients under certain conditions.
2234 PH/07 The error message for a badly-placed control=no_multiline_responses left
2235 "_responses" off the end of the name.
2237 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2239 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2240 (without spaces) instead of just copying the configuration text.
2242 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2244 PH/11 Added $max_received_linelength.
2246 PH/12 Added +ignore_defer and +include_defer to host lists.
2248 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2249 way in which PCRE > 7.0 is built.
2251 PH/14 Implemented queue_only_load_latch.
2253 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2254 MAIL command. The effect was to mangle the value on 64-bit systems.
2256 PH/16 Another patch from the Sieve maintainer.
2258 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2260 PH/18 If a system quota error occurred while trying to create the file for
2261 a maildir delivery, the message "Mailbox is full" was not appended to the
2262 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2263 only to a quota excession during the actual writing of the file.
2265 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2266 characters?) which causes problems in log lines. The DN values are now
2267 passed through string_printing() before being added to log lines.
2269 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2270 and InterBase are left for another time.)
2272 PH/21 Added message_body_newlines option.
2274 PH/22 Guard against possible overflow in moan_check_errorcopy().
2276 PH/23 POSIX allows open() to be a macro; guard against that.
2278 PH/24 If the recipient of an error message contained an @ in the local part
2279 (suitably quoted, of course), incorrect values were put in $domain and
2280 $local_part during the evaluation of errors_copy.
2286 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2287 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2290 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2291 issue a MAIL command.
2293 PH/02 In an ACL statement such as
2295 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2297 if a client was not listed at all, or was listed with a value other than
2298 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2299 the condition was not true (as it should be), so access was not denied.
2300 The bug was that the ! inversion was incorrectly passed on to the second
2301 item. This has been fixed.
2303 PH/03 Added additional dnslists conditions == and =& which are different from
2304 = and & when the dns lookup returns more than one IP address.
2306 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2307 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2309 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2310 FSYNC, which compiles an option called disable_fsync that allows for
2311 bypassing fsync(). The documentation is heavily laced with warnings.
2313 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2315 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2316 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2317 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2318 including adding "make clean"; (3) Added -fPIC when compiling the test
2319 dynamically loaded module, to get rid of a warning.
2321 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2322 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2323 = 0s. The bug is otherwise harmless.
2325 PH/07 There was a bug in the dovecot authenticator such that the value of
2326 $auth1 could be overwritten, and so not correctly preserved, after a
2327 successful authentication. This usually meant that the value preserved by
2328 the server_setid option was incorrect.
2330 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2332 PH/09 Installed PCRE release 7.0.
2334 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2335 run for batched SMTP input. It is now run at the start of every message
2336 in the batch. While fixing this I discovered that the process information
2337 (output by running exiwhat) was not always getting set for -bs and -bS
2338 input. This is fixed, and it now also says "batched" for BSMTP.
2340 PH/11 Added control=no_pipelining.
2342 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2343 patch, slightly modified), and move the expansion of helo_data till after
2344 the connection is made in the smtp transport (so it can use these
2347 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2349 PH/14 Added log_selector = +pid.
2351 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2353 PH/16 Add ${if forany and ${if forall.
2355 PH/17 Added dsn_from option to vary the From: line in DSNs.
2357 PH/18 Flush SMTP output before performing a callout, unless control =
2358 no_callout_flush is set.
2360 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2361 was true (the default) a successful delivery failed to delete the retry
2362 item, thus causing premature timeout of the address. The bug is now
2365 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2367 PH/21 Long custom messages for fakedefer and fakereject are now split up
2368 into multiline responses in the same way that messages for "deny" and
2369 other ACL rejections are.
2371 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2372 with slight modification.
2374 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2375 draft, changing the syntax and factoring some duplicate code".
2377 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2378 for deliveries of the second and subsequent messages over the same SMTP
2381 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2382 ${reduce, with only minor "tidies".
2384 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2386 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2387 expansion side effects.
2389 PH/27 When a message times out after an over-quota error from an Exim-imposed
2390 quota, the bounce message says "mailbox is full". This message was not
2391 being given when it was a system quota that was exceeded. It now should
2394 MH/03 Made $recipients available in local_scan(). local_scan() already has
2395 better access to the recipient list through recipients_list[], but
2396 $recipients can be useful in postmaster-provided expansion strings.
2398 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2399 in the case of a MAIL command with additional options following the
2400 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2401 were accidentally chopped off.
2403 PH/29 SMTP synchronization checks are implemented when a command is read -
2404 there is a check that no more input is waiting when there shouldn't be
2405 any. However, for some commands, a delay in an ACL can mean that it is
2406 some time before the response is written. In this time, more input might
2407 arrive, invalidly. So now there are extra checks after an ACL has run for
2408 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2409 pipelining has not been advertised.
2411 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2413 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2414 returned in the SMTP response when smtp_return_error_details was false.
2415 This has been fixed.
2417 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2418 instead of the C I/O that was originally supplied, because problems were
2419 reported on Solaris.
2421 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2422 Exim which did not show up earlier: it was assuming that a call to
2423 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2424 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2425 was a macro that became an assignment, so it seemed to work. This has
2426 changed to a proper function call with a void return, hence the compile
2427 error. Exim's code has been fixed.
2429 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2432 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2434 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2436 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2437 because it turns out to be too big or there is a timeout) there is no
2438 "Completed" line in the log. When some messages of this type were
2439 selected by exigrep, they were listed as "not completed". Others were
2440 picked up by some special patterns. I have improved the selection
2441 criteria to be more general.
2443 PH/38 The host_find_failed option in the manualroute router can now be set
2444 to "ignore", to completely ignore a host whose IP address cannot be
2445 found. If all hosts are ignored, the behaviour is controlled by the new
2446 host_all_ignored option.
2448 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2449 homing or because of multiple MX records with /mx) generated more than
2450 one IP address, and the following item turned out to be the local host,
2451 all the secondary addresses of the first item were incorrectly removed
2452 from the list, along with the local host and any following hosts (which
2453 is what is supposed to happen).
2455 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2456 whoever called Exim into the -H file. In the case of the daemon it was
2457 behaving confusingly. When first started, it used values for whoever
2458 started the daemon, but after a SIGHUP it used the Exim user (because it
2459 calls itself on a restart). I have changed the code so that it now always
2462 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2463 message are rejected with the same error (e.g. no authentication or bad
2464 sender address), and a DATA command is nevertheless sent (as can happen
2465 with PIPELINING or a stupid MUA), the error message that was given to the
2466 RCPT commands is included in the rejection of the DATA command. This is
2467 intended to be helpful for MUAs that show only the final error to their
2470 PH/42 Another patch from the Sieve maintainer.
2472 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2473 Eximstats - Fixed some broken HTML links and added missing column headers
2475 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2476 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2478 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2480 PH/43 Yet another patch from the Sieve maintainer.
2482 PH/44 I found a way to check for a TCP/IP connection going away before sending
2483 the response to the final '.' that terminates a message, but only in the
2484 case where the client has not sent further data following the '.'
2485 (unfortunately, this is allowed). However, in many cases there won't be
2486 any further data because there won't be any more messages to send. A call
2487 to select() can be used: if it shows that the input is "ready", there is
2488 either input waiting, or the socket has been closed. An attempt to read
2489 the next input character can distinguish the two cases. Previously, Exim
2490 would have sent an OK response which the client would never have see.
2491 This could lead to message repetition. This fix should cure that, at
2492 least in a lot of common cases.
2494 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2495 advertised in response to EHLO.
2501 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2502 fixed by 4.65/MH/01 (is this a record?) are fixed:
2504 (i) An empty string was always treated as zero by the numeric comparison
2505 operators. This behaviour has been restored.
2507 (ii) It is documented that the numeric comparison operators always treat
2508 their arguments as decimal numbers. This was broken in that numbers
2509 starting with 0 were being interpreted as octal.
2511 While fixing these problems I realized that there was another issue that
2512 hadn't been noticed. Values of message_size_limit (both the global option
2513 and the transport option) were treated as octal if they started with 0.
2514 The documentation was vague. These values are now always treated as
2515 decimal, and I will make that clear in the documentation.
2521 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2522 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2525 MH/01 Don't check that the operands of numeric comparison operators are
2526 integers when their expansion is in "skipping" mode (fixes bug
2527 introduced by 4.64-PH/07).
2529 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2530 child addresses, Exim now panics and dies. Previously, because the count
2531 is held in a short int, deliveries were likely to be lost. As such a
2532 large number of recipients for a single message is ridiculous
2533 (performance will be very, very poor), I have chosen to impose a limit
2534 rather than extend the field.
2540 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2541 leftover -K file (the existence of which was triggered by #402).
2542 While we were at it, introduced process PID as part of the -K
2543 filename. This should rule out race conditions when creating
2546 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2547 processing considerably. Previous code took too long for large mails,
2548 triggering a timeout which in turn triggers #401.
2550 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2551 in the DK code in transports.c. sendfile() is not really portable,
2552 hence the _LINUX specificness.
2554 TF/01 In the add_headers option to the mail command in an Exim filter,
2555 there was a bug that Exim would claim a syntax error in any
2556 header after the first one which had an odd number of characters
2559 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2560 callout verification, Exim cached a "reject" for the entire domain. This
2561 is correct for most verifications, but it is not correct for a recipient
2562 verification with use_sender or use_postmaster set, because in that case
2563 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2564 case of MAIL FROM:<> rejection from other early rejections (e.g.
2565 rejection of HELO). When verifying a recipient using a non-null MAIL
2566 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2567 Whatever the result of the callout, the value of the domain cache is
2568 left unchanged (for any other kind of callout, getting as far as trying
2569 RCPT means that the domain itself is ok).
2571 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2574 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2575 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2576 session, the daemon ignores ECONNECT errors and logs others; it now
2577 ignores EPIPE as well.
2579 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2580 (quoted-printable decoding).
2582 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2583 later the small subsequent patch to fix an introduced bug.
2585 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2587 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2589 PH/08 An error is now given if message_size_limit is specified negative.
2591 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2592 to be given (somewhat) arbitrary names.
2594 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2597 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2598 miscellaneous code fixes
2600 PH/10 Added the log_reject_target ACL modifier to specify where to log
2603 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2604 hostname. This is wrong, because it relates to the incoming message (and
2605 probably the interface on which it is arriving) and not to the outgoing
2606 callout (which could be using a different interface). This has been
2607 changed to use the value of the helo_data option from the smtp transport
2608 instead - this is what is used when a message is actually being sent. If
2609 there is no remote transport (possible with a router that sets up host
2610 addresses), $smtp_active_hostname is used.
2612 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2613 tweaks were necessary in order to get it to work (see also 21 below):
2614 (a) The code assumed that strncpy() returns a negative number on buffer
2615 overflow, which isn't the case. Replaced with Exim's string_format()
2617 (b) There were several signed/unsigned issues. I just did the minimum
2618 hacking in of casts. There is scope for a larger refactoring.
2619 (c) The code used strcasecmp() which is not a standard C function.
2620 Replaced with Exim's strcmpic() function.
2621 (d) The code set only $1; it now sets $auth1 as well.
2622 (e) A simple test gave the error "authentication client didn't specify
2623 service in request". It would seem that Dovecot has changed its
2624 interface. Fortunately there's a specification; I followed it and
2625 changed what the client sends and it appears to be working now.
2627 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2630 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2631 address A is aliased to B and C, where B exists and C does not. Without
2632 -v the output is "A verified" because verification stops after a
2633 successful redirection if more than one address is generated. However,
2634 with -v the child addresses are also verified. Exim was outputting "A
2635 failed to verify" and then showing the successful verification for C,
2636 with its parentage. It now outputs "B failed to verify", showing B's
2637 parentage before showing the successful verification of C.
2639 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2640 look up a TXT record in a specific list after matching in a combined
2643 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2644 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2645 they consult the DNS. I had assumed they would set it the way they
2646 wanted; and indeed my experiments on Linux seem to show that in some
2647 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2648 To be on the safe side, however, I have now made the interface to
2649 host_find_byname() similar to host_find_bydns(), with an argument
2650 containing the DNS resolver options. The host_find_byname() function now
2651 sets these options at its start, just as host_find_bydns() does. The smtp
2652 transport options dns_qualify_single and dns_search_parents are passed to
2653 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2654 of host_find_byname() use the default settings of RES_DEFNAMES
2655 (qualify_single) but not RES_DNSRCH (search_parents).
2657 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2658 spool_read_header() do less string testing, by means of a preliminary
2659 switch on the second character of optional "-foo" lines. (This is
2660 overdue, caused by the large number of possibilities that now exist.
2661 Originally there were few.) While I was there, I also converted the
2662 str(n)cmp tests so they don't re-test the leading "-" and the first
2663 character, in the hope this might squeeze out yet more improvement.
2665 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2666 flag allowing group syntax was set by the header_syntax check but not
2667 turned off, possible causing trouble later; (2) The flag was not being
2668 set at all for the header_verify test, causing "group"-style headers to
2669 be rejected. I have now set it in this case, and also caused header_
2670 verify to ignore an empty address taken from a group. While doing this, I
2671 came across some other cases where the code for allowing group syntax
2672 while scanning a header line wasn't quite right (mostly, not resetting
2673 the flag correctly in the right place). These bugs could have caused
2674 trouble for malformed header lines. I hope it is now all correct.
2676 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2677 with the "reply" argument non-NULL. The code, however (which originally
2678 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2679 but it didn't always do it. This confused somebody who was copying the
2680 code for some other use. I have removed all the tests.
2682 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2683 feature that was used to support insecure browsers during the U.S. crypto
2684 embargo. It requires special client support, and Exim is probably the
2685 only MTA that supported it -- and would never use it because real RSA is
2686 always available. This code has been removed, because it had the bad
2687 effect of slowing Exim down by computing (never used) parameters for the
2688 RSA_EXPORT functionality.
2690 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2691 authenticator to fail if there's a tab character in the incoming data
2692 (there should never be unless someone is messing about, as it's supposed
2693 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2694 if the connection is using TLS or if the remote IP is the same as the
2695 local IP, and the "valid-client-cert option" if a client certificate has
2698 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2699 authenticators. This can be used for authorization after authentication
2700 succeeds. (In the case of plaintext, it servers for both authentication
2703 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2704 if any retry times were supplied.
2706 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2707 connection, where there is no client IP address to check. In this
2708 situation, the verify now always succeeds.
2710 PH/25 Applied John Jetmore's -Mset patch.
2712 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2714 PH/27 In a string expansion for a processed (not raw) header when multiple
2715 headers of the same name were present, leading whitespace was being
2716 removed from all of them, but trailing whitespace was being removed only
2717 from the last one. Now trailing whitespace is removed from each header
2718 before concatenation. Completely empty headers in a concatenation (as
2719 before) are ignored.
2721 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2722 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2724 PH/29 [Removed. This was a change that I later backed out, and forgot to
2725 correct the ChangeLog entry (that I had efficiently created) before
2726 committing the later change.]
2728 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2729 address errors (4xx response to RCPT) over the same connection as other
2730 messages routed to the same hosts. Such deliveries are always "forced",
2731 so retry times are not inspected. This resulted in far too many retries
2732 for the affected addresses. The effect occurred only when there were more
2733 hosts than the hosts_max_try setting in the smtp transport when it had
2734 the 4xx errors. Those hosts that it had tried were not added to the list
2735 of hosts for which the message was waiting, so if all were tried, there
2736 was no problem. Two fixes have been applied:
2738 (i) If there are any address or message errors in an SMTP delivery, none
2739 of the hosts (tried or untried) are now added to the list of hosts
2740 for which the message is waiting, so the message should not be a
2741 candidate for sending over the same connection that was used for a
2742 successful delivery of some other message. This seems entirely
2743 reasonable: after all the message is NOT "waiting for some host".
2744 This is so "obvious" that I'm not sure why it wasn't done
2745 previously. Hope I haven't missed anything, but it can't do any
2746 harm, as the worst effect is to miss an optimization.
2748 (ii) If, despite (i), such a delivery is accidentally attempted, the
2749 routing retry time is respected, so at least it doesn't keep
2750 hammering the server.
2752 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2753 in ${readsocket because some servers need this prod.
2755 PH/32 Added some extra debug output when updating a wait-xxx database.
2757 PH/33 The hint "could be header name not terminated by colon", which has been
2758 given for certain expansion errors for a long time, was not being given
2759 for the ${if def:h_colon_omitted{... case.
2761 PH/34 The spec says: "With one important exception, whenever a domain list is
2762 being scanned, $domain contains the subject domain." There was at least
2763 one case where this was not true.
2765 PH/35 The error "getsockname() failed: connection reset by peer" was being
2766 written to the panic log as well as the main log, but it isn't really
2767 panic-worthy as it just means the connection died rather early on. I have
2768 removed the panic log writing for the ECONNRESET error when getsockname()
2771 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2772 runs only) independently of the message's sender address. This meant
2773 that, if the 4xx error was in fact related to the sender, a different
2774 message to the same recipient with a different sender could confuse
2775 things. In particular, this can happen when sending to a greylisting
2776 server, but other circumstances could also provoke similar problems.
2777 I have changed the default so that the retry time for these errors is now
2778 based a combination of the sender and recipient addresses. This change
2779 can be overridden by setting address_retry_include_sender=false in the
2782 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2783 remote server are returned as part of bounce messages. This was not
2784 happening for LMTP over a pipe (the lmtp transport), but now it is the
2785 same for both kinds of LMTP.
2787 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2788 in header lines that were in fact CNAMEs. This is no longer the case.
2790 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2791 and queue runs started by the daemon processed all messages. This has
2792 been fixed so that -R and -S can now usefully be given with -q<time>.
2794 PH/40 Import PCRE release 6.7 (fixes some bugs).
2796 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2798 PH/42 Give an error if -q is specified more than once.
2800 PH/43 Renamed the variables $interface_address and $interface_port as
2801 $received_ip_address and $received_port, to make it clear that these
2802 values apply to message reception, and not to the outgoing interface when
2803 a message is delivered. (The old names remain recognized, of course.)
2805 PH/44 There was no timeout on the connect() call when using a Unix domain
2806 socket in the ${readsocket expansion. There now is.
2808 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2809 be meaningful with "accept".
2811 SC/01 Eximstats V1.43
2812 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2814 SC/02 Eximstats V1.44
2815 Use a glob alias rather than an array ref in the generated
2816 parser. This improves both readability and performance.
2818 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2819 Collect SpamAssassin and rejection statistics.
2820 Don't display local sender or destination tables unless
2821 there is data to show.
2822 Added average volumes into the top table text output.
2824 SC/04 Eximstats V1.46
2825 Collect data on the number of addresses (recipients)
2826 as well as the number of messages.
2828 SC/05 Eximstats V1.47
2829 Added 'Message too big' to the list of mail rejection
2830 reasons (thanks to Marco Gaiarin).
2832 SC/06 Eximstats V1.48
2833 Mainlog lines which have GMT offsets and are too short to
2834 have a flag are now skipped.
2836 SC/07 Eximstats V1.49 (Alain Williams)
2837 Added the -emptyok flag.
2839 SC/08 Eximstats V1.50
2840 Fixes for obtaining the IP address from reject messages.
2842 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2843 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2844 whitespace changes from 4.64-PH/27
2846 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2849 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2850 are found, allow negative numbers in numeric criteria)
2852 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2854 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2855 to match changes made in 4.64-PH/43
2857 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2859 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2860 "domain" is actually an IP address, and force a failure. This locks out
2861 those revolvers/nameservers that support "A-for-A" lookups, in
2862 contravention of the specifications.
2864 PH/48 When a host name was looked up from an IP address, and the subsequent
2865 forward lookup of the name timed out, the host name was left in
2866 $sender_host_name, contrary to the specification.
2868 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2869 restricted to single-key lookups, Exim was not diagnosing an error if
2870 * or *@ was used with a query-style lookup.
2872 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2874 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2875 long ago, but noone interested enough thought of it. Let's just say that
2876 the "1.1" means that there are some new functions that weren't there at
2877 some point in the past.
2879 PH/51 Error processing for expansion failure of helo_data from an smtp
2880 transport during callout processing was broken.
2882 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2883 tested/used via the -bh/-bhc/-bs options.
2885 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2886 bug, fixed in subsequent PCRE releases).
2888 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2889 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2891 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2897 SC/01 Use a glob alias rather than an array ref in eximstats generated
2898 parser. This improves both readability and performance.
2900 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2901 Don't display local sender or destination tables in eximstats unless
2902 there is data to show.
2903 Added average volumes into the eximstats top table text output.
2905 SC/03 Collect data on the number of addresses (recipients) as well
2906 as the number of messages in eximstats.
2908 TF/01 Correct an error in the documentation for the redirect router. Exim
2909 does (usually) call initgroups() when daemonizing.
2911 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2912 with consistent privilege compared to when running as a daemon.
2914 TF/03 Note in the spec that $authenticated_id is not set for local
2915 submissions from trusted users.
2917 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2918 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2920 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2921 by adding some example configuration directives to the default
2922 configuration file. A little bit of work is required to uncomment the
2923 directives and define how usernames and passwords are checked, but
2924 there is now a framework to start from.
2926 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2927 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2928 without this. I don't know how relevant this is to other LDAP libraries.
2930 PH/02 Add the verb name to the "unknown ACL verb" error.
2932 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2934 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2936 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2937 directory not expanded when it should be if an expanded home directory
2938 was set for the address (which is overridden by the transport).
2940 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2943 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2944 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2945 because it is too late at that time, and has no effect.
2947 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2948 security issue with \' (bugzilla #107). I could not use the
2949 PQescapeStringConn() function, because it needs a PGconn value as one of
2952 PH/08 When testing addresses using -bt, indicate those final addresses that
2953 are duplicates that would not cause an additional delivery. At least one
2954 person was confused, thinking that -bt output corresponded to deliveries.
2955 (Suppressing duplicates isn't a good idea as you lose the information
2956 about possibly different redirections that led to the duplicates.)
2958 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2959 systems where poll() doesn't work, in particular OS X.
2961 PH/10 Added more information to debugging output for retry time not reached.
2963 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2964 operations in malware.c.
2966 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2969 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2970 syslog_duplication set false, log lines that would normally be written
2971 both the the main log and to the reject log were not written to syslog at
2974 PH/14 In the default configuration, change the use of "message" in ACL warn
2975 statements to "add_header".
2977 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2978 not followed by a command (e.g. "seen endif").
2980 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2981 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2984 PH/17 Added extra conditions to the default value of delay_warning_condition
2988 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2989 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2990 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2993 The Auto-Submitted: and various List- headers are standardised, whereas I
2994 don't think Precedence: ever was.
2996 PH/18 Refactored debugging code in route_finduser() to show more information,
2997 in particular, the error code if getpwnam() issues one.
2999 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3000 This is apparently needed in addition to the PH/07 change above to avoid
3001 any possible encoding problems.
3003 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3004 but not after initializing Perl.
3006 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3007 output them only if debugging. By default they are written stderr,
3008 apparently, which is not desirable.
3010 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3013 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3016 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3018 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3019 authenticated or an ident call has been made. Suppress the default
3020 values for $authenticated_id and $authenticated_sender (but permit -oMai
3021 and -oMas) when testing with -bh.
3023 PH/24 Re-jigged the order of the tests in the default configuration so that the
3024 tests for valid domains and recipients precede the DNS black list and CSA
3025 tests, on the grounds that those ones are more expensive.
3027 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3028 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3029 This bug exists in every version of Exim that I still have, right back to
3032 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3033 However, an attempt to turn on case-sensitivity in a regex key by
3034 including (?-i) didn't work because the subject string was already
3035 lowercased, and the effects were non-intuitive. It turns out that a
3036 one-line patch can be used to allow (?-i) to work as expected.
3042 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3043 other effects) broke the use of negated acl sub-conditions.
3045 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3048 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3049 "Deny" causes Exim to reject the incoming connection with a 554 error.
3050 Unfortunately, if there is a major crisis, such as a disk failure,
3051 tcp-wrappers gives "deny", whereas what one would like would be some
3052 kind of temporary error. A kludge has been added to help with this.
3053 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3054 554 error is used if errno is still zero or contains ENOENT (which occurs
3055 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3058 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3060 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3061 errors. Otherwise a message that provokes a temporary error (when other
3062 messages do not) can cause a whole host to time out.
3064 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3065 addresses were routed directly to files or pipes from a redirect router.
3066 File deliveries just didn't batch; pipe deliveries might have suffered
3069 PH/06 A failure to get a lock for a hints database would erroneously always say
3070 "Failed to get write lock", even when it was really a read lock.
3072 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3073 of 0600. This has been changed to use the value of the lockfile_mode
3074 option (which defaults to 0600).
3076 PH/08 Applied small patch from the Sieve maintainer.
3078 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3079 folder from quota calculations, a direct delivery into this folder messed
3080 up the contents of the maildirsize file. This was because the regex was
3081 used only to exclude .Trash (or whatever) when the size of the mailbox
3082 was calculated. There was no check that a delivery was happening into an
3083 excluded directory. This bug has been fixed by ignoring all quota
3084 processing for deliveries into excluded directories.
3086 PH/10 Added the maildirfolder_create_regex option to appendfile.
3092 PH/01 The code for finding all the local interface addresses on a FreeBSD
3093 system running IPv6 was broken. This may well have applied to all BSD
3094 systems, as well as to others that have similar system calls. The broken
3095 code found IPv4 interfaces correctly, but gave incorrect values for the
3096 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3097 that it would not match correctly against @[] and not recognize the IPv6
3100 PH/02 The ipliteral router was not recognizing addresses of the form user@
3101 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3103 PH/03 Added disable_ipv6.
3105 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3106 decoded form, because it is most often used to construct To: headers
3107 lines in autoreplies, and the decoded form may well be syntactically
3108 invalid. However, $reply_address has leading white space removed, and all
3109 newlines turned into spaces so that the autoreply transport does not
3112 PH/05 If group was specified without a user on a router, and no group or user
3113 was specified on a transport, the group from the router was ignored.
3115 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3116 for visible compile-time settings that can be used to change these
3117 numbers, for those that want even more. Backwards compatibility with old
3118 spool files has been maintained. However, going back to a previous Exim
3119 release will lost any variables that are in spool files.
3121 PH/07 Two small changes when running in the test harness: increase delay when
3122 passing a TCP/IP connection to a new process, in case the original
3123 process has to generate a bounce, and remove special handling of
3124 127.0.0.2 (sic), which is no longer necessary.
3126 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3127 be the same on different OS.
3129 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3132 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3133 whether --show-vars was specified or not
3135 JJ/02 exipick: Added support for new ACL variable spool format introduced
3138 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3139 syntactically invalid From: or Reply-to: line, and a filter used this to
3140 generate an autoreply, and therefore failed to obtain an address for the
3141 autoreply, Exim could try to deliver to a non-existent relative file
3142 name, causing unrelated and misleading errors. What now happens is that
3143 it logs this as a hard delivery error, but does not attempt to create a
3146 PH/11 The exinext utility has a -C option for testing purposes, but although
3147 the given file was scanned by exinext itself; it wasn't being passed on
3148 when Exim was called.
3150 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3151 an end-of-file indication when reading a command response.
3153 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3154 compiled. In many other places in Exim, IPv6 addresses are always
3155 recognized, so I have changed this. It also means that IPv4 domain
3156 literals of the form [IPV4:n.n.n.n] are now always recognized.
3158 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3159 used if the router is not running as root, for example, when verifying at
3160 ACL time, or when using -bh. The debugging output from this situation was
3161 non-existent - all you got was a failure to exec. I have made two
3164 (a) Failures to set uid/gid, the current directory, or a process leader
3165 in a subprocess such as that created by queryprogram now generate
3166 suitable debugging output when -d is set.
3168 (b) The queryprogram router detects when it is not running as root,
3169 outputs suitable debugging information if -d is set, and then runs
3170 the subprocess without attempting to change uid/gid.
3172 PH/15 Minor change to Makefile for building test_host (undocumented testing
3175 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3176 additional section of a DNS packet that returns MX or SRV records.
3177 Instead, it always explicitly searches for A/AAAA records. This avoids
3178 major problems that occur when a DNS server includes only records of one
3179 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3180 fixed another bug: if SRV records were looked up and the corresponding
3181 address records were *not* found in the additional section, the port
3182 values from the SRV records were lost.
3184 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3185 using the correct key (the original address) when searching the retry
3186 rules in order to find which one to use for generating the retry hint.
3188 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3189 adding the default one. Similarly, if it contains a Reply-To: header, the
3190 errors_reply_to option, if set, is not used.
3192 PH/19 When calculating a retry time, Exim used to measure the "time since
3193 failure" by looking at the "first failed" field in the retry record. Now
3194 it does not use this if it is later than than the arrival time of the
3195 message. Instead it uses the arrival time. This makes for better
3196 behaviour in cases where some deliveries succeed, thus re-setting the
3197 "first failed" field. An example is a quota failure for a huge message
3198 when small messages continue to be delivered. Without this change, the
3199 "time since failure" will always be short, possible causing more frequent
3200 delivery attempts for the huge message than are intended.
3201 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3203 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3204 $1, $2, $3) because the numerical variables can be reset during some
3205 expansion items (e.g. "match"), thereby losing the authentication data.
3207 PH/21 Make -bV show the size of off_t variables so that the test suite can
3208 decide whether to run tests for quotas > 2G.
3210 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3211 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3212 filecount value is greater than 2G or if a quota value is greater than 2G
3213 on a system where the size of off_t is not greater than 4, a panic error
3216 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3217 never match. The debug and -bh output now contains an explicit error
3218 message indicating a malformed IPv4 address or mask.
3220 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3221 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3222 PH/23 above applies.
3224 PH/25 Do not write to syslog when running in the test harness. The only
3225 occasion when this arises is a failure to open the main or panic logs
3226 (for which there is an explicit test).
3228 PH/26 Added the /no_tell option to "control=freeze".
3230 PH/27 If a host name lookup failed very early in a connection, for example, if
3231 the IP address matched host_lookup and the reverse lookup yielded a name
3232 that did not have a forward lookup, an error message of the form "no IP
3233 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3234 could be logged. Now it outputs the IP address instead of "NULL".
3236 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3237 allows the sender and the authenticated sender to be set when
3238 submitting a message from within Exim. Since child_open_exim() is
3239 documented for local_scan(), the new function should be too.
3241 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3242 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3243 results in an empty string is now treated as unset.
3245 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3247 PH/31 Added sender_verify_fail logging option.
3249 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3250 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3251 tidied the source and removed it altogether.
3253 PH/33 When a queue run was abandoned because the load average was too high, a
3254 log line was always written; now it is written only if the queue_run log
3255 selector is set. In addition, the log line for abandonment now contains
3256 information about the queue run such as the pid. This is always present
3257 in "start" and "stop" lines but was omitted from the "abandon" line.
3259 PH/34 Omit spaces between a header name and the colon in the error message that
3260 is given when verify = headers_syntax fails (if there are lots of them,
3261 the message gets confusing).
3263 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3264 names, as there are now some PTR records that contain slashes. This check
3265 is only to protect against broken name servers that fall over on strange
3266 characters, so the fact that it applies to all lookups doesn't matter.
3268 PH/36 Now that the new test suite is complete, we can remove some of the
3269 special code in Exim that was needed for the old test suite. For example,
3270 sorting DNS records because real resolvers return them in an arbitrary
3271 order. The new test suite's fake resolver always returns records in the
3274 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3275 bounces) except when queue_only is set, to avoid logging races between
3276 the different processes.
3278 PH/38 Panic-die if .include specifies a non-absolute path.
3280 PH/39 A tweak to the "H" retry rule from its user.
3282 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3283 a label. They prevented compilation on older perls.
3285 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3286 a warning to be raised on newish perls.
3288 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3289 on queue. Changes to match documented behaviour of showing count of
3290 messages matching specified criteria.
3292 PH/40 Changed the default ident timeout from 30s to 5s.
3294 PH/41 Added support for the use of login_cap features, on those BSD systems
3295 that have them, for controlling the resources used by pipe deliveries.
3297 PH/42 The content-scanning code uses fopen() to create files in which to put
3298 message data. Previously it was not paying any attention to the mode of
3299 the files. Exim runs with umask(0) because the rest of the code creates
3300 files with open(), and sets the required mode explicitly. Thus, these
3301 files were ending up world-writeable. This was not a big issue, because,
3302 being within the spool directory, they were not world-accessible. I have
3303 created a function called modefopen, which takes an additional mode
3304 argument. It sets umask(777), creates the file, chmods it to the required
3305 mode, then resets the umask. All the relevant calls to fopen() in the
3306 content scanning code have been changed to use this function.
3308 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3309 to 24 hours. This avoids potential overflow problems when processing G
3310 and H retry rules. I suspect nobody ever tinkers with this value.
3312 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3314 PH/45 When the plaintext authenticator is running as a client, the server's
3315 challenges are checked to ensure they are valid base64 strings. By
3316 default, the authentication attempt is cancelled if an invalid string is
3317 received. Setting client_ignore_invalid_base64 true ignores these errors.
3318 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3319 they are received. Thus, the responses can be made to depend on the
3320 challenges. If an invalid string is ignored, an empty string is placed in
3323 PH/46 Messages that are created by the autoreply transport now contains a
3324 References: header, in accordance with RFCs 2822 and 3834.
3326 PH/47 Added authenticated_sender_force to the smtp transport.
3328 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3330 PH/49 Installed latest patch from the Sieve maintainer.
3332 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3333 was also set, the appendfile transport was unnecessarily scanning a
3334 directory of message files (e.g. for maildir delivery) to find the count
3335 of files (along with the size), even though it did not need this
3336 information. It now does the scan only if it needs to find either the
3337 size of the count of files.
3339 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3341 PH/52 Two bugs concerned with error handling when the smtp transport is
3344 (i) Exim was not creating retry information for temporary errors given
3345 for individual recipients after the DATA command when the smtp transport
3346 was used in LMTP mode. This meant that they could be retried too
3347 frequently, and not timed out correctly.
3349 (ii) Exim was setting the flag that allows error details to be returned
3350 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3351 recipients that were returned after the DATA command.
3353 PH/53 This is related to PH/52, but is more general: for any failing address,
3354 when detailed error information was permitted to be returned to the
3355 sender, but the error was temporary, then after the final timeout, only
3356 "retry timeout exceeded" was returned. Now it returns the full error as
3357 well as "retry timeout exceeded".
3359 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3360 do this, and (what is worse) MTAs that accept it.
3362 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3363 will now be deprecated.
3365 PH/56 New os.c-cygwin from the Cygwin maintainer.
3367 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3368 formats (previously only available in exim formats via -bpr, -bpru,
3369 and -bpra. Now also available in native and exiqgrep formats)
3371 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3372 with very large, slow to parse queues
3374 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3376 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3378 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3379 responses to authentication challenges, though it was showing the
3380 challenges; (ii) I've removed the CR characters from the debug output for
3383 PH/58 Allow for the insertion of a newline as well as a space when a string
3384 is turned into more than one encoded-word during RFC 2047 encoding. The
3385 Sieve code now uses this.
3387 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3388 data_4xx, lost_connection, tls_required.
3390 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3391 message was being sent as an SMTP response.
3393 PH/61 Add -l and -k options to exicyclog.
3395 PH/62 When verifying, if an address was redirected to one new address, so that
3396 verification continued, and the new address failed or deferred after
3397 having set something in $address_data, the value of $address_data was not
3398 passed back to the ACL. This was different to the case when no
3399 redirection occurred. The value is now passed back in both cases.
3401 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3402 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3403 use login_cap.h, so on its own it isn't the distinguishing feature. The
3404 new name refers directly to the setclassresources() function.
3406 PH/65 Added configuration files for NetBSD3.
3408 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3410 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3411 is preferred over IPv4.
3413 PH/68 The bounce_return_message and bounce_return_body options were not being
3414 honoured for bounces generated during the reception of non-SMTP messages.
3415 In particular, this applied to messages rejected by the ACL. This bug has
3416 been fixed. However, if bounce_return_message is true and bounce_return_
3417 body is false, the headers that are returned for a non-SMTP message
3418 include only those that have been read before the error was detected.
3419 (In the case of an ACL rejection, they have all been read.)
3421 PH/69 The HTML version of the specification is now built in a directory called
3422 spec_html instead of spec.html, because the latter looks like a path with
3423 a MIME-type, and this confuses some software.
3425 PH/70 Catch two compiler warnings in sieve.c.
3427 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3428 function verify_get_ident() calls ip_connect() to connect a socket, but
3429 if the "connect()" function timed out, ip_connect() used to close the
3430 socket. However, verify_get_ident() also closes the socket later, and in
3431 between Exim writes to the log, which may get opened at this point. When
3432 the socket was closed in ip_connect(), the log could get the same file
3433 descriptor number as the socket. This naturally causes chaos. The fix is
3434 not to close the socket in ip_connect(); the socket should be closed by
3435 the function that creates it. There was only one place in the code where
3436 this was missing, in the iplookup router, which I don't think anybody now
3437 uses, but I've fixed it anyway.
3439 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3440 well as to direct DNS lookups. Otherwise the handling of names in host
3441 lists is inconsistent and therefore confusing.
3447 PH/01 Two changes to the default runtime configuration:
3449 (1) Move the checks for relay_from_hosts and authenticated clients from
3450 after to before the (commented out) DNS black list checks.
3452 (2) Add control=submission to the relay_from_hosts and authenticated
3453 clients checks, on the grounds that messages accepted by these
3454 statements are most likely to be submissions.
3456 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3458 (1) Generate an error if the third argument for the ${prvs expansion is
3461 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3464 (3) Reset the variables that are obtained from the first argument of
3465 ${prvscheck and used in the second argument before leaving the code,
3466 because their memory is reclaimed, so using them afterwards may do
3469 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3470 one (it's much easier than Tom thought :-).
3472 (5) Because of (4), we can now allow for the use of $prvscheck_result
3473 inside the third argument.
3475 PH/03 For some reason, the default setting of PATH when running a command from
3476 a pipe transport was just "/usr/bin". I have changed it to
3479 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3480 anything to be listed in the output from -bV.
3482 PH/05 When a filter generated an autoreply, the entire To: header line was
3483 quoted in the delivery log line, like this:
3485 => >A.N.Other <ano@some.domain> <original@ddress> ...
3487 This has been changed so that it extracts the operative address. There
3488 may be more than one such address. If so, they are comma-separated, like
3491 => >ano@some.domain,ona@other.domain <original@ddress> ...
3493 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3494 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3495 not being looked up in the rDNS to get a host name, Exim was showing the
3496 IP address twice in Received: lines, even though the IP addresses were
3497 identical. For example:
3499 Received: from [1.2.3.4] (helo=[1.2.3.4])
3501 However, if the real host name was known, it was omitting the HELO data
3502 if it matched the actual IP address. This has been tidied up so that it
3503 doesn't show the same IP address twice.
3505 PH/07 When both +timestamp and +memory debugging was on, the value given by
3506 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3507 called by the debug printing, thereby overwriting the timestamp buffer.
3508 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3510 PH/08 When the original message was included in an autoreply transport, it
3511 always said "this is a copy of the message, including all the headers",
3512 even if body_only or headers_only was set. It now gives an appropriate
3515 PH/09 Applied a patch from the Sieve maintainer which:
3517 o fixes some comments
3518 o adds the (disabled) notify extension core
3519 o adds some debug output for the result of if/elsif tests
3520 o points to the current vacation draft in the documentation
3521 and documents the missing references header update
3525 o fixes a bug in processing the envelope test (when testing
3526 multiple envelope elements, the last element determined the
3529 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3530 Electronic Mail") by including:
3532 Auto-submitted: auto-generated
3534 in the messages that it generates (bounce messages and others, such as
3535 warnings). In the case of bounce messages for non-SMTP messages, there was
3536 also a typo: it was using "Auto_submitted" (underscore instead of
3537 hyphen). Since every message generated by Exim is necessarily in response
3538 to another message, thes have all been changed to:
3540 Auto-Submitted: auto-replied
3542 in accordance with these statements in the RFC:
3544 The auto-replied keyword:
3546 - SHOULD be used on messages sent in direct response to another
3547 message by an automatic process,
3549 - MUST NOT be used on manually-generated messages,
3551 - MAY be used on Delivery Status Notifications (DSNs) and Message
3552 Disposition Notifications (MDNs),
3554 - MUST NOT be used on messages generated by automatic or periodic
3555 processes, except for messages which are automatic responses to
3558 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3559 to the default Received: header definition.
3561 PH/12 Added log selector acl_warn_skipped (default on).
3563 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3564 variables because (a) they are in the wrong storage pool and (b) even if
3565 they were copied, it wouldn't work properly because of the caching.
3567 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3568 checking when decoding. Apparently there are clients that generate
3569 overlong encoded strings. Why am I not surprised?
3571 PH/15 If the first argument of "${if match_address" was not empty, but did not
3572 contain an "@" character, Exim crashed. Now it writes a panic log message
3573 and treats the condition as false.
3575 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3577 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3578 extension "envelope-auth". The code is finished and in agreement with
3579 other implementations, but there is no documentation so far and in fact,
3580 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3581 not changing the active code.
3583 Print executed "if" and "elsif" statements when debugging is used. This
3584 helps a great deal to understand what a filter does.
3586 Document more things not specified clearly in RFC3028. I had all this
3587 sorted out, when out of a sudden new issues came to my mind. Oops."
3589 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3592 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3593 canonical form (as documented). However, after a host name lookup from
3594 the IP address, check_host() was doing a simple string comparison with
3595 addresses acquired from the DNS when checking that the found name did
3596 have the original IP as one of its addresses. Since any found IPv6
3597 addresses are likely to be in abbreviated form, the comparison could
3598 fail. Luckily, there already exists a function for doing the comparison
3599 by converting both addresses to binary, so now that is used instead of
3600 the text comparison.
3602 PH/20 There was another similar case to PH/19, when a complete host name was
3603 given in a host list; looking up its IP address could give an abbreviated
3604 form, whereas the current host's name might or might not be abbreviated.
3605 The same fix has been applied.
3611 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3612 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3615 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3616 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3618 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3620 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3621 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3622 Also, for "Auto-Submitted", treat anything other than "no" as
3623 non-personal, in accordance with RFC 3834. (Previously it treated
3624 anything starting "auto-" as non-personal.)
3626 TF/01 The control=submission/name=... option had a problem with syntax
3627 errors if the name included a slash character. The /name= option
3628 now slurps the rest of the string, so it can include any characters
3629 but it must come last in the list of options (after /sender_retain
3632 PH/05 Some modifications to the interface to the fake nameserver for the new
3640 TK/01 Added the "success_on_redirect" address verification option. See
3641 NewStuff for rationale and an example.
3643 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3645 PH/02 Patch to exigrep to allow it to work on syslog lines.
3647 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3648 fread() to skip over the body file's header line, because in Cygwin the
3649 header line is locked and is inaccessible.
3651 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3652 co-exist for some time) to make it clear that it is the Exim ID that is
3653 referenced, not the Message-ID: header line.
3655 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3656 string_format() function, because snprintf() does not exist on all
3659 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3660 ${stat: expansion item.
3662 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3663 protocol synchronization error", to keep the pedants happy.
3665 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3666 well as for IRIX systems, when gcc is being used. See the host.c source
3669 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3671 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3674 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3675 transport and to the smtp transport in LMTP mode.
3677 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3679 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3680 run a filter in a subprocess. This could lead to confusion in subsequent
3681 lookups in the parent process. There should also be a search_tidyup() at
3682 the end of the subprocess.
3684 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3685 only if the host matched helo_try_verify_hosts, which caused the
3686 verification to occur when the EHLO/HELO command was issued. The ACL just
3687 tested the remembered result. Now, if a previous verification attempt has
3688 not happened, "verify = helo" does it there and then.
3690 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3692 TK/03 Fix log output including CR from clamd.
3694 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3695 exist provoked a memory error which could cause a segfault.
3697 PH/15 Installed PCRE 6.2
3699 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3701 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3702 of the problem. Specifically, suggested +O2 rather than +O1 for the
3705 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3707 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3708 record, the port was not correctly logged when the outgoing_port log
3709 selector was set (it logged the transort's default port).
3711 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3712 fallback_hosts, and "hosts" in the smtp transport.
3714 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3715 host errors such as "Connection refused".
3717 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3718 authentication with radiusclient 0.4.9:
3720 - Error returned from rc_read_config was caught wrongly
3721 - Username/password not passed on to radius server due to wrong length.
3723 The presumption is that some radiusclient API changes for 4.51/PH/17
3724 were not taken care of correctly. The code is still untested by me (my
3725 Linux distribution still has 0.3.2 of radiusclient), but it was
3726 contributed by a Radius user.
3728 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3729 expanding the "port" option of the smtp transport.
3731 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3732 while reading a MIME header. Thanks to Tom Hughes for a patch.
3734 PH/24 Include config.h inside local_scan.h so that configuration settings are
3737 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3738 that in an ACL for RCPT (for example), you can examine exactly what was
3741 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3742 commands, but it was not correctly comparing the address with the actual
3743 client host address. Thus, it would show the EHLO address in Received:
3744 header lines when this was not necessary.
3746 PH/27 Added the % operator to ${eval:}.
3748 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3749 it should be ignoring failures (because with -C, for example, it has lost
3750 privilege). It wasn't ignoring creation failures other than "already
3753 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3756 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3757 because an input line was too long, either on its own, or by virtue of
3758 too many continuations, the temporary file was not being removed, and the
3759 return code was incorrect.
3761 PH/31 Missing "BOOL" in function definition in filtertest.c.
3763 PH/32 Applied Sieve patches from the maintainer.
3765 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3767 PH/33 Added "verify = not_blind".
3769 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3770 Local/Makefile (with some defaults set). These are used in built scripts
3771 such as exicyclog, but they have never been used in the exim_install
3772 script (though there are many overriding facilities there). I have
3773 arranged that the exim_install script now takes note of these two
3776 PH/35 Installed configuration files for Dragonfly.
3778 PH/36 When a locally submitted message by a trusted user did not contain a
3779 From: header, and the sender address was obtained from -f or from an SMTP
3780 MAIL command, and the trusted user did not use -F to supply a sender
3781 name, $originator_name was incorrectly used when constructing a From:
3782 header. Furthermore, $originator_name was used for submission mode
3783 messages from external hosts without From: headers in a similar way,
3784 which is clearly wrong.
3786 PH/37 Added control=suppress_local_fixups.
3788 PH/38 When log_selector = +received_sender was set, and the addition of the
3789 sender made the log line's construction buffer exactly full, or one byte
3790 less than full, an overflow happened when the terminating "\n" was
3793 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3794 when the result of a list match is failure because a DNS lookup failed.
3796 PH/40 RM_COMMAND is now used in the building process.
3798 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3799 the "build-* directories that it finds.
3801 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3802 address in a domain literal was a prefix of an interface address.
3804 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3805 when verifying a sender address, unless rewrite_headers is false.
3807 PH/44 Wrote a long comment about why errors_to addresses are verified as
3808 recipients, not senders.
3810 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3811 the ratelimit ACL was added.
3813 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3815 PH/46 Added extra information about PostgreSQL errors to the error string.
3817 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3818 avoiding the need to install special zones in a real server. This is
3819 backwards compatible; if it can't find the fake resolver, it drops back.
3820 Thus, both old and new test suites can be run.
3822 TF/02 Added util/ratelimit.pl
3824 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3825 clock is set back in time.
3827 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3828 Candler <B.Candler@pobox.com>.
3830 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3831 OK for addresses that are the result of redirections.
3833 PH/48 A number of further additions for the benefit of the new test suite,
3834 including a fake gethostbyname() that interfaces to the fake DNS resolver
3837 TF/06 The fix for widen_domains has also been applied to qualify_single and
3838 search_parents which are the other dnslookup options that can cause
3841 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3844 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3846 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3847 service" if it encountered an MX record with an empty target hostname.
3848 The message is now "an MX or SRV record indicated no SMTP service".
3850 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3851 if the DNS of the sending site is misconfigured. This is quite a
3852 common situation. This change restores the behaviour of treating a
3853 helo verification defer as a failure.
3855 PH/51 If self=fail was set on a router, the bounce message did not include the
3856 actual error message.
3862 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3864 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3865 command itself ended in error, the underlying message about the transport
3866 filter timeout was being overwritten with the pipe command error. Now the
3867 underlying error message should be appended to the second error message.
3869 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3871 PH/02 Reduce the amount of output that "make" produces by default. Full output
3872 can still be requested.
3874 PH/03 The warning log line about a condition test deferring for a "warn" verb
3875 was being output only once per connection, rather than after each
3876 occurrence (because it was using the same function as for successful
3877 "warn" verbs). This seems wrong, so I have changed it.
3879 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3880 it should not have, which might have caused a crash in the right
3881 circumstances, but probably never did.
3883 PH/04 Installed a modified version of Tony Finch's patch to make submission
3884 mode fix the return path as well as the Sender: header line, and to
3885 add a /name= option so that you can make the user's friendly name appear
3888 TF/03 Added the control = fakedefer ACL modifier.
3890 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3891 Mark Lowes for thorough testing.
3893 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3895 TK/03 Merged latest SRS patch from Miles Wilton.
3897 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3898 with the definition in sysexits.h (which is #included earlier).
3899 Fortunately, Exim does not actually use EX_OK. The code used to try to
3900 preserve the sysexits.h value, by assuming that macro definitions were
3901 scanned for macro replacements. I have been disabused of this notion,
3902 so now the code just undefines EX_OK before #including unistd.h.
3904 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3905 in the smtp transport. When a block could not be written in a single
3906 write() function, the timeout was being re-applied to each part-write.
3907 This seems wrong - if the receiver was accepting one byte at a time it
3908 would take for ever. The timeout is now adjusted when this happens. It
3909 doesn't have to be particularly precise.
3911 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3912 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3914 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3915 without a domain if the check to <postmaster@domain> fails.
3917 SC/01 Eximstats: added -xls and the ability to specify output files
3918 (patch written by Frank Heydlauf).
3920 SC/02 Eximstats: use FileHandles for outputting results.
3922 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3924 SC/04 Eximstats: fixed display of large numbers with -nvr option
3926 SC/05 Eximstats: fixed merging of reports with empty tables.
3928 SC/06 Eximstats: added the -include_original_destination flag
3930 SC/07 Eximstats: removed tabs and trailing whitespace.
3932 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3934 TK/06 MBOX spool code: Add real "From " MBOX separator line
3935 so the .eml file is really in mbox format (even though
3936 most programs do not really care). Patch from Alex Miller.
3938 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3939 The latter is generated from $received_to and is only set if the
3940 message has one envelope recipient. SA can use these headers,
3941 obviously out-of-the-box. Patch from Alex Miller.
3943 PH/08 The ${def test on a variable was returning false if the variable's
3944 value was "0", contrary to what the specification has always said!
3945 The result should be true unless the variable is empty.
3947 PH/09 The syntax error of a character other than { following "${if
3948 def:variable_name" (after optional whitespace) was not being diagnosed.
3949 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3950 accidental colon was present, for example, could give incorrect results.
3952 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3953 result is used (not including appendfile, where other changes are about
3956 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3957 This involved changing a lot of size variables from int to off_t. It
3958 should work with maildirs and everything.
3960 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3961 spamd dying while we are connected to it.
3963 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3966 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3967 function for BDB 4.3.
3969 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3971 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3972 the sources for PCRE have been re-arranged and more files are now
3975 PH/15 The code I had for printing potentially long long variables in PH/11
3976 above was not the best (it lost precision). The length of off_t variables
3977 is now inspected at build time, and an appropriate printing format (%ld
3978 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3979 to be "long long int" or "long int". This is needed for the internal
3980 formatting function string_vformat().
3982 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3983 the configuration file to be ":syslog", then the script "guesses" where
3984 the logs files are, rather than using the compiled in default. In our
3985 case the guess is not the same as the compiled default, so the script
3986 suddenly stopped working when I started to use syslog. The patch checks
3987 to see if log_file_path is "". If so, it attempts to read it from exim
3988 with no configuration file to get the compiled in version, before it
3989 falls back to the previous guessing code."
3991 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3992 implementing BATV in an Exim configuration. See NewStuff for the gory
3995 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3996 Makefile that are specific to HP-UX.
3998 PH/18 If the "use_postmaster" option was set for a recipient callout together
3999 with the "random" option, the postmaster address was used as the MAIL
4000 FROM address for the random test, but not for the subsequent recipient
4001 test. It is now used for both.
4003 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4004 patch removes a few documentation additions to RFC 3028, because the
4005 latest draft now contains them. It adds the new en;ascii-case comparator
4006 and a new error check for 8bit text in MIME parts. Comparator and
4007 require names are now matched exactly. I enabled the subaddress
4008 extension, but it is not well tested yet (read: it works for me)."
4010 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4011 rework some of the code of TK/09 above to avoid the hardwired use of
4012 "%lld" and "long long". Replaced the call to snprintf() with a call to
4015 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4016 records point to non-existent hosts", "retry timeout exceeded", and
4017 "retry time not reached for any host after a long failure period".
4019 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4020 experimental DomainKeys support:
4022 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4023 (2) On an error such as an illegally used "control", the wrong name for
4024 the control was given.
4026 These problems did NOT occur unless DomainKeys support was compiled.
4028 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4030 PH/24 Added ${if match_ip condition.
4032 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4033 databases so that it will be absolutely obvious if a crash occurs in the
4034 DB library. This is a regular occurrence (often caused by mis-matched
4037 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4038 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4039 picked up on a user's system that detects such things. There doesn't seem
4040 to be a gcc warning option for this - only an attribute that has to be
4041 put on the function's prototype. It seems that in Fedora Core 4 they have
4042 set this on a number of new functions. No doubt there will be more in due
4045 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4046 specify a transport. However, if an address that was verified by such a
4047 router was the subject of a callout, Exim crashed because it tried to
4048 read the rcpt_include_affixes from the non-existent transport. Now it
4049 just assumes that the setting of that option is false. This bug was
4050 introduced by 4.51/PH/31.
4052 PH/28 Changed -d+all to exclude +memory, because that information is very
4053 rarely of interest, but it makes the output a lot bigger. People tend to
4054 do -d+all out of habit.
4056 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4057 code in os-type was giving problems when libc.so lives in lib64, like on
4060 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4061 aren't the modern standard, and it seems that some systems' include files
4062 don't always have them. Exim was already checking for some of the newer
4063 ones like T_AAAA, and defining it itself. I've added checks for all the
4064 record types that Exim uses.
4066 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4067 not automatically generating a new one, as it is supposed to. This
4068 prevented TLS from working. If the file did exist, but contained invalid
4069 data, a new version was generated, as expected. It was only the case of a
4070 non-existent file that was broken.
4072 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4073 with a change in libdomainkeys > 0.64.
4075 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4076 from DNS. If the selector record carries the flag, it now has
4077 precedence over the domain-wide flag.
4079 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4081 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4082 the use of a port name that isn't defined in /etc/services) occurred, the
4083 message was deferred as in a normal delivery, and thus remained on the
4084 spool, instead of being failed because of the mua_wrapper setting. This
4085 is now fixed, and I tidied up some of the mua_wrapper messages at the
4088 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4089 the messages in a hash of arrays rather than using individual hashes.
4090 This is a bit cleaner and results in dramatic memory savings, albeit
4091 at a slight CPU cost.
4093 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4094 as requested by Marc Sherman.
4096 SC/10 Eximstats: added histograms for user specified patterns as requested
4099 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4101 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4102 fopen() in the content-scanning modules that did not already have it.
4108 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4109 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4111 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4113 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4115 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4116 to test Sieve filters that use "vacation".
4118 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4119 that changes the way the GnuTLS parameters are stored in the cache file.
4120 The new format can be generated externally. For backward compatibility,
4121 if the data in the cache doesn't make sense, Exim assumes it has read an
4122 old-format file, and it generates new data and writes a new file. This
4123 means that you can't go back to an older release without removing the
4126 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4127 work if there are any delivery delays because "one_time" forces the
4128 parent to be marked "delivered", so its unseen clone is never tried
4129 again. For this reason, Exim now forbids the simultaneous setting of
4132 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4133 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4134 entry if you want to know the details. The fix, however, neglected to
4135 consider the case when local delivery batching is involved. The test for
4136 "previously delivered" was not happening when checking to see if an
4137 address could be batched with a previous (undelivered) one; under
4138 certain circumstances this could lead to multiple deliveries to the same
4141 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4142 in its include files, and this causes problems building Exim.
4144 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4145 header_syntax) but Exim was just ignoring anything given after a slash.
4146 In particular, this caused confusion with an attempt to use "verify =
4147 reverse_host_lookup/defer_ok". An error is now given when options are
4148 supplied for verify items that do not have them. (Maybe reverse_host_
4149 lookup should have a defer_ok option, but that's a different point.)
4151 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4152 defined by RFC 821) to 2048, because there were problems with some AUTH
4153 commands, and RFC 1869 says the size should be increased for extended
4154 SMTP commands that take arguments.
4156 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4159 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4160 "unknown" error; now it says that the functionality isn't in the binary.
4162 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4163 an address' error message when a string expansion fails (syntax or
4164 whatever). Otherwise the password may appear in the log. Following change
4165 PH/42 below, there is no longer a chance of it appearing in a bounce
4168 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4170 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4171 of its addresses was ever tried. (Bugzilla bug #2.)
4173 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4174 the result incorrectly in the debug output. (It correctly added a newline
4175 to what was transported.)
4177 TF/01 Added $received_time.
4179 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4180 commented out examples of how to interface to a virus scanner and to
4181 SpamAssassin. Also added commented examples of av_scanner and
4182 spamd_address settings.
4184 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4185 and controls are allowed in which ACLs. There were a couple of minor
4186 errors. Some of the entries in the conditions table (which is a table of
4187 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4188 negation of where the condition IS allowed.
4190 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4192 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4193 header file does not have a version number, so I've had to invent a new
4194 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4195 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4196 radiusclient), but it was contributed by a Radius user.
4198 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4199 files or directories, for OpenSSL.
4201 PH/19 When an Exim process that is running as root has to create an Exim log
4202 file, it does so in a subprocess that runs as exim:exim so as to get the
4203 ownership right at creation (otherwise, other Exim processes might see
4204 the file with the wrong ownership). There was no test for failure of this
4205 fork() call, which would lead to the process getting stuck as it waited
4206 for a non-existent subprocess. Forks do occasionally fail when resources
4207 run out. I reviewed all the other calls to fork(); they all seem to check
4210 PH/20 When checking for unexpected SMTP input at connect time (before writing
4211 the banner), Exim was not dealing correctly with a non-positive return
4212 from the read() function. If the client had disconnected by this time,
4213 the result was a log entry for a synchronization error with an empty
4214 string after "input=" when read() returned zero. If read() returned -1
4215 (an event I could not check), uninitialized data bytes were printed.
4216 There were reports of junk text (parts of files, etc) appearing after
4219 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4221 PH/22 Added support for macro redefinition, and (re)definition in between
4222 driver and ACL definitions.
4224 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4225 forgetting to use the resulting value; it was using the unexpanded value.
4227 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4228 hadn't been configured. The fix is from Juergen Kreileder, who
4229 understands it better than I do:
4231 "Here's what I see happening with three configured cyrus_sasl
4232 authenticators configured (plain, login, cram-md5):
4234 On startup auth_cyrus_sasl_init() gets called for each of these.
4235 This means three calls to sasl_listmech() without a specified mech_list.
4236 => SASL tests which mechs of all available mechs actually work
4237 => three warnings about OTP not working
4238 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4240 With the patch, sasl_listmech() also gets called three times. But now
4241 SASL's mech_list option is set to the server_mech specified in the the
4242 authenticator. Or in other words, the answer from sasl_listmech()
4243 gets limited to just the mech you're testing for (which is different
4245 => the return list contains just 'plain' or 'login', 'cram-md5' or
4246 nothing depending on the value of ob->server_mech.
4248 I've just tested the patch: Authentication still works fine,
4249 unavailable mechs specified in the exim configuration are still
4250 caught, and the auth.log warnings about OTP are gone."
4252 PH/25 When debugging is enabled, the contents of the command line are added
4253 to the debugging output, even when log_selector=+arguments is not
4256 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4257 answer is "GNU", and only if the return is "GNU/something" is the answer
4260 PH/27 $acl_verify_message is now set immediately after the failure of a
4261 verification in an ACL, and so is available in subsequent modifiers. In
4262 particular, the message can be preserved by coding like this:
4264 warn !verify = sender
4265 set acl_m0 = $acl_verify_message
4267 Previously, $acl_verify_message was set only while expanding "message"
4268 and "log_message" when a very denied access.
4270 PH/28 Modified OS/os.c-Linux with
4272 -#ifndef OS_LOAD_AVERAGE
4273 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4275 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4276 nomenclature these days.)
4278 PH/29 Installed patch from the Sieve maintainer that adds the options
4279 sieve_useraddress and sieve_subaddress to the redirect router.
4281 PH/30 In these circumstances:
4282 . Two addresses routed to the same list of hosts;
4283 . First host does not offer TLS;
4284 . First host accepts first address;
4285 . First host gives temporary error to second address;
4286 . Second host offers TLS and a TLS session is established;
4287 . Second host accepts second address.
4288 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4289 and peerdn, if requested) that were in fact used only for the second
4292 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4293 attention to any local part prefix or suffix that was matched by the
4294 router that accepted the address. It now behaves in the same way as it
4295 does for delivery: the affixes are removed from the local part unless
4296 rcpt_include_affixes is set on the transport.
4298 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4299 timeout during the DATA phase of an incoming message.
4301 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4302 applied a patch sanctioned by the Sieve maintainer.
4304 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4305 the uid or gid is negative. A case of a negative gid caused this to be
4306 noticed. The fix allows for either to be negative.
4308 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4309 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4312 PH/36 The change PH/12 above was broken. Fixed it.
4314 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4315 the grounds that routing the same address twice would always produce the
4316 same answer. This might have been true once, but it is certainly no
4317 longer true now. Routing a child address may depend on the previous
4318 routing that produced that child. Some complicated redirection strategies
4319 went wrong when messages had multiple recipients, and made Exim's
4320 behaviour dependent on the order in which the addresses were given.
4322 I have moved the duplicate checking until after the routing is complete.
4323 Exim scans the addresses that are assigned to local and remote
4324 transports, and removes any duplicates. This means that more work will be
4325 done, as duplicates will always all be routed, but duplicates are
4326 presumably rare, so I don't expect this is of any significance.
4328 For deliveries to pipes, files, and autoreplies, the duplicate checking
4329 still happens during the routing process, since they are not going to be
4332 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4333 It corrects a timeout issue with spamd. This is Ian's comment: "The
4334 background is that sometimes spamd either never reads data from a
4335 connection it has accepted, or it never writes response data. The exiscan
4336 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4337 blindly assumes that writes won't block so it may never time out."
4339 PH/39 Allow G after quota size as well as K and M.
4341 PH/40 The value set for $authenticated_id in an authenticator may not contain
4342 binary zeroes or newlines because the value is written to log lines and
4343 to spool files. There was no check on this. Now the value is run through
4344 the string_printing() function so that such characters are converted to
4345 printable escape sequences.
4347 PH/41 $message_linecount is a new variable that contains the total number of
4348 lines in the message. Compare $body_linecount, which is the count for the
4351 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4352 bounce and delay warning messages, except in certain special cases, which
4355 (a) An SMTP error message from a remote host;
4356 (b) A message specified in a :fail: redirection;
4357 (c) A message specified in a "fail" command in a system filter;
4358 (d) A message specified in a FAIL return from the queryprogram router;
4359 (e) A message specified by the cannot_route_message router option.
4361 In these cases only, Exim does include the error details in bounce and
4362 warning messages. There are also a few cases where bland messages such
4363 as "unrouteable address" or "local delivery error" are given.
4365 PH/43 $value is now also set for the "else" part of a ${run expansion.
4367 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4368 being worked on, but at least Exim now implements the latest version to
4371 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4372 process to complete was treated as a delivery failure, a timeout while
4373 writing the message to the pipe was logged, but erroneously treated as a
4374 successful delivery. Such timeouts include transport filter timeouts. For
4375 consistency with the overall process timeout, these timeouts are now
4376 treated as errors, giving rise to delivery failures by default. However,
4377 there is now a new Boolean option for the pipe transport called
4378 timeout_defer, which, if set TRUE, converts the failures into defers for
4379 both kinds of timeout. A transport filter timeout is now identified in
4382 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4383 systems where "make" and "gmake" are different, calling "gmake" at top
4384 level broke things. I've arranged for the value of $(MAKE) to be passed
4385 from the Makefile to this script so that it can call the same version of
4389 A note about Exim versions 4.44 and 4.50
4390 ----------------------------------------
4392 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4393 changes of various kinds. As a consequence, a big documentation update was
4394 needed. This delayed the release for rather longer than seemed good, especially
4395 in the light of a couple of (minor) security issues. Therefore, the changes
4396 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4397 release. So 4.44 and 4.50 are in effect two different branches that both start
4400 I have left the 4.50 change log unchanged; it contains all the changes since
4401 4.43. The change log for 4.44 is below; many of its items are identical to
4402 those for 4.50. This seems to be the most sensible way to preserve the
4403 historical information.
4409 1. Minor wording change to the doc/README.SIEVE file.
4411 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4412 computation of the current number of files was incorrect.
4414 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4415 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4418 4. Give more explanation in the error message when the command for a transport
4419 filter fails to execute.
4421 5. There are several places where Exim runs a non-Exim command in a
4422 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4423 was being done only for the command run by the queryprogram router. It is
4424 now done for all such subprocesses. The other cases are: ${run, transport
4425 filters, and the commands run by the lmtp and pipe transports.
4427 6. Added CONFIGURE_GROUP build-time option.
4429 7. Some older OS have a limit of 256 on the maximum number of file
4430 descriptors. Exim was using setrlimit() to set 1000 as a large value
4431 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4432 systems. I've change it so that if it can't get 1000, it tries for 256.
4434 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4435 was an oversight, and furthermore, ever since the addition of extra
4436 controls (e.g. 4.43/32), the checks on when to allow different forms of
4437 "control" were broken. There should now be diagnostics for all cases when a
4438 control that does not make sense is encountered.
4440 9. Added the /retain_sender option to "control=submission".
4442 10. $recipients is now available in the predata ACL (oversight).
4444 11. Tidy the search cache before the fork to do a delivery from a message
4445 received from the command line. Otherwise the child will trigger a lookup
4446 failure and thereby defer the delivery if it tries to use (for example) a
4447 cached ldap connection that the parent has called unbind on.
4449 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4450 of $address_data from the recipient verification was clobbered by the
4451 sender verification.
4453 13. The value of address_data from a sender verification is now available in
4454 $sender_address_data in subsequent conditions in the ACL statement.
4456 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4458 15. Added a new option "connect=<time>" to callout options, to set a different
4461 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4462 was its contents. (It was OK if the option was not defined at all.)
4464 17. A "Completed" log line is now written for messages that are removed from
4465 the spool by the -Mrm option.
4467 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4468 information about exactly what failed.
4470 19. Added -dd to debug only the daemon process.
4472 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4473 handles timeouts, both on the server side and network timeouts. Renamed the
4474 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4476 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4477 It is now set to "smtps".
4479 22. $host_address is now set to the target address during the checking of
4480 ignore_target_hosts.
4482 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4483 being passed; this would have caused $sender_host_name to have been used if
4484 matching the list had actually called for a host name (not very likely,
4485 since this list is usually IP addresses). A host name is now passed as
4488 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4489 code that specifies a non-restarting handler (typically sigaction() in
4490 modern systems) in an attempt to fix a rare and obscure crash bug.
4492 25. Narrowed the window for a race in the daemon that could cause it to ignore
4493 SIGCHLD signals. This is not a major problem, because they are used only to
4494 wake it up if nothing else does.
4496 26. A malformed maildirsize file could cause Exim to calculate negative values
4497 for the mailbox size or file count. Odd effects could occur as a result.
4498 The maildirsize information is now recalculated if the size or filecount
4501 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4502 support for a long time. Removed HAVE_SYS_VFS_H.
4504 28. Installed the latest version of exipick from John Jetmore.
4506 29. In an address list, if the pattern was not a regular expression, an empty
4507 subject address (from a bounce message) matched only if the pattern was an
4508 empty string. Non-empty patterns were not even tested. This was the wrong
4509 because it is perfectly reasonable to use an empty address as part of a
4510 database query. An empty address is now tested by patterns that are
4511 lookups. However, all the other forms of pattern expect the subject to
4512 contain a local part and a domain, and therefore, for them, an empty
4513 address still always fails if the pattern is not itself empty.
4515 30. Exim went into a mad DNS loop when attempting to do a callout where the
4516 host was specified on an smtp transport, and looking it up yielded more
4517 than one IP address.
4519 31. Re-factored the code for checking spool and log partition space into a
4520 function that finds that data and another that does the check. The former
4521 is then used to implement four new variables: $spool_space, $log_space,
4522 $spool_inodes, and $log_inodes.
4524 32. The RFC2047 encoding function was originally intended for short strings
4525 such as real names; it was not keeping to the 75-character limit for
4526 encoded words that the RFC imposes. It now respects the limit, and
4527 generates multiple encoded words if necessary. To be on the safe side, I
4528 have increased the buffer size for the ${rfc2047: expansion operator from
4531 33. It is now permitted to omit both strings after an "if" condition; if the
4532 condition is true, the result is "true". As before, when the second string
4533 is omitted, a false condition yields an empty string. This makes it less
4534 cumbersome to write custom ACL and router conditions.
4536 34. Failure to deliver a bounce message always caused it to be frozen, even if
4537 there was an errors_to setting on the router. The errors_to setting is now
4540 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4541 canonical form (fully expanded) before being placed in
4542 $sender_host_address.
4544 36. The table in the code that translates DNS record types into text (T_A to
4545 "A" for instance) was missing entries for NS and CNAME. It is just possible
4546 that this could have caused confusion if both these types were looked up
4547 for the same domain, because the text type is used as part of Exim's
4548 per-process caching. But the chance of anyone hitting this buglet seems
4551 37. The dnsdb lookup has been extended in a number of ways.
4553 (1) There is a new type, "zns", which walks up the domain tree until it
4554 finds some nameserver records. It should be used with care.
4556 (2) There is a new type, "mxh", which is like "mx" except that it returns
4557 just the host names, not the priorities.
4559 (3) It is now possible to give a list of domains (or IP addresses) to be
4560 looked up. The behaviour when one of the lookups defers can be
4561 controlled by a keyword.
4563 (4) It is now possible to specify the separator character for use when
4564 multiple records are returned.
4566 38. The dnslists ACL condition has been extended: it is now possible to supply
4567 a list of IP addresses and/or domains to be looked up in a particular DNS
4570 39. Added log_selector=+queue_time_overall.
4572 40. When running the queue in the test harness, wait just a tad after forking a
4573 delivery process, to get repeatability of debugging output.
4575 41. Include certificate and key file names in error message when GnuTLS fails
4576 to set them up, because the GnuTLS error message doesn't include the name
4577 of the failing file when there is a problem reading it.
4579 42. Allow both -bf and -bF in the same test run.
4581 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4583 44. The "Exiscan patch" is now merged into the mainline Exim source.
4585 45. Sometimes the final signoff response after QUIT could fail to get
4586 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4587 < 0 before doing a fflush(). This bug looks as though it goes back to the
4588 introduction of TLS in release 3.20, but "sometimes" must have been rare
4589 because the tests only now provoked it.
4591 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4592 (this can affect the format of dates).
4594 47. exim_tidydb, when checking for the continued existence of a message for
4595 which it has found a message-specific retry record, was not finding
4596 messages that were in split spool directories. Consequently, it was
4597 deleting retry records that should have stayed in existence.
4599 48. Steve fixed some bugs in eximstats.
4601 49. The SPA authentication driver was not abandoning authentication and moving
4602 on to the next authenticator when an expansion was forced to fail,
4603 contradicting the general specification for all authenticators. Instead it
4604 was generating a temporary error. It now behaves as specified.
4606 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4607 (the order specifies the preference for clients). The order is now AES256,
4608 AES128, 3DES, ARCFOUR128.
4610 51. Small patch to Sieve code - explicitly set From: when generating an
4613 52. Exim crashed if a remote delivery caused a very long error message to be
4614 recorded - for instance if somebody sent an entire SpamAssassin report back
4615 as a large number of 550 error lines. This bug was coincidentally fixed by
4616 increasing the size of one of Exim's internal buffers (big_buffer) that
4617 happened as part of the Exiscan merge. However, to be on the safe side, I
4618 have made the code more robust (and fixed the comments that describe what
4621 53. Now that there can be additional text after "Completed" in log lines (if
4622 the queue_time_overall log selector is set), a one-byte patch to exigrep
4623 was needed to allow it to recognize "Completed" as not the last thing in
4626 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4627 patch that reportedly fixes this has been added. I am not expert enough to
4628 create a test for it. This is what the patch creator wrote:
4630 "I found a little strange behaviour of ldap code when working with
4631 Windows 2003 AD Domain, where users was placed in more than one
4632 Organization Units. When I tried to give exim partial DN, the exit code
4633 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4634 But simultaneously result of request was absolutely normal ldap result,
4635 so I produce this patch..."
4637 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4638 REFERENCE, so I have modified the code to exclude the patch when that macro
4641 55. Some experimental protocols are using DNS PTR records for new purposes. The
4642 keys for these records are domain names, not reversed IP addresses. The
4643 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4644 leaves it alone. Component reversal etc. now happens only for IP addresses.
4647 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4649 57. Double the size of the debug message buffer (to 2048) so that more of very
4650 long debug lines gets shown.
4652 58. The exicyclog utility now does better if the number of log files to keep
4653 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4655 59. Two changes related to the smtp_active_hostname option:
4657 (1) $smtp_active_hostname is now available as a variable.
4658 (2) The default for smtp_banner uses $smtp_active_hostname instead
4659 of $primary_hostname.
4661 60. The host_aton() function is supposed to be passed a string that is known
4662 to be a valid IP address. However, in the case of IPv6 addresses, it was
4663 not checking this. This is a hostage to fortune. Exim now panics and dies
4664 if the condition is not met. A case was found where this could be provoked
4665 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4666 components; fortuitously, this particular loophole had already been fixed
4667 by change 4.50/55 above.
4669 If there are any other similar loopholes, the new check in host_aton()
4670 itself should stop them being exploited. The report I received stated that
4671 data on the command line could provoke the exploit when Exim was running as
4672 exim, but did not say which command line option was involved. All I could
4673 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4674 running as the user.
4677 61. There was a buffer overflow vulnerability in the SPA authentication code
4678 (which came originally from the Samba project). I have added a test to the
4679 spa_base64_to_bits() function which I hope fixes it.
4682 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4683 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4685 63. The daemon start-up calls getloadavg() while still root for those OS that
4686 need the first call to be done as root, but it missed one case: when
4687 deliver_queue_load_max is set with deliver_drop_privilege. This is
4688 necessary for the benefit of the queue runner, because there is no re-exec
4689 when deliver_drop_privilege is set.
4691 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4692 This has been fixed.
4694 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4695 was in use, was not putting the data itself into the right store pool;
4696 consequently, it could be overwritten for a subsequent message in the same
4697 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4700 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4702 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4703 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4704 the function were treating the return as a boolean value, which happened to
4705 work because 0=false and not-0=true, but is not correct code.
4707 68. The host_aton() function was not handling scoped IPv6 addresses (those
4708 with, for example, "%eth0" on the end) correctly.
4710 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4711 negated items (that is, ~something) in unsigned ints. Some compilers
4712 apparently mutter when there is no cast.
4714 70. If an address verification called from an ACL failed, and did not produce a
4715 user-specific message (i.e. there was only a "system" message), nothing was
4716 put in $acl_verify_message. In this situation, it now puts the system
4719 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4720 session; change 4.31/43 added the unwanted input to the log line - except
4721 that it did not do this in the start of session case. It now does.
4723 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4724 This is wrong and can cause the other end to generate a synchronization
4725 error if it is another Exim or anything else that does the synchronization
4726 check. A QUIT command is no longer sent after a timeout.
4728 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4729 during host lookups.
4731 74. The defer_ok option of callout verification was not working if it was used
4732 when verifying addresses in header lines, that is, for this case:
4734 verify = header_sender/callout=defer_ok
4736 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4737 those file descriptors could be used for SMTP connections. If anything
4738 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4739 could be sent to the SMTP client, causing chaos. The daemon now opens
4740 stdin, stdout, and stderr to /dev/null when it puts itself into the
4743 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4744 log by default. The user can override this with suitable Perl magic.
4746 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4747 the log message when discard triggers, was not working for the DATA ACL or
4748 for the non-SMTP ACL.
4750 78. Error message wording change in sieve.c.
4752 79. If smtp_accept_max_per_host was set, the number of connections could be
4753 restricted to fewer than expected, because the daemon was trying to set up
4754 a new connection before checking whether the processes handling previous
4755 connections had finished. The check for completed processes is now done
4756 earlier. On busy systems, this bug wouldn't be noticed because something
4757 else would have woken the daemon, and it would have reaped the completed
4760 80. If a message was submitted locally by a user whose login name contained one
4761 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4762 It caused a spool format error. I have fixed the spool reading code. A
4763 related problem was that the "from" clause in the Received: line became
4764 illegal because of the space(s). It is now covered by ${quote_local_part.
4766 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4769 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4771 83. Patch from Sieve maintainer to fix unterminated string problem in
4772 "vacation" handling.
4774 84. Some minor changes to the Linux configuration files to help with other
4775 OS variants using glibc.
4777 85. One more patch for Sieve to update vacation handling to latest spec.
4780 ----------------------------------------------------
4781 See the note above about the 4.44 and 4.50 releases.
4782 ----------------------------------------------------
4788 1. Change 4.43/35 introduced a bug that caused file counts to be
4789 incorrectly computed when quota_filecount was set in an appendfile
4792 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4793 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4796 3. Give more explanation in the error message when the command for a transport
4797 filter fails to execute.
4799 4. There are several places where Exim runs a non-Exim command in a
4800 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4801 was being done only for the command run by the queryprogram router. It is
4802 now done for all such subprocesses. The other cases are: ${run, transport
4803 filters, and the commands run by the lmtp and pipe transports.
4805 5. Some older OS have a limit of 256 on the maximum number of file
4806 descriptors. Exim was using setrlimit() to set 1000 as a large value
4807 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4808 systems. I've change it so that if it can't get 1000, it tries for 256.
4810 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4811 was an oversight, and furthermore, ever since the addition of extra
4812 controls (e.g. 4.43/32), the checks on when to allow different forms of
4813 "control" were broken. There should now be diagnostics for all cases when a
4814 control that does not make sense is encountered.
4816 7. $recipients is now available in the predata ACL (oversight).
4818 8. Tidy the search cache before the fork to do a delivery from a message
4819 received from the command line. Otherwise the child will trigger a lookup
4820 failure and thereby defer the delivery if it tries to use (for example) a
4821 cached ldap connection that the parent has called unbind on.
4823 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4824 of $address_data from the recipient verification was clobbered by the
4825 sender verification.
4827 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4828 was its contents. (It was OK if the option was not defined at all.)
4830 11. A "Completed" log line is now written for messages that are removed from
4831 the spool by the -Mrm option.
4833 12. $host_address is now set to the target address during the checking of
4834 ignore_target_hosts.
4836 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4837 being passed; this would have caused $sender_host_name to have been used if
4838 matching the list had actually called for a host name (not very likely,
4839 since this list is usually IP addresses). A host name is now passed as
4842 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4843 code that specifies a non-restarting handler (typically sigaction() in
4844 modern systems) in an attempt to fix a rare and obscure crash bug.
4846 15. Narrowed the window for a race in the daemon that could cause it to ignore
4847 SIGCHLD signals. This is not a major problem, because they are used only to
4848 wake it up if nothing else does.
4850 16. A malformed maildirsize file could cause Exim to calculate negative values
4851 for the mailbox size or file count. Odd effects could occur as a result.
4852 The maildirsize information is now recalculated if the size or filecount
4855 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4856 support for a long time. Removed HAVE_SYS_VFS_H.
4858 18. Updated exipick to current release from John Jetmore.
4860 19. Allow an empty sender to be matched against a lookup in an address list.
4861 Previously the only cases considered were a regular expression, or an
4864 20. Exim went into a mad DNS lookup loop when doing a callout where the
4865 host was specified on the transport, if the DNS lookup yielded more than
4868 21. The RFC2047 encoding function was originally intended for short strings
4869 such as real names; it was not keeping to the 75-character limit for
4870 encoded words that the RFC imposes. It now respects the limit, and
4871 generates multiple encoded words if necessary. To be on the safe side, I
4872 have increased the buffer size for the ${rfc2047: expansion operator from
4875 22. Failure to deliver a bounce message always caused it to be frozen, even if
4876 there was an errors_to setting on the router. The errors_to setting is now
4879 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4880 canonical form (fully expanded) before being placed in
4881 $sender_host_address.
4883 24. Updated eximstats to version 1.33
4885 25. Include certificate and key file names in error message when GnuTLS fails
4886 to set them up, because the GnuTLS error message doesn't include the name
4887 of the failing file when there is a problem reading it.
4889 26. Expand error message when OpenSSL has problems setting up cert/key files.
4892 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4893 (this can affect the format of dates).
4895 28. exim_tidydb, when checking for the continued existence of a message for
4896 which it has found a message-specific retry record, was not finding
4897 messages that were in split spool directories. Consequently, it was
4898 deleting retry records that should have stayed in existence.
4900 29. eximstats updated to version 1.35
4901 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4902 1.35 - bugfix such that pie charts by volume are generated correctly
4904 30. The SPA authentication driver was not abandoning authentication and moving
4905 on to the next authenticator when an expansion was forced to fail,
4906 contradicting the general specification for all authenticators. Instead it
4907 was generating a temporary error. It now behaves as specified.
4909 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4910 (the order specifies the preference for clients). The order is now AES256,
4911 AES128, 3DES, ARCFOUR128.
4913 31. Small patch to Sieve code - explicitly set From: when generating an
4916 32. Exim crashed if a remote delivery caused a very long error message to be
4917 recorded - for instance if somebody sent an entire SpamAssassin report back
4918 as a large number of 550 error lines. This bug was coincidentally fixed by
4919 increasing the size of one of Exim's internal buffers (big_buffer) that
4920 happened as part of the Exiscan merge. However, to be on the safe side, I
4921 have made the code more robust (and fixed the comments that describe what
4924 33. Some experimental protocols are using DNS PTR records for new purposes. The
4925 keys for these records are domain names, not reversed IP addresses. The
4926 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4927 leaves it alone. Component reversal etc. now happens only for IP addresses.
4930 34. The host_aton() function is supposed to be passed a string that is known
4931 to be a valid IP address. However, in the case of IPv6 addresses, it was
4932 not checking this. This is a hostage to fortune. Exim now panics and dies
4933 if the condition is not met. A case was found where this could be provoked
4934 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4935 components; fortuitously, this particular loophole had already been fixed
4936 by change 4.50/55 or 4.44/33 above.
4938 If there are any other similar loopholes, the new check in host_aton()
4939 itself should stop them being exploited. The report I received stated that
4940 data on the command line could provoke the exploit when Exim was running as
4941 exim, but did not say which command line option was involved. All I could
4942 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4943 running as the user.
4946 35. There was a buffer overflow vulnerability in the SPA authentication code
4947 (which came originally from the Samba project). I have added a test to the
4948 spa_base64_to_bits() function which I hope fixes it.
4951 36. The daemon start-up calls getloadavg() while still root for those OS that
4952 need the first call to be done as root, but it missed one case: when
4953 deliver_queue_load_max is set with deliver_drop_privilege. This is
4954 necessary for the benefit of the queue runner, because there is no re-exec
4955 when deliver_drop_privilege is set.
4957 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4958 was in use, was not putting the data itself into the right store pool;
4959 consequently, it could be overwritten for a subsequent message in the same
4960 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4963 38. Sometimes the final signoff response after QUIT could fail to get
4964 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4965 < 0 before doing a fflush(). This bug looks as though it goes back to the
4966 introduction of TLS in release 3.20, but "sometimes" must have been rare
4967 because the tests only now provoked it.
4973 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4974 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4975 Now it returns an integer. A number of calls were still expecting a T/F
4976 return. Fortuitously, in all cases, the tests worked in OK situations,
4977 which is the norm. However, things would have gone wrong on any write
4978 failures on the smtp file descriptor. This function is used when sending
4979 messages over SMTP and also when doing verify callouts.
4981 2. When Exim is called to do synchronous delivery of a locally submitted
4982 message (the -odf or -odi options), it no longer closes stderr before doing
4985 3. Implemented the mua_wrapper option.
4987 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4989 5. Implemented the functions header_remove(), header_testname(),
4990 header_add_at_position(), and receive_remove_recipient(), and exported them
4993 6. If an ACL "warn" statement specified the addition of headers, Exim already
4994 inserted X-ACL-Warn: at the start if there was no header name. However, it
4995 was not making this test for the second and subsequent header lines if
4996 there were newlines in the string. This meant that an invalid header could
4997 be inserted if Exim was badly configured.
4999 7. Allow an ACL "warn" statement to add header lines at the start or after all
5000 the Received: headers, as well as at the end.
5002 8. Added the rcpt_4xx retry error code.
5004 9. Added postmaster_mailfrom=xxx to callout verification option.
5006 10. Added mailfrom=xxxx to the callout verification option, for verify=
5009 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5010 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5012 12. Inserted some casts to stop certain compilers warning when using pointer
5013 differences as field lengths or precisions in printf-type calls (mostly
5014 affecting debugging statements).
5016 13. Added optional readline() support for -be (dynamically loaded).
5018 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5019 same clock tick as a message's arrival, so that its received time was the
5020 same as the "first fail" time on the retry record, and that message
5021 remained on the queue past the ultimate address timeout, every queue runner
5022 would try a delivery (because it was past the ultimate address timeout) but
5023 after another failure, the ultimate address timeout, which should have then
5024 bounced the address, did not kick in. This was a "< instead of <=" error;
5025 in most cases the first failure would have been in the next clock tick
5026 after the received time, and all would be well.
5028 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5029 being recognized when the domain list was tested by the match_domain
5030 condition in an expansion string.
5032 16. Added the ${str2b64: operator.
5034 17. Exim was always calling setrlimit() to set a large limit for the number of
5035 processes, without checking whether the existing limit was already
5036 adequate. (It did check for the limit on file descriptors.) Furthermore,
5037 errors from getrlimit() and setrlimit() were being ignored. Now they are
5038 logged to the main and panic logs, but Exim does carry on, to try to do its
5039 job under whatever limits there are.
5041 18. Imported PCRE 5.0.
5043 19. Trivial typo in log message " temporarily refused connection" (the leading
5046 20. If the log selector return_path_on_delivery was set and an address was
5047 redirected to /dev/null, the delivery process crashed because it assumed
5048 that a return path would always be set for a "successful" delivery. In this
5049 case, the whole delivery is bypassed as an optimization, and therefore no
5052 21. Internal re-arrangement: the function for sending a challenge and reading
5053 a response while authentication was assuming a zero-terminated challenge
5054 string. It's now changed to take a pointer and a length, to allow for
5055 binary data in such strings.
5057 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5059 23. Exim was not respecting finduser_retries when seeking the login of the
5060 uid under which it was called; it was always trying 10 times. (The default
5061 setting of finduser_retries is zero.) Also, it was sleeping after the final
5062 failure, which is pointless.
5064 24. Implemented tls_on_connect_ports.
5066 25. Implemented acl_smtp_predata.
5068 26. If the domain in control=submission is set empty, Exim assumes that the
5069 authenticated id is a complete email address when it generates From: or
5070 Sender: header lines.
5072 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5073 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5074 chown and chgrp in /bin and hostname in /usr/bin.
5076 28. Exim was keeping the "process log" file open after each use, just as it
5077 does for the main log. This opens the possibility of it remaining open for
5078 long periods when the USR1 signal hits a daemon. Occasional processlog
5079 errors were reported, that could have been caused by this. Anyway, it seems
5080 much more sensible not to leave this file open at all, so that is what now
5083 29. The long-running daemon process does not normally write to the log once it
5084 has entered its main loop, and it closes the log before doing so. This is
5085 so that log files can straightforwardly be renamed and moved. However,
5086 there are a couple of unusual error situations where the daemon does write
5087 log entries, and I had neglected to close the log afterwards.
5089 30. The text of an SMTP error response that was received during a remote
5090 delivery was being truncated at 512 bytes. This is too short for some of
5091 the long messages that one sometimes sees. I've increased the limit to
5094 31. It is now possible to make retry rules that apply only when a message has a
5095 specific sender, in particular, an empty sender.
5097 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5098 it possible to be selective about when SMTP synchronization is enforced.
5100 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5102 32. Implemented hosts_connection_nolog.
5104 33. Added an ACL for QUIT.
5106 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5109 35. Added mailbox_size and mailbox_filecount to appendfile.
5111 36. Added control = no_multiline_responses to ACLs.
5113 37. There was a bug in the logic of the code that waits for the clock to tick
5114 in the case where the clock went backwards by a substantial amount such
5115 that the microsecond fraction of "now" was more than the microsecond
5116 fraction of "then" (but the whole seconds number was less).
5118 38. Added support for the libradius Radius client library this is found on
5119 FreeBSD (previously only the radiusclient library was supported).
5125 1. When certain lookups returned multiple values in the form name=value, the
5126 quoting of the values was not always being done properly. Specifically:
5127 (a) If the value started with a double quote, but contained no whitespace,
5129 (b) If the value contained whitespace other than a space character (i.e.
5130 tabs or newlines or carriage returns) it was not quoted.
5131 This fix has been applied to the mysql and pgsql lookups by writing a
5132 separate quoting function and calling it from the lookup code. The fix
5133 should probably also be applied to nisplus, ibase and oracle lookups, but
5134 since I cannot test any of those, I have not disturbed their existing code.
5136 2. A hit in the callout cache for a specific address caused a log line with no
5137 reason for rejecting RCPT. Now it says "Previous (cached) callout
5138 verification failure".
5140 3. There was an off-by-one bug in the queryprogram router. An over-long
5141 return line was truncated at 256 instead of 255 characters, thereby
5142 overflowing its buffer with the terminating zero. As well as fixing this, I
5143 have increased the buffer size to 1024 (and made a note to document this).
5145 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5146 when Exim is waiting for an SMTP response from a remote server, Exim
5147 restarts its select() call on the socket, thereby resetting its timeout.
5148 This is not a problem when such interrupts are rare. Somebody set up a cron
5149 job to run exiwhat every 2 minutes, which is less than the normal select()
5150 timeout (5 or 10 minutes). This meant that the select() timeout never
5151 kicked in because it was always reset. I have fixed this by comparing the
5152 time when an interrupt arrives with the time at the start of the first call
5153 to select(). If more time than the timeout has elapsed, the interrupt is
5154 treated as a timeout.
5156 5. Some internal re-factoring in preparation for the addition of Sieve
5157 extensions (by MH). In particular, the "personal" test is moved to a
5158 separate function, and given an option for scanning Cc: and Bcc: (which is
5159 not set for Exim filters).
5161 6. When Exim created an email address using the login of the caller as the
5162 local part (e.g. when creating a From: or Sender: header line), it was not
5163 quoting the local part when it contained special characters such as @.
5165 7. Installed new OpenBSD configuration files.
5167 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5168 try to make them clearer.
5170 9. Callout options, other than the timeout value, were being ignored when
5171 verifying sender addresses in header lines. For example, when using
5173 verify = header_sender/callout=no_cache
5175 the cache was (incorrectly) being used.
5177 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5178 only the Cygwin environment.
5180 11. When return_path_on_delivery was set as a log selector, if different remote
5181 addresses in the same message used different return paths and parallel
5182 remote delivery occurred, the wrong values would sometimes be logged.
5183 (Whenever a remote delivery process finished, the return path value from
5184 the most recently started remote delivery process was logged.)
5186 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5187 lines when AUTH and/or TLS are in use. This is the "received protocol"
5188 field. Exim used to use "asmtp" for authenticated SMTP, without any
5189 indication (in the protocol name) for TLS use. Now it follows the RFC and
5190 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5191 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5192 names appear in log lines as well as in Received: header lines.
5194 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5195 extensions, and comparison tests, and to fix some bugs.
5197 14. Changes to the "personal" filter test:
5199 (1) The test was buggy in that it was just doing the equivalent of
5200 "contains" tests on header lines. For example, if a user's address was
5201 anne@some.where, the "personal" test would incorrectly be true for
5203 To: susanne@some.where
5205 This test is now done by extracting each address from the header in turn,
5206 and checking the entire address. Other tests that are part of "personal"
5207 are now done using regular expressions (for example, to check local parts
5208 of addresses in From: header lines).
5210 (2) The list of non-personal local parts in From: addresses has been
5211 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5212 taken from the Sieve specification recommendations.
5214 (3) If the message contains any header line starting with "List-" it is
5215 treated as non-personal.
5217 (4) The test for "circular" in the Subject: header line has been removed
5218 because it now seems ill-conceived.
5220 15. Minor typos in src/EDITME comments corrected.
5222 16. Installed latest exipick from John Jetmore.
5224 17. If headers_add on a router specified a text string that was too long for
5225 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5226 of string_sprintf() is now avoided.
5228 18. $message_body_size was not set (it was always zero) when running the DATA
5229 ACL and the local_scan() function.
5231 19. For the "mail" command in an Exim filter, no default was being set for
5232 the once_repeat time, causing a random time value to be used if "once" was
5233 specified. (If the value happened to be <= 0, no repeat happened.) The
5234 default is now 0s, meaning "never repeat". The "vacation" command was OK
5235 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5236 (I found it when inspecting the code).
5238 20. There is now an overall timeout for performing a callout verification. It
5239 defaults to 4 times the callout timeout, which applies to individual SMTP
5240 commands during the callout. The overall timeout applies when there is more
5241 than one host that can be tried. The timeout is checked before trying the
5242 next host. This prevents very long delays if there are a large number of
5243 hosts and all are timing out (e.g. when the network connections are timing
5244 out). The value of the overall timeout can be changed by specifying an
5245 additional sub-option for "callout", called "maxwait". For example:
5247 verify = sender/callout=5s,maxwait=20s
5249 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5250 to the end before writing, but this should make it even safer).
5252 22. Exim was forgetting that it had advertised PIPELINING for the second and
5253 subsequent messages on an SMTP connection. It was also not resetting its
5254 memory on STARTTLS and an internal HELO.
5256 23. When Exim logs an SMTP synchronization error within a session, it now
5257 records whether PIPELINING has been advertised or not.
5259 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5260 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5261 rather than long int.
5263 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5265 26. Added the never_mail option to autoreply.
5271 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5272 crash if the getsockname() call failed; this can happen if a connection is
5273 closed very soon after it is established. The problem was simply in the
5274 order in which certain operations were done, causing Exim to try to write
5275 to the SMTP stream before it had set up the file descriptor. The bug has
5276 been fixed by making things happen in the correct order.
5282 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5283 before the connection was closed, thus losing the rejection response.
5285 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5286 some early Solaris releases, but causes trouble in current releases where
5287 socklen_t is defined.
5289 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5292 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5295 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5296 that releasing the top of it at the end releases what was used for sub-
5297 expansions (unless the block got too big). However, discard this block if
5298 the first thing is a variable or header, so that we can use its block when
5299 it is dynamic (useful for very large $message_headers, for example).
5301 6. Lookups now cache *every* query, not just the most recent. A new, separate
5302 store pool is used for this. It can be recovered when all lookup caches are
5303 flushed. Lookups now release memory at the end of their result strings.
5304 This has involved some general refactoring of the lookup sources.
5306 7. Some code has been added to the store_xxx() functions to reduce the amount
5307 of flapping under certain conditions.
5309 8. log_incoming_interface used to affect only the <= reception log lines. Now
5310 it causes the local interface and port to be added to several more SMTP log
5311 lines, for example "SMTP connection from", and rejection lines.
5313 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5315 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5317 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5318 could be overwritten at the end of the current message (or the start of a
5319 new message if it was set in a HELO ACL). The value is now preserved for
5320 the duration of the SMTP connection.
5322 12. If a transport had a headers_rewrite setting, and a matching header line
5323 contained an unqualified address, that address was qualified, even if it
5324 did not match any rewriting rules. The underlying bug was that the values
5325 of the flags that permit the existence of unqualified sender and recipient
5326 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5327 non-local messages, and by -bnq for local messages) were not being
5328 preserved with the message after it was received.
5330 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5331 "next input=" as part of the text comprising the host identity instead of
5332 the correct text. The code was using the same buffer for two different
5333 strings. However, depending on which order the printing function evaluated
5334 its arguments, the bug did not always show up. Under Linux, for example, my
5335 test suite worked just fine.
5337 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5338 doesn't work with some older versions of Perl. It has been changed to "my",
5339 which in any case is probably the better facility to use.
5341 15. A really picky compiler found some instances of statements for creating
5342 error messages that either had too many or two few arguments for the format
5345 16. The size of the buffer for calls to the DNS resolver has been increased
5346 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5347 for addresses that have a lot of PTR records. This alleviates a problem; it
5348 does not fully solve it.
5350 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5351 buffer now truncates the list and logs the incident, which is the same
5352 action as happens when Exim is looking up a host name and its aliases.
5353 Previously in this situation something unpredictable would happen;
5354 sometimes it was "internal error: store_reset failed".
5356 18. If a server dropped the connection unexpectedly when an Exim client was
5357 using GnuTLS and trying to read a response, the client delivery process
5358 crashed while trying to generate an error log message.
5360 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5361 string, for example:
5363 warn message = H1: something\nH2: something
5365 the text was added as a single header line from Exim's point of view
5366 though it ended up OK in the delivered message. However, searching for the
5367 second and subsequent header lines using $h_h2: did not work. This has been
5368 fixed. Similarly, if a system filter added multiple headers in this way,
5369 the routers could not see them.
5371 20. Expanded the error message when iplsearch is called with an invalid key to
5372 suggest using net-iplsearch in a host list.
5374 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5375 ACLs are no longer actually imposed (and a message to that effect is
5378 22. If a "gecos" field in a passwd entry contained escaped characters, in
5379 particular, if it contained a \" sequence, Exim got it wrong when building
5380 a From: or a Sender: header from that name. A second bug also caused
5381 incorrect handling when an unquoted " was present following a character
5382 that needed quoting.
5384 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5385 was not being matched caselessly.
5387 24. Arranged for all hyphens in the exim.8 source to be escaped with
5390 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5391 properly. Recipient callout cache records were still being keyed to include
5392 the sender, even when use_sender was set false. This led to far more
5393 callouts that were necessary. The sender is no longer included in the key
5394 when use_sender is false.
5396 26. Added "control = submission" modifier to ACLs.
5398 27. Added the ${base62d: operator to decode base 62 numbers.
5400 28. dnsdb lookups can now access SRV records.
5402 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5403 the configuration file.
5405 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5406 (-v) mode. This makes the output for a verbose queue run more intelligible.
5408 31. Added a use_postmaster feature to recipient callouts.
5410 32. Added the $body_zerocount variable, containing the number of binary zero
5411 bytes in the message body.
5413 33. The time of last modification of the "new" subdirectory is now used as the
5414 "mailbox time last read" when there is a quota error for a maildir
5417 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5419 35. Added +ignore_unknown as a special item in host lists.
5421 36. Code for decoding IPv6 addresses in host lists is now included, even if
5422 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5423 address was recognized as an IP address, but was then not correctly decoded
5424 into binary, causing unexpected and incorrect effects when compared with
5431 1. Very minor rewording of debugging text in manualroute to say "list of
5432 hosts" instead of "hostlist".
5434 2. If verify=header_syntax was set, and a header line with an unqualified
5435 address (no domain) and a large number of spaces between the end of the
5436 name and the colon was received, the reception process suffered a buffer
5437 overflow, and (when I tested it) crashed. This was caused by some obsolete
5438 code that should have been removed. The fix is to remove it!
5440 3. When running in the test harness, delay a bit after writing a bounce
5441 message to get a bit more predictability in the log output.
5443 4. Added a call to search_tidyup() just before forking a reception process. In
5444 theory, someone could use a lookup in the expansion of smtp_accept_max_
5445 per_host which, without the tidyup, could leave open a database connection.
5447 5. Added the variables $recipient_data and $sender_data which get set from a
5448 lookup success in an ACL "recipients" or "senders" condition, or a router
5449 "senders" option, similar to $domain_data and $local_part_data.
5451 6. Moved the writing of debug_print from before to after the "senders" test
5454 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5455 problems for message scanning, either using a data ACL, or using
5456 local_scan() because the Received: header was not generated till after they
5457 were called (in order to set the time as the time of reception completion).
5458 I have revised the way this works. The header is now generated after the
5459 body is received, but before the ACL or local_scan() are called. After they
5460 are run, the timestamp in the header is updated.
5466 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5467 before starting a queue runner without re-exec. This happened only when
5468 deliver_drop_privilege was set or when the Exim user was set to root. The
5469 effect of the bug was that timeouts during subsequent deliveries caused
5470 crashes instead of being properly handled. The handler is now left at its
5471 default (and expected) setting.
5473 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5474 message, again when deliver_drop_privilege is set or Exim is run as root.
5475 The bug described in (1) was not present in this case, but the tidying up
5476 of the other signals was missing. I have made the two cases consistent.
5478 3. The ignore_target_hosts setting on a manualroute router was being ignored
5479 for hosts that were looked up using the /MX notation.
5481 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5484 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5485 operated on the sender address. After changing the $sender_address to <>
5486 for the sender address verify, Exim was re-instated it as the original
5487 (before rewriting) address, but remembering that it had rewritten it, so it
5488 wasn't rewriting it again. This bug also had the effect of breaking the
5489 sender address verification caching when the sender address was rewritten.
5491 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5492 This has been changed so that if the ip literal address matches
5493 ignore_target_hosts, the router declines.
5495 7. Added expansion conditions match_domain, match_address, and match_local_
5496 part (NOT match_host).
5498 8. The placeholder for the Received: header didn't have a length field set.
5500 9. Added code to Exim itself and to exim_lock to test for a specific race
5501 condition that could lead to file corruption when using MBX delivery. The
5502 issue is with the lockfile that is created in /tmp. If this file is removed
5503 after a process has opened it but before that process has acquired a lock,
5504 there is the potential for a second process to recreate the file and also
5505 acquire a lock. This could lead to two Exim processes writing to the file
5506 at the same time. The added code performs the same test as UW imapd; it
5507 checks after acquiring the lock that its file descriptor still refers to
5508 the same named file.
5510 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5511 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5514 11. Added the smtp_active_hostname option. If used, this will typically be made
5515 to depend on the incoming interface address. Because $interface_address is
5516 not set up until the daemon has forked a reception process, error responses
5517 that can happen earlier (such as "too many connections") no longer contain
5520 12. If an expansion in a condition on a "warn" statement fails because a lookup
5521 defers, the "warn" statement is abandoned, and the next ACL statement is
5522 processed. Previously this caused the whole ACL to be aborted.
5524 13. Added the iplsearch lookup type.
5526 14. Added ident_timeout as a log selector.
5528 15. Added tls_certificate_verified as a log selector.
5530 16. Added a global option tls_require_ciphers (compare the smtp transport
5531 option of the same name). This controls incoming TLS connections.
5533 17. I finally figured out how to make tls_require_ciphers do a similar thing
5534 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5535 before starting the TLS session.
5537 18. Tabs are now shown as \t in -bP output.
5539 19. If the log selector return_path_on_delivery was set, Exim crashed when
5540 bouncing a message because it had too many Received: header lines.
5542 20. If two routers both had headers_remove settings, and the first one included
5543 a superfluous trailing colon, the final name in the first list and the
5544 first name in the second list were incorrectly joined into one item (with a
5545 colon in the middle).
5551 1. Added -C and -D options to the exinext utility, mainly to make it easier
5552 to include in the automated testing, but these could be helpful when
5553 multiple configurations are in use.
5555 2. The exinext utility was not formatting the output nicely when there was
5556 an alternate port involved in the retry record key, nor when there was a
5557 message id as well (for retries that were specific to a specific message
5558 and a specific host). It was also confused by IPv6 addresses, because of
5559 the additional colons they contain. I have fixed the IPv4 problem, and
5560 patched it up to do a reasonable job for IPv6.
5562 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5563 delivery, the log line now contains "pipelined" if PIPELINING was used.
5565 4. An SMTP transport process used to panic and die if the bind() call to set
5566 an explicit outgoing interface failed. This has been changed; it is now
5567 treated in the same way as a connect() failure.
5569 5. A reference to $sender_host_name in the part of a conditional expansion
5570 that was being skipped was still causing a DNS lookup. This no longer
5573 6. The def: expansion condition was not recognizing references to header lines
5574 that used bh_ and bheader_.
5576 7. Added the _cache feature to named lists.
5578 8. The code for checking quota_filecount in the appendfile transport was
5579 allowing one more file than it should have been.
5581 9. For compatibility with Sendmail, the command line option
5589 and sets the incoming protocol and host name (for trusted callers). The
5590 host name and its colon can be omitted when only the protocol is to be set.
5591 Note the Exim already has two private options, -pd and -ps, that refer to
5592 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5593 "s", but I don't think that's a major issue.
5595 10. A number of refactoring changes to the code, none of which should affect
5598 (a) The number of logging options was getting close to filling up the
5599 32-bit word that was used as a bit map. I have split them into two classes:
5600 those that are passed in the argument to log_write(), and those that are
5601 only ever tested independently outside of that function. These are now in
5602 separate 32-bit words, so there is plenty of room for expansion again.
5603 There is no change in the user interface or the logging behaviour.
5605 (b) When building, for example, log lines, the code previously used a
5606 macro that called string_cat() twice, in order to add two strings. This is
5607 not really sufficiently general. Furthermore, there was one instance where
5608 it was actually wrong because one of the argument was used twice, and in
5609 one call a function was used. (As it happened, calling the function twice
5610 did not affect the overall behaviour.) The macro has been replaced by a
5611 function that can join an arbitrary number of extra strings onto a growing
5614 (c) The code for expansion conditions now uses a table and a binary chop
5615 instead of a serial search (which was left over from when there were very
5616 few conditions). Also, it now recognizes conditions like "pam" even when
5617 the relevant support is not compiled in: a suitably worded error message is
5618 given if an attempt is made to use such a condition.
5620 11. Added ${time_interval:xxxxx}.
5622 12. A bug was causing one of the ddress fields not to be passed back correctly
5623 from remote delivery subprocesses. The field in question was not being
5624 subsequently used, so this caused to problems in practice.
5626 13. Added new log selectors queue_time and deliver_time.
5628 14. Might have fixed a bug in maildirsizefile handling that threw up
5629 "unexpected character" debug warnings, and recalculated the data
5630 unnecessarily. In any case, I expanded the warning message to give more
5633 15. Added the message "Restricted characters in address" to the statements in
5634 the default ACL that block characters like @ and % in local parts.
5636 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5637 Three changes have been made:
5639 (a) There was a serious bug; a negative response to MAIL caused the whole
5640 recipient domain to be cached as invalid, thereby blocking all messages
5641 to all local parts at the same domain, from all senders. This bug has
5642 been fixed. The domain is no longer cached after a negative response to
5643 MAIL if the sender used is not empty.
5645 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5648 (c) A new callout option, "use_sender" has been added for people who want
5649 the modified behaviour.
5655 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5658 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5659 indeed breaks things for older releases.
5661 3. Added additional logging to the case where there is a problem reading data
5662 from a filter that is running in a subprocess using a pipe, in order to
5663 try to track down a specific problem.
5665 4. Testing facility fudge: when running in the test harness and attempting
5666 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5667 getting "No route to host". Convert this to a timeout.
5669 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5672 6. Some OS don't have socklen_t but use size_t instead. This affects the
5673 fifth argument of getsockopt() amongst other things. This is now
5674 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5675 can be set for individual OS. I have set it for SunOS5, OSF1, and
5676 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5677 some earlier ones do not.
5679 7. Change 4.30/15 was not doing the test caselessly.
5681 8. The standard form for an IPv6 address literal was being rejected by address
5682 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5683 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5684 this, as well as the form without the "IPv6" on the front (but only when
5685 address literals are enabled, of course).
5687 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5689 10. Exim crashed if a message with an empty sender address specified by -f
5690 encountered a router with an errors_to setting. This could be provoked only
5691 by a command such as
5695 where an empty string was supplied; "<>" did not hit this bug.
5697 11. Installed PCRE release 4.5.
5699 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5700 remained set. It is now erased.
5702 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5703 times from message ids (which are base 36 rather than the normal 62).
5705 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5706 were being counted as actual protocol errors, and logged if the log
5707 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5708 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5709 and DATA following a set of rejected RCPTs do not count as protocol errors.
5710 In other words, Exim assumes they were pipelined, though this may not
5711 actually be the case. Of course, in all cases the client gets an
5712 appropriate error code.
5714 15. If a lookup fails in an ACL condition, a message about the failure may
5715 be available; it is used if testing the ACL cannot continue, because most
5716 such messages specify what the cause of the deferral is. However, some
5717 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5718 that caused an old message to be retained and used if a later statement
5719 caused a defer, replacing the real cause of the deferral.
5721 16. If an IP address had so many PTR records that the DNS lookup buffer
5722 was not large enough to hold them, Exim could crash while trying to process
5723 the truncated data. It now detects and logs this case.
5725 17. Further to 4.21/58, another change has been made: if (and only if) the
5726 first line of a message (the first header line) ends with CRLF, a bare LF
5727 in a subsequent header line has a space inserted after it, so as not to
5728 terminate the header.
5730 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5731 unnecessarily, used atoi() instead of strtol(), and didn't check the
5732 termination when getting file sizes from file names by regex.
5734 19. Completely re-implemented the support for maildirsize files, in the light
5735 of a number of problems with the previous contributed implementation
5736 (4.30/29). In particular:
5738 . If the quota is zero, the maildirsize file is maintained, but no quota is
5741 . If the maildir directory does not exist, it is created before any attempt
5742 to write a maildirsize file.
5744 . The quota value in the file is just a cache; if the quota is changed in
5745 the transport, the new value overrides.
5747 . A regular expression is available for excluding directories from the
5750 20. The autoreply transport checks the characters in options that define the
5751 message's headers; it allows continued headers, but it was checking with
5752 isspace() after an embedded newline instead of explicitly looking for a
5755 21. If all the "regular" hosts to which an address was routed had passed their
5756 expiry times, and had not reached their retry times, the address was
5757 bounced, even if fallback hosts were defined. Now Exim should go on to try
5760 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5761 equivalent code in the SMTP transport. Some hosts send humungous responses
5762 to HELO/EHLO, more than 1024 it seems.
5764 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5765 gives compiler warnings in some environments. I've now done it "properly",
5768 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5769 (because of problems with the built-in one) was declared to return uschar *
5770 instead of char *, causing compiler failure.
5772 25. Fixed a file descriptor leak when processing alias/forward files.
5774 26. Fixed a minor format string issue in dbfn.c.
5776 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5778 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5779 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5782 29. When a sender address is verified, it is cached, to save repeating the test
5783 when there is more than one recipient in a message. However, when the
5784 verification involves a callout, it is possible for different callout
5785 options to be set for different recipients. It is too complicated to keep
5786 track of this in the cache, so now Exim always runs a verification when a
5787 callout is required, relying on the callout cache for the optimization.
5788 The overhead is duplication of the address routing, but this should not be
5791 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5792 to be verified with callout=postmaster, and the main callout worked but the
5793 postmaster check failed, the verification correctly failed. However, if a
5794 subsequent RCPT command asked for sender verification *without* the
5795 postmaster check, incorrect caching caused this verification also to fail,
5798 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5799 it was not caching the DNS options (qualify_single, search_parents) that
5800 were used when the lookup failed. A subsequent lookup with different
5801 options therefore always gave the same answer, though there were cases
5802 where it should not have. (Example: a "domains = !$mx_any" option on a
5803 dnslookup router: the "domains" option is always processed without any
5804 widening, but the router might have qualify_single set.) Now Exim uses the
5805 cached value only when the same options are set.
5807 32. Added John Jetmore's "exipick" utility to the distribution.
5809 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5810 than a timeout (e.g. a certificate is required, and is not provided), an
5811 Exim server now closes the connection immediately. Previously it waited for
5812 the client to close - but if the client is SSL, it seems that they each
5813 wait for each other, leading to a delay before one of them times out.
5815 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5816 maintained 0.8.x compatibility because I don't think many are using it, and
5817 it is clearly obsolete.
5819 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5822 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5823 client certificate was expired. A simple patch fixes this, though I don't
5824 understand the full logic of why the verify callback is called multiple
5827 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5828 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5829 which causes problems with some clients (such as the Certicom SSL Plus
5830 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5831 disables the coutermeasure allowing Eudora to connect."
5833 38. Exim was not checking that a write() to a log file succeeded. This could
5834 lead to Bad Things if a log got too big, in particular if it hit a file
5835 size limit. Exim now panics and dies if it cannot write to a log file, just
5836 as it does if it cannot open a log file.
5838 39. Modified OS/Makefile-Linux so that it now contains
5840 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5842 The two -D definitions ensure that Exim is compiled with large file
5843 support, which makes it possible to handle log files that are bigger than
5846 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5847 instance) a domain was checked against a named list that involved a lookup,
5848 causing $domain_data to be set, then another domain was checked against the
5849 same list, then the first domain was re-checked, the value of $domain_data
5850 after the final check could be wrong. In particular, if the second check
5851 failed, it could be set empty. This bug probably also applied to
5854 41. The strip_trailing_dot option was not being applied to the address given
5855 with the -f command-line option.
5857 42. The code for reading a message's header from the spool was incrementing
5858 $received_count, but never initializing it. This meant that the value was
5859 incorrect (doubled) while delivering a message in the same process in which
5860 it was received. In the most common configuration of Exim, this never
5861 happens - a fresh exec is done - but it can happen when
5862 deliver_drop_privilege is set.
5864 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5865 it now includes up to 150 characters of the unexpected data in the log
5868 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5869 and building data strings. The size of both of these buffers was 10 000
5870 bytes - far larger than anybody would *ever* want, thought I. Needless to
5871 say, somebody hit the limit. I have increased the maximum line length to
5872 20 000 and the maximum data length of concatenated lines to 100 000. I have
5873 also fixed two bugs, because there was no checking on these buffers. Tsk,
5874 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5875 buffer is too small.
5877 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5878 lsearch lookups. Now it does.
5880 46. When parsing a route_list item in a manualroute router, a fixed-length
5881 buffer was used for the list of hosts. I made this 1024 bytes long,
5882 thinking that nobody would ever have a list of hosts that long. Wrong.
5883 Somebody had a whole pile of complicated expansion conditions, and the
5884 string was silently truncated, leading to an expansion error. It turns out
5885 that it is easier to change to an unlimited length (owing to other changes
5886 that have happened since this code was originally written) than to build
5887 structure for giving a limitation error. The length of the item that
5888 expands into the list of hosts is now unlimited.
5890 47. The lsearch lookup could not handle data where the length of text line was
5891 more than 4095 characters. Such lines were truncated, leading to shortened
5892 data being returned. It should now handle lines of any length.
5894 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5895 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5898 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5899 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5900 no longer happens in comment lines. A list of replacements is now placed
5901 at the head of all of the source files, except those whose only change is
5902 to replace PERL_COMMAND in the very first #! line.
5904 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5905 messages on the queue, with a bottom-up merge sort, using code contributed
5906 by Michael Haardt. This should make operations like -bp somewhat faster on
5907 large queues. It won't affect queue runners, except when queue_run_in_order
5910 51. Installed eximstats 1.31 in the distribution.
5912 52. Added support for SRV lookups to the dnslookup router.
5914 53. If an ACL referred to $message_body or $message_body_end, the value was not
5915 reset for any messages that followed in the same SMTP session.
5917 54. The store-handling optimization for building very long strings was not
5918 differentiating between the different store pools. I don't think this
5919 actually made any difference in practice, but I've tidied it.
5921 55. While running the routers to verify a sender address, $sender_address
5922 was still set to the sender address. This is wrong, because when routing to
5923 send a bounce to the sender, it would be empty. Therefore, I have changed
5924 it so that, while verifying a sender address, $sender_address is set to <>.
5925 (There is no change to what happens when verifying a recipient address.)
5927 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5928 target A or AAAA records (if not already returned) without resetting the
5929 qualify_single or search_parents options of the DNS resolver. These are
5930 inappropriate in this case because the targets of MX and SRV records must
5931 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5932 target that, when qualified, matched something in the local domain. These
5933 two options are now turned off when doing these lookups.
5935 57. It seems that at least some releases of Reiserfs (which does not have the
5936 concept of a fixed number of inodes) returns zero and not -1 for the
5937 number of available inodes. This interacted badly with check_spool_inodes,
5938 which assumed that -1 was the "no such thing" setting. What I have done is
5939 to check that the total number of inodes is greater than zero before doing
5940 the test of how many are available.
5942 58. When a "warn" ACL statement has a log_message modifier, the message is
5943 remembered, and not repeated. This is to avoid a lot of repetition when a
5944 message has many recipients that cause the same warning to be written.
5945 However, Exim was preserving the list of already written lines for an
5946 entire SMTP session, which doesn't seem right. The memory is now reset if a
5947 new message is started.
5949 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5950 debugging output unless log_rewrite was also set.
5952 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5953 of (int)(handle) when we know that handle contains (void *)(-1).
5955 61. The Exim daemon panic-logs an error return when it closes the incoming
5956 connection. However "connection reset by peer" seems to be common, and
5957 isn't really an error worthy of noting specially, so that particular error
5960 62. When Exim is trying to find all the local interfaces, it used to panic and
5961 die if the ioctl to get the interface flags failed. However, it seems that
5962 on at least one OS (Solaris 9) it is possible to have an interface that is
5963 included in the list of interfaces, but for which you get a failure error
5964 for this call. This happens when the interface is not "plumbed" into a
5965 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5966 failure of the "get flags" call assumes that the interface is down.
5968 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5969 makes life easier for people who are doing arithmetic on fields extracted
5970 from dates, where you often get leading zeros that should not be
5971 interpreted as octal.
5973 64. Added qualify_domain to the redirect router, to override the global
5976 65. If a pathologically long header line contained very many addresses (the
5977 report of this problem mentioned 10 000) and each of them was rewritten,
5978 Exim could use up a very large amount of memory. (It kept on making new
5979 copies of the header line as it rewrote, and never released the old ones.)
5980 At the expense of a bit more processing, the header rewriting function has
5981 been changed so that it no longer eats memory in this way.
5983 66. The generation of the Received: header has been moved from the time that a
5984 message starts to be received, to the time that it finishes. The timestamp
5985 in the Received: header should now be very close to that of the <= log
5986 line. There are two side-effects of this change:
5988 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5989 logged header lines no longer include the local Received: line, because
5990 it has not yet been created. The same applies to a copy of the message
5991 that is returned to a non-SMTP sender when a message is rejected.
5993 (b) When a filter file is tested using -bf, no additional Received: header
5994 is added to the test message. After some thought, I decided that this
5997 This change does not affect the value of $received_for. It is still set
5998 after address rewriting, but before local_scan() is called.
6000 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6002 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6003 gave an unhelpful panic error message, and a defer error. I have managed to
6004 change this behaviour so that it now rejects any supplied certificate,
6005 which seems right, as the list of acceptable certificates is empty.
6007 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6008 gave an unhelpful defer error. I have not managed to make this reject any
6009 supplied certificates, but the error message it gives is "no certificate
6010 supplied", which is not helpful.
6012 70. exigrep's output now also includes lines that are not associated with any
6013 message, but which match the given pattern. Implemented by a patch from
6014 Martin Sluka, which also tidied up the Perl a bit.
6016 71. Recipient callout verification, like sender verification, was using <> in
6017 the MAIL FROM command. This isn't really the right thing, since the actual
6018 sender may affect whether the remote host accepts the recipient or not. I
6019 have changed it to use the actual sender in the callout; this means that
6020 the cache record is now keyed on a recipient/sender pair, not just the
6021 recipient address. There doesn't seem to be a real danger of callout loops,
6022 since a callout by the remote host to check the sender would use <>.
6023 [SEE ABOVE: changed after hitting problems.]
6025 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6026 temporary errors. However, in the case of such a code being given after
6027 the end of a data transmission (i.e. after ".") Exim was failing to write
6028 a retry record for the message. (Yes, there was some broken host that was
6029 actually sending 8xx at this point.)
6031 73. An unknown lookup type in a host list could cause Exim to panic-die when
6032 the list was checked. (An example that provoked this was putting <; in the
6033 middle of a list instead of at the start.) If this happened during a DATA
6034 ACL check, a -D file could be left lying around. This kind of configuration
6035 error no longer causes Exim to die; instead it causes a defer error. The
6036 incident is still logged to the main and panic logs.
6038 74. Buglet left over from Exim 3 conversion. The message "too many messages
6039 in one connection" was written to the rejectlog but not the mainlog, except
6040 when address rewriting (yes!) was being logged.
6042 75. Added write_rejectlog option.
6044 76. When a system filter was run not as root (that is, when system_filter_user
6045 was set), the values of the $n variables were not being returned to the
6046 main process; thus, they were not subsequently available in the $sn
6049 77. Added +return_path_on_delivery log selector.
6051 78. A connection timeout was being treated differently from recipients deferred
6052 when testing hosts_max_try with a message that was older than the host's
6053 retry timeout. (The host should not be counted, thus allowing all hosts to
6054 be tried at least once before bouncing.) This may have been the cause of an
6055 occasionally reported bug whereby a message would remain on the queue
6056 longer than the retry timeout, but would be bounced if a delivery was
6057 forced. I say "may" because I never totally pinned down the problem;
6058 setting up timeout/retry tests is difficult. See also the next item.
6060 79. The ultimate address timeout was not being applied to errors that involved
6061 a combination of host plus message (for example, a timeout on a MAIL
6062 command). When an address resolved to a number of possible hosts, and they
6063 were not all tried for each delivery (e.g. because of hosts_max_try), a
6064 message could remain on the queue longer than the retry timeout.
6066 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6069 81. Fixed an obscure SMTP outgoing bug which required at least the following
6070 conditions: (a) there was another message waiting for the same server;
6071 (b) the server returned 5xx to all RCPT commands in the first message so
6072 that the message was not completed; (c) the server dropped the connection
6073 or gave a negative response to the RSET that Exim sends to abort the
6074 transaction. The observed case was a dropped connection after DATA that had
6075 been sent in pipelining mode. That is, the server had advertised PIPELINING
6076 but was not implementing it correctly. The effect of the bug was incorrect
6077 behaviour, such as trying another host, and this could lead to a crash.
6083 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6084 and daemon.c were passed as pointers to ints; they should have been
6085 pointers to socklen_t variables (which are typically unsigned ints).
6087 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6090 3. Fixed a really odd bug that affected only the testing scheme; patching a
6091 certain fixed string in the binary changed the value of another string that
6092 happened to be identical to the end of the original first string.
6094 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6095 name", it returns that address as the IP address. On some operating
6096 systems (e.g. Solaris), it also passes back the IP address string as the
6097 "host name". However, on others (e.g. Linux), it passes back an empty
6098 string. Exim wasn't checking for this, and was changing the host name to an
6099 empty string, assuming it had been canonicalized.
6101 5. Although rare, it is permitted to have more than one PTR record for a given
6102 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6103 all the names associated with an address, because they do in Solaris.
6104 However, it seems that they do not in Linux for data that comes from the
6105 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6106 I found this out when I moved to a new Linux workstation and tried to run
6107 the Exim test suite.
6109 To get round this problem I have changed the code so that it now does its
6110 own call to the DNS to look up PTR records when searching for a host name.
6111 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6112 addresses that are only in /etc/hosts are still found.
6114 This behaviour is, however, controlled by an option called host_lookup_
6115 order, which defaults to "bydns:byaddr". If people want to use the other
6116 order, or indeed, just use one or the other means of lookup, they can
6117 specify it in this variable.
6119 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6120 some operating systems, this comes back from gethostbyaddr() as an empty
6121 string, and this is what Exim used to test for. However, it seems that in
6122 other systems, "." is yielded. Exim now tests for this case too.
6124 7. The values of check_spool_space and check_log_space are now held internally
6125 as a number of kilobytes instead of an absolute number of bytes. If a
6126 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6127 kilobyte. This means that much larger values can be stored.
6129 8. Exim monitor: an attempt to get the action menu when not actually pointing
6130 at a message produces an empty menu entitled "No message selected". This
6131 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6132 no entries in it ("Shell widget menu has zero width and/or height"). So I
6133 have added a single, blank menu entry in this case.
6135 9. Added ${quote_local_part.
6137 10. MIME decoding is now applied to the contents of Subject: header lines when
6140 11. Now that a reference to $sender_host_address automatically causes a reverse
6141 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6142 host lookup before query-style lookups in lists that might use this
6143 variable. This has therefore been abolished, and the "net-" prefix is no
6144 longer necessary for query-style lookups.
6146 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6147 have been a typo for LFLAGS, so it has been changed.
6149 13. The install script calls Exim with "-C /dev/null" in order to find the
6150 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6151 to be output. However, since Exim outputs its version number before the
6152 error, it didn't break the script. It just looked ugly. I fixed this by
6153 always allowing "-C /dev/null" if the caller is root.
6155 14. Ignore overlarge ACL variable number when reading spool file - insurance
6156 against a later release with more variables having written the file.
6158 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6159 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6160 this, as well as the form without the "IPv6" on the front.
6162 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6163 OS/Makefile-Darwin file.
6165 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6166 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6168 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6169 variable. However, some OS use ints for inodes. I've added cast to long int
6170 to get rid of the compiler warning.
6172 19. I had forgotten to lock out "/../" in configuration file names when
6173 ALT_CONFIG_PREFIX was set.
6175 20. Routers used for verification do not need to specify transports. However,
6176 if such a router generated a host list, and callout was configured, Exim
6177 crashed, because it could not find a port number from the (non-existent)
6178 transport. It now assumes port 25 in this circumstance.
6180 21. Added the -t option to exigrep.
6182 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6183 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6184 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6185 them is set, LOOKUP_LSEARCH is forced.
6187 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6189 24. Added sender and host information to the "rejected by local_scan()" log
6190 line; previously there was no indication of these.
6192 25. Added .include_if_exists.
6194 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6195 Linux. It turns out that not all file systems support this. Apparently some
6196 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6197 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6198 not supported on the file descriptor, is now ignored when Exim is trying to
6199 sync a directory. This applies only to Linux.
6201 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6203 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6204 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6207 29. Added support for maildirsize files from supplied patch (modified a bit).
6209 30. The use of :fail: followed by an empty string could lead Exim to respond to
6210 sender verification failures with (e.g.):
6212 550 Verification failed for <xxx>
6213 550 Sender verify failed
6215 where the first response line was missing the '-' that indicates it is not
6216 the final line of the response.
6218 31. The loop for finding the name of the user that called Exim had a hardwired
6219 limit of 10; it now uses the value of finduser_retries, which is used for
6220 all other user lookups.
6222 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6225 33. Exim was neglecting to zero errno before one call of strtol() when
6226 expanding a string and expecting an integer value. On some systems this
6227 resulted in spurious "integer overflow" errors. Also, it was casting the
6228 result into an int without checking.
6230 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6231 did not work. The code looks as if it has *never* worked, though it appears
6232 to have been documented since at least release 1.62. I have made it work.
6234 35. The "timeout_DNS" error in retry rules, also documented since at least
6235 1.62, also never worked. As it isn't clear exactly what this means, and
6236 clearly it isn't a major issue, I have abolished the feature by treating it
6237 as "timeout", and writing a warning to the main and panic logs.
6239 36. The display of retry rules for -brt wasn't always showing the error code
6242 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6243 timeout_connect_A, timeout_connect_MX.
6245 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6246 to the empty sender.
6248 39. The daemon was not analysing the content of -oX till after it had closed
6249 stderr and disconnected from the controlling terminal. This meant that any
6250 syntax errors were only noted on the panic log, and the return code from
6251 the command was 0. By re-arranging the code a little, I've made the
6252 decoding happen first, so such errors now appear on stderr, and the return
6253 code is 1. However, the actual setting up of the sockets still happens in
6254 the disconnected process, so errors there are still only recorded on the
6257 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6258 connections (as happens on some IP stacks) was logged at start up time as
6259 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6260 it from "IPv6 and IPv4", which means that two separate sockets are being
6263 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6264 says whether AF_INET or AF_INET6 was passed as an argument.
6266 42. Exiwhat output was messed up when time zones were included in log
6269 43. Exiwhat now gives more information about the daemon's listening ports,
6270 and whether -tls-on-connect was used.
6272 44. The "port" option of the smtp transport is now expanded.
6274 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6275 silently ignored. Now an error message is written to the main and panic
6278 46. There's a new ACL modifier called "logwrite" which writes to a log file
6279 as soon as it is encountered.
6281 47. Added $local_user_uid and $local_user_gid at routing time.
6283 48. Exim crashed when trying to verify a sender address that was being
6286 49. Exim was recognizing only a space character after ".include". It now also
6287 recognizes a tab character.
6289 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6290 extracting the relevant information from the specification. The man page no
6291 longer contains scrambled data for the -d option, and I've added a section
6292 at the front about calling Exim under different names.
6294 51. Added "extra_headers" argument to the "mail" command in filter files.
6296 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6299 53. Installed eximstats 1.29.
6301 54. Added transport_filter_timeout as a generic transport option.
6303 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6304 Cc: header lines. This was required by RFC 822, but it not required by RFC
6307 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6308 incoming messages that did not have them. Now it does so only if the
6309 message originates locally, that is, if there is no associated remote host
6310 address. When Resent- header lines are present, this applies to the Resent-
6311 lines rather than the non-Resent- lines.
6313 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6314 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6316 58. Messages for configuration errors now include the name of the main
6317 configuration file - useful now that there may be more than one file in a
6318 list (.included file names were always shown).
6320 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6321 for those rare installations that do not start the daemon as root or run it
6322 setuid root. I've cut out the call to initgroups() if the daemon is not
6325 60. The Exim user and group can now be bound into the binary as text strings
6326 that are looked up at the start of Exim's processing.
6328 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6330 62. Added $mailstore_basename variable.
6332 63. Installed patch to sieve.c from Michael Haardt.
6334 64. When Exim failed to open the panic log after failing to open the main log,
6335 the original message it was trying to log was written to stderr and debug
6336 output, but if they were not available (the usual case in production), it
6337 was lost. Now it is written to syslog before the two lines that record the
6338 failures to open the logs.
6340 65. Users' Exim filters run in subprocesses under the user's uid. It is
6341 possible for a "deliver" command or an alias in a "personal" command to
6342 provoke an address rewrite. If logging of address rewriting is configured,
6343 this fails because the process is not running as root or exim. There may be
6344 a better way of dealing with this, but for the moment (because 4.30 needs
6345 to be released), I have disabled address rewrite logging when running a
6346 filter in a non-root, non-exim process.
6352 1. The buildconfig auxiliary program wasn't quoting the value set for
6353 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6354 not defined. This bug was masked in 4.22 by the effect that was fixed in
6357 2. Some messages that were rejected after a message id was allocated were
6358 shown as "incomplete" by exigrep. It no longer does this for messages that
6359 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6361 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6362 have allow_domain_literals set, the ID did not get logged in the <= line.
6363 Domain literals are now always recognized in Message-ID: header lines.
6365 4. The first argument for a ${extract expansion item is the key name or field
6366 number. Leading and trailing spaces in this item were not being ignored,
6367 causing some misleading effects.
6369 5. When deliver_drop_privilege was set, single queue runner processes started
6370 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6371 same command in the process it spins off) were not dropping privilege.
6373 6. When the daemon running as "exim" started a queue runner, it always
6374 re-executed Exim in the spun-off process. This is a waste of effort when
6375 deliver_drop_privilege is set. The new process now just calls the
6376 queue-runner function directly.
6382 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6385 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6386 ignored. Though the use of -oP was forcing the writing of a pid file, it
6387 was always written to the default place.
6389 3. If the message "no IP address found for host xxxx" is generated during
6390 incoming verification, it is now followed by identification of the incoming
6391 connection (so you can more easily find what provoked it).
6393 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6395 5. Added some features to "harden" Exim a bit more against certain attacks:
6397 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6398 be put in Local/Makefile. This is like the never_users runtime option,
6399 but it cannot be overridden. The default setting is "root".
6401 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6402 prefix string with which any file named in a -C command line option
6405 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6406 is retained for -C and -D only if the caller of Exim is root. Without
6407 it, the exim user may also use -C and -D and retain privilege.
6409 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6410 command line option is disabled.
6412 6. Macro names set by the -D option must start with an upper case letter, just
6413 like macro names defined in the configuration file.
6415 7. Added "dereference=" facility to LDAP.
6417 8. Two instances of the typo "uknown" in the source files are fixed.
6419 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6420 the Configure-Makefile script screwed up while processing it.
6422 10. Incorporated PCRE 4.4.
6424 11. The SMTP synchronization check was not operating right at the start of an
6425 SMTP session. For example, it could not catch a HELO sent before the client
6426 waited for the greeting. There is now a check for outstanding input at the
6427 point when the greeting is written. Because of the duplex, asynchronous
6428 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6429 way, but not yet received, when the check is performed.
6431 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6432 on TCP/IP sockets, because this apparently causes some broken clients to
6435 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6436 unchanged) from the Cygwin maintainer.
6438 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6439 was supported instead of testing for mbx. Effectively a typo.
6441 15. The spa authenticator server code was not checking that the input it
6442 received was valid base64.
6444 16. The debug output line for the "set" modifier in ACLs was not showing the
6445 name of the variable that was being set.
6447 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6449 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6450 lookup on its own. Something else was needed to trigger the lookup. For
6451 example, a match in host_lookup or the need for a host name in a host list.
6452 Now, if $sender_host_name is referenced and the host name has not yet been
6453 looked up, a lookup is performed. If the lookup fails, the variable remains
6454 empty, and $host_lookup_failed is set to "1".
6456 19. Added "eqi" as a case-independent comparison operator.
6458 20. The saslauthd authentication condition could segfault if neither service
6459 nor realm was specified.
6461 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6462 error message that was logged was misleading, and incoming SMTP
6463 connections were dropped. The message is now more accurate, and temporary
6464 errors are given to SMTP connections.
6466 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6467 (except RSET) with a 421 error, until QUIT is received. However, it was
6468 failing to send a response to QUIT.
6470 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6471 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6472 ACL is now run after the helo_try_verify_hosts code.
6474 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6475 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6476 case-independent, so other case variants are also recognized). Apparently
6477 some systems use these upper case variants.
6479 25. If more than two messages were waiting for the same host, and a transport
6480 filter was specified for the transport, Exim sent two messages over the
6481 same TCP/IP connection, and then failed with "socket operation on non-
6482 socket" when it tried to send the third.
6484 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6486 27. The extern definition of crypt16() in expand.c was not being excluded when
6487 the OS had its own crypt16() function.
6489 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6490 as a preferred synonym for return_size_limit, both as an option and as an
6493 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6495 30. Changed the default configuration ACL to relax the local part checking rule
6496 for addresses that are not in any local domains. For these addresses,
6497 slashes and pipe symbols are allowed within local parts, but the sequence
6498 /../ is explicitly forbidden.
6500 31. SPA server authentication was not clearing the challenge buffer before
6503 32. log_message in a "warn" ACL statement was writing to the reject log as
6504 well as to the main log, which contradicts the documentation and doesn't
6505 seem right (because no rejection is happening). So I have stopped it.
6507 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6508 However, I am unable to do any testing of this.
6510 34. Fixed an infelicity in the appendfile transport. When checking directories
6511 for a mailbox, to see if any needed to be created, it was accidentally
6512 using path names with one or more superfluous leading slashes; tracing
6513 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6515 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6516 contents are added to the log line that is written for every discarded
6517 recipient. (Previously a log_message setting was ignored.)
6519 36. The ${quote: operator now quotes the string if it is empty.
6521 37. The install script runs exim in order to find its version number. If for
6522 some reason other than non-existence or emptiness, which it checks, it
6523 could not run './exim', it was installing it with an empty version number,
6524 i.e. as "exim-". This error state is now caught, and the installation is
6527 38. An argument was missing from the function that creates an error message
6528 when Exim fails to connect to the socket for saslauthd authentication.
6529 This could cause Exim to crash, or give a corrupted message.
6531 39. Added isip, isip4, and isip6 to ${if conditions.
6533 40. The ACL variables $acl_xx are now saved with the message, and can be
6534 accessed later in routers, transports, and filters.
6536 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6537 strings in the file are not string-expanded.
6539 42. If a MAIL command specified a SIZE value that was too large to fit into an
6540 int variable, the check against message_size_limit failed. Such values are
6541 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6542 day this will have to be increased, but I don't think I want to be around
6543 when emails are that large.
6550 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6551 iconv() is not standard in FreeBSD.
6553 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6554 IPv6 enabled. The observed symptom was a segmentation fault on return from
6555 the function os_common_find_running_interfaces() in src/os.c.
6557 3. In the check_special_case() function in daemon.c I had used "errno" as an
6558 argument name, which causes warnings on some systems. This was basically a
6559 typo, since it was named "eno" in the comments!
6561 4. The code that waits for the clock to tick (at a resolution of some fraction
6562 of a second) so as to ensure message-id uniqueness was always waiting for
6563 at least one whole tick, when it could have waited for less. [This is
6564 almost certainly not relevant at current processor speeds, where it is
6565 unlikely to ever wait at all. But we try to future-proof.]
6567 5. The function that sleeps for a time interval that includes fractions of a
6568 second contained a race. It did not block SIGALRM between setting the
6569 timer, and suspending (a couple of lines later). If the interval was short
6570 and the sigsuspend() was delayed until after it had expired, the suspension
6571 never ended. On busy systems this could lead to processes getting stuck for
6574 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6575 process, before it forks any delivery processes. The open lookup caching
6576 mechanism meant that the open file or database connection was passed into
6577 the delivery process. The problem was that delivery processes always tidy
6578 up cached lookup data. This could cause a problem for the next delivery
6579 process started by the queue runner, because the external queue runner
6580 process does not know about the closure. So the next delivery process
6581 still has data in the lookup cache. In the case of a file lookup, there was
6582 no problem because closing a file descriptor in a subprocess doesn't affect
6583 the parent. However, if the lookup was caching a connection to a database,
6584 the connection was closed, and the second delivery process was likely to
6585 see errors such as "PGSQL: query failed: server closed the connection
6586 unexpectedly". The problem has been fixed by closing all cached lookups
6587 in a queue runner before running a delivery process.
6589 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6590 seem to have the "const" qualifier which it has on other OS. I've
6593 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6594 *of the same type* (client or server) with the same public name that an
6595 error should be diagnosed.
6597 9. When Exim looked up a host name for an IP address, but failed to find the
6598 original IP address when looking up the host name (a safety check), it
6599 output the message "<ip address> does not match any IP for NULL", which was
6600 confusing, to say the least. The bug was that the host name should have
6601 appeared instead of "NULL".
6603 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6604 user that is built into the binary, and the -C or -D options is used, root
6605 privilege is dropped before the configuration file is read. In addition,
6606 logging is switched to stderr instead of the normal log files. If the
6607 configuration then re-defines the Exim user, the unprivileged environment
6608 is probably not what is expected, so Exim logs a panic warning message (but
6611 However, if deliver_drop_privilege is set, the unprivileged state may well
6612 be exactly what is intended, so the warning has been cut out in that case,
6613 and Exim is allowed to try to write to its normal log files.
6619 1. smtp_return_error_details was not giving details for temporary sender
6620 or receiver verification errors.
6622 2. Diagnose a configuration error if two authenticators have the same public
6625 3. Exim used not to create the message log file for a message until the first
6626 delivery attempt. This could be confusing when incoming messages were held
6627 for policy or load reasons. The message log file is now created at the time
6628 the message is received, and an initial "Received" line is written to it.
6630 4. The automatically generated man page for command line options had a minor
6631 bug that caused no ill effects; however, a more serious problem was that
6632 the procedure for building the man page automatically didn't always
6633 operate. Consequently, release 4.20 contains an out-of-date version. This
6634 shouldn't happen again.
6636 5. When building Exim with embedded Perl support, the script that builds the
6637 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6638 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6640 6. The freeze_tell option was not being used for messages that were frozen on
6641 arrival, either by an ACL or by local_scan().
6643 7. Added the smtp_incomplete_transaction log selector.
6645 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6646 was accepting AUTH without a new EHLO.
6648 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6649 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6652 10. Logging of TCP/IP connections (when configured) now happens in the main
6653 daemon process instead of the child process, so that the TCP/IP connection
6654 count is more accurate (but it can never be perfect).
6656 11. The use of "drop" in a nested ACL was not being handled correctly in the
6657 outer ACL. Now, if condition failure induced by the nested "drop" causes
6658 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6659 or "require"), the connection is dropped.
6661 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6662 that yield "discard" can now be used with an "accept" or a "discard" verb,
6663 but an error is generated for any others (because I can't see a useful way
6664 to define what should happen).
6666 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6667 are now processed in the same way as lines in the Exim configuration file.
6668 In particular, continuation lines are supported.
6670 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6672 15. Added -ti meaning -t -i.
6674 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6675 domains, and warn by logging if others are found.
6677 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6678 returned from GIFCONF when Exim is trying to find the list of interfaces on
6679 a host. The code in os.c has been modified to copy each ifreq to an aligned
6680 structure in all cases.
6682 Also, in some cases, the returned ifreq's were being copied to a 'struct
6683 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6684 means the last couple of bytes of an IPv6 address could be chopped if the
6685 ifreq contained only a normal sockaddr (14 bytes storage).
6687 18. Named domain lists were not supported in the hosts_treat_as_local option.
6688 An entry such as +xxxx was not recognized, and was treated as a literal
6691 19. Ensure that header lines added by a DATA ACL are included in the reject log
6692 if the ACL subsequently rejects the message.
6694 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6695 MD5 (which is deprecated).
6697 21. When testing a filter file using -bf, Exim was writing a message when it
6698 took the sender from a "From " line in the message, but it was not doing so
6699 when it took $return_path from a Return-Path: header line. It now does.
6701 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6702 with a valid header line field name (a series of printing characters
6703 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6705 23. Changed "disc" in the source to "disk" to conform to the documentation and
6706 the book and for uniformity.
6708 24. Ignore Sendmail's -Ooption=value command line item.
6710 25. When execve() failed while trying to run a command in a pipe transport,
6711 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6712 could be confused with a return value of 69 from the command itself. This
6713 has been changed to 127, the value the shell returns if it is asked to run
6714 a non-existent command. The wording for the related log line suggests a
6715 non-existent command as the problem.
6717 26. If received_header_text expands to an empty string, do not add a Received:
6718 header line to the message. (Well, it adds a token one on the spool, but
6719 marks it "old" so that it doesn't get used or transmitted.)
6721 27. Installed eximstats 1.28 (addition of -nt option).
6723 28. There was no check for failure on the call to getsockname() in the daemon
6724 code. This can fail if there is a shortage of resources on the system, with
6725 ENOMEM, for example. A temporary error is now given on failure.
6727 29. Contrary to the C standard, it seems that in some environments, the
6728 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6729 program. Exim now does this explicitly; it affects the formatting of
6730 timestamps using strftime().
6732 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6733 complaints. I've now initialized all the variables, to avoid this.
6735 32. Header lines added by a system filter were not being "seen" during
6736 transport-time rewrites.
6738 33. The info_callback() function passed to OpenSSL is set up with type void
6739 (*)(SSL *, int, int), as described somewhere. However, when calling the
6740 function (actually a macro) that sets it up, the type void(*)() is
6741 expected. I've put in a cast to prevent warnings from picky compilers.
6743 34. If a DNS black list lookup found a CNAME record, but there were no A
6744 records associated with the domain it pointed at, Exim crashed.
6746 35. If a DNS black list lookup returned more than one A record, Exim ignored
6747 all but the first. It now scans all returned addresses if a particular IP
6748 value is being sought. In this situation, the contents of the
6749 $dnslist_value variable are a list of all the addresses, separated by a
6752 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6753 to accept a host name and all its aliases if the forward lookup for any of
6754 them yielded the IP address of the incoming connection. Now it accepts only
6755 those names whose forward lookup yields the correct IP address. Any other
6756 names are discarded. This closes a loophole whereby a rogue DNS
6757 administrator could create reverse DNS records to break through a
6758 wildcarded host restriction in an ACL.
6760 37. If a user filter or a system filter that ran in a subprocess used any of
6761 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6762 the wrong values were passed to the pipe command ($thisaddress had the
6763 value of $0, $0 had the value of $1, etc). This bug was introduced by
6764 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6766 38. Improved the line breaking for long SMTP error messages from ACLs.
6767 Previously, if there was no break point between 40 and 75 characters, Exim
6768 left the rest of the message alone. Two changes have been made: (a) I've
6769 reduced the minimum length to 35 characters; (b) if it can't find a break
6770 point between 35 and 75 characters, it looks ahead and uses the first one
6771 that it finds. This may give the occasional overlong line, but at least the
6772 remaining text gets split now.
6774 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6775 file descriptors might be low, and that setting 1000 would always raise it.
6776 It turns out that in some environments, the limit is already over 1000 and
6777 that lowering it causes trouble. So now Exim takes care not to decrease it.
6779 40. When delivering a message, the value of $return_path is set to $sender_
6780 address at the start of routing (routers may change the value). By an
6781 oversight, this default was not being set up when an address was tested by
6782 -bt or -bv, which affected the outcome if any router or filter referred to
6785 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6786 log only when "message" or "log_message" are set. However, if one of the
6787 conditions was an address verification, or a call to a nested ACL, the
6788 messages generated by the underlying test were being passed through. This
6789 no longer happens. The underlying message is available in $acl_verify_
6790 message for both "message" and "log_message" expansions, so it can be
6791 passed through if needed.
6793 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6794 new expansion $bh_ to give the encoded byte string without charset
6795 translation. Translation happens only if iconv() is available; HAVE_ICONV
6796 indicates this at build time. HEADERS_CHARSET gives the charset to
6797 translate to; headers_charset can change it in the configuration, and
6798 "headers charset" can change it in an individual filter file.
6800 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6801 that creates RFC 2047 encoded "words" labels them as that charset instead
6802 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6803 expansion operator; (ii) when Exim creates a From: line for a local
6804 message; (iii) when a header line is rewritten to include a "phrase" part.
6806 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6807 buggy, causing it to skip the first lines of messages whose message ID
6808 ended in 'D'. This would not have bitten before Exim release 4.14, because
6809 message IDs were unlikely to end in 'D' before then. The effect was to have
6810 incorrect size information for certain domains.
6812 45. #include "config.h" was missing at the start of the crypt16.c module. This
6813 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6816 46. If there was a timeout during a "random" callout check, Exim treated it as
6817 a failure of the random address, and carried on sending RSET and the real
6818 address. If the delay was just some slowness somewhere, the response to the
6819 original RCPT would be taken as a response to RSET and so on, causing
6820 mayhem of various kinds.
6822 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6823 when I implemented it. It didn't allow for the fact that some option values
6824 may legitimately be negative (e.g. size_addition), and it didn't even do
6825 the right test for positive values.
6827 48. Domain names in DNS records are case-independent. Exim always looks them up
6828 in lower case. Some resolvers return domain names in exactly the case they
6829 appear in the zone file, that is, they may contain uppercase letters. Not
6830 all resolvers do this - some return always lower case. Exim was treating a
6831 change of case by a resolver as a change of domain, similar to a widening
6832 of a domain abbreviation. This triggered its re-routing code and so it was
6833 trying to route what was effectively the same domain again. This normally
6834 caused routing to fail (because the router wouldn't handle the domain
6835 twice). Now Exim checks for this case specially, and just changes the
6836 casing of the domain that it ultimately uses when it transmits the message
6839 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6842 50. If a filter generated a file delivery with a non-absolute name (possible if
6843 no home directory exists for the router), the forbid_file option was not
6846 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6847 the existing equality matching.
6849 52. Exim was using ints instead of ino_t variables in some places where it was
6850 dealing with inode numbers.
6852 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6853 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6854 called TMPDIR, and if it finds it is different, it changes its value.
6856 54. The smtp_printf() function is now made available to local_scan() so
6857 additional output lines can be written before returning. There is also an
6858 smtp_fflush() function to enable the detection of a dropped connection.
6859 The variables smtp_input and smtp_batched_input are exported to
6862 55. Changed the default runtime configuration: the message "Unknown user"
6863 has been removed from the ACL, and instead placed on the localuser router,
6864 using the cannot_route_message feature. This means that any verification
6865 failures that generate their own messages won't get overridden. Similarly,
6866 the "Unrouteable address" message that was in the ACL for unverifiable
6867 relay addresses has also been removed.
6869 56. Added hosts_avoid_esmtp to the smtp transport.
6871 57. The exicyclog script was not checking for the esoteric option
6872 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6873 will work only if exicyclog is run under the appropriate euid.
6875 58. Following a discussion on the list, the rules by which Exim recognises line
6876 endings on incoming messages have been changed. The -dropcr and drop_cr
6877 options are now no-ops, retained only for backwards compatibility. The
6878 following line terminators are recognized: LF CRLF CR. However, special
6879 processing applies to CR:
6881 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6882 nor a local message in the state where . is a terminator.
6884 (ii) If a bare CR is encountered in a header line, an extra space is added
6885 after the line terminator so as not to end the header. The reasoning
6886 behind this is that bare CRs in header lines are most likely either
6887 to be mistakes, or people trying to play silly games.
6889 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6890 was being incorrectly given as 18 bytes larger than it should have been.
6891 This is a VOB (very old bug).
6893 60. This may never have affected anything current, but just in case it has:
6894 When the local host is found other than at the start of a list of hosts,
6895 the local host, those with the same MX, and any that follow, are discarded.
6896 When the list in question was part of a longer list of hosts, the following
6897 hosts (not currently being processed) were also being discarded. This no
6898 longer happens. I'm not sure if this situation could ever has previously
6901 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6904 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6905 header. This is something that is recommended in a new Internet Draft, and
6906 is something that is documented as being done by Sendmail. There are two
6907 possible values. For messages generated by the autoreply transport, Exim
6910 Auto-Submitted: auto-replied
6912 whereas for all other generated messages (e.g. bounces) it adds
6914 Auto-Submitted: auto-generated
6916 63. The "personal" condition in filters now includes a test for the
6917 Auto-Submitted: header. If it contains the string "auto-" the message it
6918 not considered personal.
6920 64. Added rcpt_include_affixes as a generic transport option.
6922 65. Added queue_only_override (default true).
6924 66. Added the syslog_duplication option.
6926 67. If what should have been the first header line of a message consisted of
6927 a space followed by a colon, Exim was mis-interpreting it as a header line.
6928 It isn't of course - it is syntactically invalid and should therefore be
6929 treated as the start of the message body. The misbehaviour could have
6930 caused a number of strange effects, including loss of data in subsequent
6931 header lines, and spool format errors.
6933 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6934 client host had authenticated. This control can now be exercised by an ACL
6935 for more flexibility.
6937 69. By default, callouts do not happen when testing with -bh. There is now a
6938 variant, -bhc, which does actually run the callout code, including
6939 consulting and updating the callout cache.
6941 70. Added support for saslauthd authentication, courtesy of Alexander
6944 71. If statvfs() failed on the spool or log directories while checking their
6945 size for availability, Exim confusingly gave the error "space shortage".
6946 Furthermore, in debugging mode it crashed with a floating point exception.
6947 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6948 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6949 serious problem, Exim now writes to the main and panic logs when this
6950 happens, with details of the failure. It then refuses to accept the
6951 incoming message, giving the message "spool directory problem" or "log
6952 directory problem" with a 421 code for SMTP messages.
6954 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6955 0, 1, and 2 exist, because some OS complain for execs without them (see
6956 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6957 descriptors. However, the code omitted to check that the open succeeded,
6958 causing mysterious errors if for some reason the permissions on /dev/null
6959 got screwed. Now Exim writes a message to the main and panic logs, and
6960 bombs out if it can't open /dev/null.
6962 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6963 interact so that it is all more flexible. It is supposed to remain
6964 backwards compatible. Also added extra_local_interfaces.
6966 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6967 to bomb out with an assertion failure - to the client this appears as a
6968 connection drop. This problem occurs in the part of the code that was taken
6969 from the Samba project. Fortunately, the assertion is in a very simple
6970 function, so I have fixed this by reproducing the function inline in the
6971 one place where it is called, and arranging for authentication to fail
6972 instead of killing the process with assert().
6974 75. The SPA client code was not working when the server requested OEM rather
6975 than Unicode encoding.
6977 76. Added code to make require_files with a specific uid setting more usable in
6978 the case where statting the file as root fails - usually a non-root-mounted
6979 NFS file system. When this happens and the failure is EACCES, Exim now
6980 forks a subprocess and does the per-uid checking as the relevant uid.
6982 77. Added process_log_path.
6984 78. If log_file_path was not explicitly set, a setting of check_log_space or
6985 check_log_inodes was ignored.
6987 79. If a space check for the spool or log partitions fails, the incident is now
6988 logged. Of course, in the latter case the data may get lost...
6990 80. Added the %p formatting code to string_format() so that it can be used to
6991 print addresses in debug_print(). Adjusted all the address printing in the
6992 debugging in store.c to use %p rather than %d.
6994 81. There was a concern that a line of code in smtp_in.c could overflow a
6995 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6996 initially expressed, the concern was not well-founded, because trailing
6997 spaces are removed early. However, if the trailing spaces were followed by
6998 a NULL, they did not get removed, so the overflow was possible. Two fixes
7001 (a) I re-wrote the offending code in a cleaner fashion.
7002 (b) If an incoming SMTP command contains a NULL character, it is rejected
7005 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7006 runs initgroups(), so that if the Exim user is in any additional groups,
7007 they will be used during message reception.
7013 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob