1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/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.
55 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
56 a timeout on read on a GnuTLS initiating connection, resulting in the
57 initiating connection being dropped. This mattered most when the callout
58 was marked defer_ok. Fix to keep the two timeout-detection methods
61 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
63 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
64 metadata, resulting in a crash in free().
66 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
67 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
68 Broken also in d185889f4, with init system revamp.
74 JH/01 Rework error string handling in TLS interface so that the caller in
75 more cases is responsible for logging. This permits library-sourced
76 string to be attached to addresses during delivery, and collapses
77 pairs of long lines into single ones.
79 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
80 during configuration. Wildcards are allowed and expanded.
82 JH/02 Rework error string handling in DKIM to pass more info back to callers.
83 This permits better logging.
85 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
86 do not close it down and have the child transport start it up again on
87 the passed-on TCP connection. Instead, proxy the child (and any
88 subsequent ones) for TLS via a unix-domain socket channel. Logging is
89 affected: the continued delivery log lines do not have any DNSSEC, TLS
90 Certificate or OCSP information. TLS cipher information is still logged.
92 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
93 identical IP addresses on different listening ports. Will also affect
96 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
97 add noisy ifdef guards to special-case this sillyness.
98 Patch from Bernd Kuhls.
100 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
101 than 255 are no longer allowed.
103 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
104 Disable the session-cache too, which might reduce our load. Since we
105 currrectly use a new context for every connection, both as server and
106 client, there is no benefit for these.
107 GnuTLS appears to not support tickets server-side by default (we don't
108 call gnutls_session_ticket_enable_server()) but client side is enabled
109 by default on recent versions (3.1.3 +) unless the PFS priority string
112 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
113 <https://reproducible-builds.org/specs/source-date-epoch/>.
115 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
116 the check for any unsuccessful recipients did not notice the limit, and
117 erroneously found still-pending ones.
119 JH/08 Pipeline CHUNKING command and data together, on kernels that support
120 MSG_MORE. Only in-clear (not on TLS connections).
122 JH/09 Avoid using a temporary file during transport using dkim. Unless a
123 transport-filter is involved we can buffer the headers in memory for
124 creating the signature, and read the spool data file once for the
125 signature and again for transmission.
127 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
128 4.77 as the kernel support then wasn't solid, having issues in 64bit
129 mode. Now, it's been long enough. Add support for FreeBSD also.
131 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
132 case where the routing stage had gathered several addresses to send to
133 a host before calling the transport for the first, we previously failed
134 to close down TLS in the old transport process before passing the TCP
135 connection to the new process. The new one sent a STARTTLS command
136 which naturally failed, giving a failed delivery and bloating the retry
137 database. Investigation and fix prototype from Wolfgang Breyha.
139 JH/12 Fix check on SMTP command input synchronisation. Previously there were
140 false-negatives in the check that the sender had not preempted a response
141 or prompt from Exim (running as a server), due to that code's lack of
142 awareness of the SMTP input buffering.
144 PP/04 Add commandline_checks_require_admin option.
145 Exim drops privileges sanely, various checks such as -be aren't a
146 security problem, as long as you trust local users with access to their
147 own account. When invoked by services which pass untrusted data to
148 Exim, this might be an issue. Set this option in main configuration
149 AND make fixes to the calling application, such as using `--` to stop
152 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
153 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
154 responses to those, into a single TLS record each way (this usually means
155 a single packet). As a side issue, smtp_enforce_sync now works on TLS
158 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
159 affects you only if you're dancing at the edge of the param size limits.
160 If you are, and this message makes sense to you, then: raise the
161 configured limit or use OpenSSL 1.1. Nothing we can do for older
164 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
165 cmdline element to get the documented default one. Previously it was
168 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
169 are unsuable for non-message-specific reasons - eg. connection timeout,
170 banner-time rejection.
172 JH/16 Fix logging of delivery remote port, when specified by router, under
175 PP/06 Repair manualroute's ability to take options in any order, even if one
176 is the name of a transport.
179 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
181 JH/17 Change the list-building routines interface to use the expanding-string
182 triplet model, for better allocation and copying behaviour.
184 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
185 Previously it was constructed the first time a possibly-matching string
186 was met in the configuration file input during startup; now it is done
189 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
190 compatible one, to avoid the (poorly documented) possibility of a config
191 file in the working directory redirecting the DB files, possibly correpting
192 some existing file. CVE-2017-10140 assigned for BDB.
194 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
195 cache-hot. Previously, although the result was properly cached, the
196 initial verify call returned a defer.
198 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
199 the main verify for receipient in uncached-mode.
201 JH/22 Retire historical build files to an "unsupported" subdir. These are
202 defined as "ones for which we have no current evidence of testing".
204 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
205 if present. Previously it was ignored.
207 JH/24 Start using specified-initialisers in C structure init coding. This is
208 a C99 feature (it's 2017, so now considered safe).
210 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
211 if was a fixed-sized field and bitmask ops via macros; it is now more
214 PP/07 GitHub PR 56: Apply MariaDB build fix.
215 Patch provided by Jaroslav Škarvada.
217 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
218 during Coverity cleanups [4.87 JH/47]
219 Diagnosis and fix provided by Michael Fischer v. Mollard.
221 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
222 the right size to place the terminating semicolon on its own folded
223 line, the header hash was calculated to an incorrect value thanks to
224 the (relaxed) space the fold became.
226 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
227 and confused the parent.
229 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
230 which could crash as a result. This could lead to undeliverable messages.
232 JH/28 Logging: "next input sent too soon" now shows where input was truncated
235 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
236 matters on fast-turnover and PID-randomising systems, which were getting
237 out-of-order delivery.
239 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
240 a possibly-overlapping copy. The symptom was that "Remote host closed
241 connection in response to HELO" was logged instead of the actual 4xx
244 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
245 Previously only that bufferd was discarded, resulting in SYMTP command
248 JH/32 DKIM: when a message has multiple signatures matching an identity given
249 in dkim_verify_signers, run the dkim acl once for each. Previously only
250 one run was done. Bug 2189.
252 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
253 "panic the current process" to "deliberately defer". The panic log is
254 still written with the problem list name; the mail and reject logs now
255 get a temp-reject line for the message that was being handled, saying
256 something like "domains check lookup or other defer". The SMTP 451
257 message is still "Temporary local problem".
259 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
260 A crafted sequence of BDAT commands could result in in-use memory beeing
261 freed. CVE-2017-16943.
263 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
264 from SMTP input. Previously it was always done; now only done for DATA
265 and not BDAT commands. CVE-2017-16944.
267 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
268 to the message (such as an overlong header line). Previously this was
269 not done and we did not exit BDAT mode. Followon from the previous item
270 though a different problem.
276 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
277 than -2003 did; needs libidn2 in addition to libidn.
279 JH/02 The path option on a pipe transport is now expanded before use.
281 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
282 Patch provided by "Björn", documentation fix added too.
284 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
285 missing a wire-to-host endian conversion.
287 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
288 close after a BDAT command line could be taken as a following command,
289 giving a synch failure. Fix by only checking for synch immediately
290 before acknowledging the chunk.
292 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
293 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
294 macro. Patches provided by Josh Soref.
296 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
297 Previously we did not; the RFC seems ambiguous and VRFY is not listed
298 by IANA as a service extension. However, John Klensin suggests that we
301 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
302 the dkim code may be unix-mode line endings rather than smtp wire-format
303 CRLF, so prepend a CR to any bare LF.
305 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
306 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
308 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
309 the first were themselves being wrongly included in the feed into dkim
310 processing; with most chunk sizes in use this resulted in an incorrect
311 body hash calculated value.
313 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
314 DKIM signature block, for verification. Although advised against by
315 standards it is specifically not ruled illegal.
317 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
319 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
320 missing a body hash (the bh= tag).
322 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
323 It seems that HAProxy sends the Proxy Protocol information in clear and
324 only then does a TLS startup, so do the same.
326 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
327 TCP connections (such as for Spamd) unless the daemon successfully set
328 Fast Open mode on its listening sockets. This fixes breakage seen on
329 too-old kernels or those not configured for Fast Open, at the cost of
330 requiring both directions being enabled for TFO, and TFO never being used
331 by non-daemon-related Exim processes.
333 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
334 endings, at least on the first header line. Try to canonify any that get
335 past that check, despite the cost.
337 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
338 now limited to an arbitrary five deep, while parsing addresses with the
339 strip_excess_angle_brackets option enabled.
341 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
342 instead leave the unprompted TLS handshake in socket buffer for the
343 TLS library to consume.
345 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
347 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
349 JH/16 Drop variables when they go out of scope. Memory management drops a whole
350 region in one operation, for speed, and this leaves assigned pointers
351 dangling. Add checks run only under the testsuite which checks all
352 variables at a store-reset and panics on a dangling pointer; add code
353 explicitly nulling out all the variables discovered. Fixes one known
354 bug: a transport crash, where a dangling pointer for $sending_ip_address
355 originally assigned in a verify callout, is re-used.
357 PP/06 Drop '.' from @INC in various Perl scripts.
359 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
361 PP/08 Reduce a number of compilation warnings under clang; building with
362 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
363 should be warning-free.
365 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
367 HS/01 Fix portability problems introduced by PP/08 for platforms where
368 realloc(NULL) is not equivalent to malloc() [SunOS et al].
370 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
371 chunk. This allows us to accept broken chunked messages. We need a more
372 general solution here.
374 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
375 already-broken messages in the queue.
377 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
379 JH/19 Fix reference counting bug in routing-generated-address tracking.
385 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
386 supports it and a size is available (ie. the sending peer gave us one).
388 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
389 years of being deprecated). The replacements are the ACLs
390 acl_smtp_mime and acl_not_smtp_mime.
392 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
393 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
394 or even an in-clear connection were permitted. Now, if the host lookup
395 was dnssec and dane was requested then the host is only used if the
396 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
397 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
398 if one fails this test.
399 This means that a poorly-configured remote DNS will make it incommunicado;
400 but it protects against a DNS-interception attack on it.
402 JH/04 Bug 1810: make continued-use of an open smtp transport connection
403 non-noisy when a race steals the message being considered.
405 JH/05 If main configuration option tls_certificate is unset, generate a
406 self-signed certificate for inbound TLS connections.
408 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
409 in rewrites and routers.
411 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
412 and logged a warning sing 4.83; now they are a configuration file error.
414 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
415 (lacking @domain). Apply the same qualification processing as RCPT.
417 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
419 JH/10 Support ${sha256:} applied to a string (as well as the previous
422 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
423 a cutthrough deliver is pending, as we always want to make a connection.
424 This also avoids re-routing the message when later placing the cutthrough
425 connection after a verify cache hit.
426 Do not update it with the verify result either.
428 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
429 when routing results in more than one destination address.
431 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
432 signing (which inhibits the cutthrough capability). Previously only
433 the presence of an option was tested; now an expansion evaluating as
434 empty is permissible (obviously it should depend only on data available
435 when the cutthrough connection is made).
437 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
438 the relevant preceding SMTP command did not note the pipelining mode.
440 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
441 Previously they were not counted.
443 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
444 as one having no matching records. Previously we deferred the message
445 that needed the lookup.
447 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
448 distinguished as "(=".
450 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
451 for missing MX records. Previously it only worked for missing A records.
453 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
455 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
456 after the data-go-ahead and data-ack. Patch from Jason Betts.
458 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
459 even for a "none" policy. Patch from Tony Meyer.
461 JH/22 Fix continued use of a connection for further deliveries. If a port was
462 specified by a router, it must also match for the delivery to be
465 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
466 When one of the recipients of a message was unsuitable for the connection
467 (has no matching addresses), we lost track of needing to mark it
468 deferred. As a result mail would be lost.
470 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
472 JH/25 Decoding ACL controls is now done using a binary search; the source code
473 takes up less space and should be simpler to maintain. Merge the ACL
474 condition decode tables also, with similar effect.
476 JH/26 Fix problem with one_time used on a redirect router which returned the
477 parent address unchanged. A retry would see the parent address marked as
478 delivered, so not attempt the (identical) child. As a result mail would
481 JH/27 Fix a possible security hole, wherein a process operating with the Exim
482 UID can gain a root shell. Credit to http://www.halfdog.net/ for
483 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
486 JH/28 Enable {spool,log} filesystem space and inode checks as default.
487 Main config options check_{log,spool}_{inodes,space} are now
488 100 inodes, 10MB unless set otherwise in the configuration.
490 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
491 Previously it only applied to the main-section connection policy
494 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
496 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
497 by me. Added RFC7919 DH primes as an alternative.
499 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
500 are not in the system include path.
502 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
503 GnuTLS, when a session startup failed (eg because the client disconnected)
504 Exim did stdio operations after fclose. This was exposed by a recent
505 change which nulled out the file handle after the fclose.
507 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
508 signed directly by the cert-signing cert, rather than an intermediate
509 OCSP-signing cert. This is the model used by LetsEncrypt.
511 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
513 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
514 an incoming connection.
516 HS/02 Bug 1802: Do not half-close the connection after sending a request
519 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
520 fallback to "prime256v1".
522 JH/34 SECURITY: Use proper copy of DATA command in error message.
523 Could leak key material. Remotely exploitable. CVE-2016-9963.
529 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
530 and 3.4.4 - once the server is enabled to respond to an OCSP request
531 it does even when not requested, resulting in a stapling non-aware
532 client dropping the TLS connection.
534 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
535 support variable-length bit vectors. No functional change.
537 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
538 The I= interface field on outgoing lines is now after the H= remote
539 host field, same as incoming lines. There is a separate
540 outgoing_interface log selector which allows you to disable the
543 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
544 If not running log_selector +smtp_connection the mainlog would be held
545 open indefinitely after a "too many connections" event, including to a
546 deleted file after a log rotate. Leave the per net connection logging
547 leaving it open for efficiency as that will be quickly detected by the
548 check on the next write.
550 HS/01 Bug 1671: Fix post transport crash.
551 Processing the wait-<transport> messages could crash the delivery
552 process if the message IDs didn't exist for some reason. When
553 using 'split_spool_directory=yes' the construction of the spool
554 file name failed already, exposing the same netto behaviour.
556 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
557 mime_regex ACL conditions.
559 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
560 to DSN fail messages (bounces): remote IP, remote greeting, remote response
561 to HELO, local diagnostic string.
563 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
564 log line to debug. Even when configured with a tls authenticator many
565 client connections are expected to not authenticate in this way, so
566 an authenticate fail is not an error.
568 HS/02 Add the Exim version string to the process info. This way exiwhat
569 gives some more detail about the running daemon.
571 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
572 matter for fast-change records such as DNSBLs.
574 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
575 retry record, even if constant. There may be multiple transports with
576 different interface settings and the retry behaviour needs to be kept
579 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
581 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
583 JH/10 Bug 840: fix log_defer_output option of pipe transport
585 JH/11 Bug 830: use same host for all RCPTS of a message, even under
586 hosts_randomize. This matters a lot when combined with mua_wrapper.
588 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
589 ${quote_pgsql:<string>} operator.
591 JH/13 Bug 1708: avoid misaligned access in cached lookup.
593 JH/14 Change header file name for freeradius-client. Relevant if compiling
594 with Radius support; from the Gentoo tree and checked under Fedora.
596 JH/15 Bug 1712: Introduce $prdr_requested flag variable
598 JH/16 Bug 1714: Permit an empty string as expansion result for transport
599 option transport_filter, meaning no filtering.
601 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
603 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
604 defaults to "*" (all hosts). The variable is now available when not built
605 with TLS, default unset, mainly to enable keeping the testsuite sane.
606 If a server certificate is not supplied (via tls_certificate) an error is
607 logged, and clients will find TLS connections fail on startup. Presumably
608 they will retry in-clear.
609 Packagers of Exim are strongly encouraged to create a server certificate
610 at installation time.
612 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
613 with the $config_file variable.
615 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
616 in transport context, after the attempt, and per-recipient. The latter type
617 is per host attempted. The event data is the error message, and the errno
618 information encodes the lookup type (A vs. MX) used for the (first) host,
619 and the trailing two digits of the smtp 4xx response.
621 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
622 to write to mainlog (or rejectlog, paniclog) in the window between file
623 creation and permissions/ownership being changed. Particularly affects
624 installations where exicyclog is run as root, rather than exim user;
625 result is that the running daemon panics and dies.
627 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
629 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
630 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
631 "pri" and "weight". Note that the previous implicit priority given by the
632 list order is no longer honoured.
634 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
637 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
638 by defining SUPPORT_SOCKS.
640 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
641 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
642 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
643 variables are renamed to proxy_{local,external}_{address,port}.
645 JH/27 Move Internationalisation support from Experimental to mainline, enabled
646 for a build by defining SUPPORT_I18N
648 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
649 of the query string, and make ${quote_redis:} do that quoting.
651 JH/29 Move Events support from Experimental to mainline, enabled by default
652 and removable for a build by defining DISABLE_EVENT.
654 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
656 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
657 cached by the daemon.
659 JH/32 Move Redis support from Experimental to mainline, enabled for a build
660 by defining LOOKUP_REDIS. The libhiredis library is required.
662 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
663 keys are given for lookup.
665 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
666 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
667 only supported when built with TLS support. The PolarSSL SHA routines
668 are still used when the TLS library is too old for convenient support.
670 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
671 openssl_options), for security. OpenSSL forces this from version 1.1.0
672 server-side so match that on older versions.
674 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
675 allocation for $value could be released as the expansion processing
676 concluded, but leaving the global pointer active for it.
678 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
679 and to use the domains and local_parts ACL conditions.
681 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
682 incorrectly not doubled on cutthrough transmission, hence seen as a
683 body-termination at the receiving system - resulting in truncated mails.
684 Commonly the sender saw a TCP-level error, and retransmitted the message
685 via the normal store-and-forward channel. This could result in duplicates
686 received - but deduplicating mailstores were liable to retain only the
687 initial truncated version.
689 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
691 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
693 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
694 we're in there, support oversigning also; bug 1309.
696 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
698 HS/04 Add support for keep_environment and add_environment options.
700 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
701 either intentional arithmetic overflow during PRNG, or testing config-
704 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
705 delivery resulted in actual delivery. Cancel cutthrough before DATA
708 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
709 rejects a recipient: pass the reject to the originator.
711 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
712 Many were false-positives and ignorable, but it's worth fixing the
715 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
716 for the new environment-manipulation done at startup. Move the routines
717 from being local to tls.c to being global via the os.c file.
719 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
720 an extract embedded as result-arg for a map, the first arg for extract
721 is unavailable so we cannot tell if this is a numbered or keyed
722 extraction. Accept either.
728 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
731 JH/02 The smtp transport option "multi_domain" is now expanded.
733 JH/03 The smtp transport now requests PRDR by default, if the server offers
736 JH/04 Certificate name checking on server certificates, when exim is a client,
737 is now done by default. The transport option tls_verify_cert_hostnames
738 can be used to disable this per-host. The build option
739 EXPERIMENTAL_CERTNAMES is withdrawn.
741 JH/05 The value of the tls_verify_certificates smtp transport and main options
742 default to the word "system" to access the system default CA bundle.
743 For GnuTLS, only version 3.0.20 or later.
745 JH/06 Verification of the server certificate for a TLS connection is now tried
746 (but not required) by default. The verification status is now logged by
747 default, for both outbound TLS and client-certificate supplying inbound
750 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
753 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
754 Status Notification (bounce) messages are now MIME format per RFC 3464.
755 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
756 under the control of the dsn_advertise_hosts option, and routers may
757 have a dsn_lasthop option.
759 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
760 default, modifiable by a malware= option. The list separator for
761 the options can now be changed in the usual way. Bug 68.
763 JH/10 The smtp_receive_timeout main option is now expanded before use.
765 JH/11 The incoming_interface log option now also enables logging of the
766 local interface on delivery outgoing connections.
768 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
769 if the interface and destination host and port all match.
771 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
774 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
775 Patch from Andrew Lewis.
777 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
778 now supports optional time-restrictions, weighting, and priority
779 modifiers per server. Patch originally by <rommer@active.by>.
781 JH/16 The spamd_address main option now supports a mixed list of local
782 and remote servers. Remote servers can be IPv6 addresses, and
783 specify a port-range.
785 JH/17 Bug 68: The spamd_address main option now supports an optional
786 timeout value per server.
788 JH/18 Bug 1581: Router and transport options headers_add/remove can
789 now have the list separator specified.
791 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
794 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
797 JH/21 Support for the A6 type of dns record is withdrawn.
799 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
800 rather than the verbs used.
802 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
803 from 255 to 1024 chars.
805 JH/24 Verification callouts now attempt to use TLS by default.
807 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
808 are generic router options now. The defaults didn't change.
810 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
811 Original patch from Alexander Shikoff, worked over by JH.
813 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
814 files if ZCAT_COMMAND is not executable.
816 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
818 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
820 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
821 Normally benign, it bites when the pair was led to by a CNAME;
822 modern usage is to not canonicalize the domain to a CNAME target
823 (and we were inconsistent anyway for A-only vs AAAA+A).
825 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
827 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
828 when evaluating $sender_host_dnssec.
830 JH/31 Check the HELO verification lookup for DNSSEC, adding new
831 $sender_helo_dnssec variable.
833 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
835 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
837 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
839 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
840 documented as working, but never had. Support all but $spam_report.
842 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
843 added for tls authenticator.
845 HS/03 Add perl_taintmode main config option
851 TL/01 When running the test suite, the README says that variables such as
852 no_msglog_check are global and can be placed anywhere in a specific
853 test's script, however it was observed that placement needed to be near
854 the beginning for it to behave that way. Changed the runtest perl
855 script to read through the entire script once to detect and set these
856 variables, reset to the beginning of the script, and then run through
857 the script parsing/test process like normal.
859 TL/02 The BSD's have an arc4random API. One of the functions to induce
860 adding randomness was arc4random_stir(), but it has been removed in
861 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
862 function when detected.
864 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
865 cause callback expansion.
867 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
868 syntax errors in an expansion can be treated as a string instead of
869 logging or causing an error, due to the internal use of bool_lax
870 instead of bool when processing it.
872 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
873 server certificates when making smtp deliveries.
875 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
877 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
879 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
881 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
882 Merged patch from Sebastian Wiedenroth.
884 JH/05 Fix results-pipe from transport process. Several recipients, combined
885 with certificate use, exposed issues where response data items split
886 over buffer boundaries were not parsed properly. This eventually
887 resulted in duplicates being sent. This issue only became common enough
888 to notice due to the introduction of connection certificate information,
889 the item size being so much larger. Found and fixed by Wolfgang Breyha.
891 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
892 size buffer was used, resulting in syntax errors when an expansion
895 JH/07 Add support for directories of certificates when compiled with a GnuTLS
896 version 3.3.6 or later.
898 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
899 is EXPERIMENTAL_EVENT, the main-configuration and transport options
900 both become "event_action", the variables become $event_name, $event_data
901 and $event_defer_errno. There is a new variable $verify_mode, usable in
902 routers, transports and related events. The tls:cert event is now also
903 raised for inbound connections, if the main configuration event_action
906 TL/06 In test suite, disable OCSP for old versions of openssl which contained
907 early OCSP support, but no stapling (appears to be less than 1.0.0).
909 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
910 server certificate names available under the smtp transport option
911 "tls_verify_cert_hostname" now do not permit multi-component wildcard
914 JH/10 Time-related extraction expansions from certificates now use the main
915 option "timezone" setting for output formatting, and are consistent
916 between OpenSSL and GnuTLS compilations. Bug 1541.
918 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
919 encoded parameter in the incoming message. Bug 1558.
921 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
922 include certificate info, eximon was claiming there were spoolfile
925 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
927 JH/14 Log delivery-related information more consistently, using the sequence
928 "H=<name> [<ip>]" wherever possible.
930 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
931 are problematic for Debian distribution, omit them from the release
934 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
936 JH/16 Fix string representation of time values on 64bit time_t architectures.
939 JH/17 Fix a null-indirection in certextract expansions when a nondefault
940 output list separator was used.
945 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
946 checkers that were complaining about end of non-void function with no
949 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
950 This was a regression introduced in 4.83 by another bugfix.
952 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
954 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
955 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
961 TF/01 Correctly close the server side of TLS when forking for delivery.
963 When a message was received over SMTP with TLS, Exim failed to clear up
964 the incoming connection properly after forking off the child process to
965 deliver the message. In some situations the subsequent outgoing
966 delivery connection happened to have the same fd number as the incoming
967 connection previously had. Exim would try to use TLS and fail, logging
968 a "Bad file descriptor" error.
970 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
971 utilities have not been installed.
973 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
974 temporary space as the ACL may create new global variables.
976 TL/01 LDAP support uses per connection or global context settings, depending
977 upon the detected version of the libraries at build time.
979 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
980 to extract and use the src ip:port in logging and expansions as if it
981 were a direct connection from the outside internet. PPv2 support was
982 updated based on HAProxy spec change in May 2014.
984 JH/02 Add ${listextract {number}{list}{success}{fail}}.
986 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
987 Properly escape header and check for NULL return.
989 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
992 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
994 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
995 characters in header names, implemented as a verify condition.
996 Contributed by Michael Fischer v. Mollard.
998 TL/05 Rename SPF condition results err_perm and err_temp to standardized
999 results permerror and temperror. Previous values are deprecated but
1000 still accepted. In a future release, err_perm and err_temp will be
1001 completely removed, which will be a backward incompatibility if the
1002 ACL tests for either of these two old results. Patch contributed by
1003 user bes-internal on the mailing list.
1005 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1007 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1008 selectors, in both main and reject logs.
1010 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1013 JH/07 Add malware type "sock" for talking to simple daemon.
1015 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1017 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1018 routers/transports under cutthrough routing.
1020 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1021 numbers. Touch up "bool" conditional to keep the same definition.
1023 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1025 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1027 JH/12 Expand items in router/transport headers_add or headers_remove lists
1028 individually rather than the list as a whole. Bug 1452.
1030 Required for reasonable handling of multiple headers_ options when
1031 they may be empty; requires that headers_remove items with embedded
1032 colons must have them doubled (or the list-separator changed).
1034 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1035 view the policy declared in the DMARC record. Currently, $dmarc_status
1036 is a combined value of both the record presence and the result of the
1039 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1041 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1042 dnslookup router and the smtp transport (applying to the forward
1045 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1046 of ldap servers used for a specific lookup. Patch provided by Heiko
1049 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1050 New variable $lookup_dnssec_authenticated for observability.
1052 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1053 Patch submitted by Lars Timman.
1055 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1057 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1058 Requires trusted mode and valid format message id, aborts otherwise.
1059 Patch contributed by Heiko Schlichting.
1061 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1062 certextract with support for various fields. Bug 1358.
1064 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1065 is requested by default, modifiable by smtp transport option
1068 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1069 operate on certificate variables to give certificate fingerprints
1070 Also new ${sha256:cert_variable}.
1072 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1074 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1075 Christian Aistleitner.
1077 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1079 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1080 file. Patch from Wolfgang Breyha.
1082 JH/25 Expand the coverage of the delivery $host and $host_address to
1083 client authenticators run in verify callout. Bug 1476.
1085 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1086 align with daemon_smtp_ports. Bug 72.
1088 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1089 support and error reporting did not work properly.
1091 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1092 and is readable. Patch from Andrew Colin Kissa.
1094 TL/14 Enhance documentation of ${run expansion and how it parses the
1095 commandline after expansion, particularly in the case when an
1096 unquoted variable expansion results in an empty value.
1098 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1100 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1101 by renaming away from it. Addresses GH issue 15, reported by
1104 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1106 TL/15 SECURITY: prevent double expansion in math comparison functions
1107 (can expand unsanitized data). Not remotely exploitable.
1114 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1116 PP/02 Make -n do something, by making it not do something.
1117 When combined with -bP, the name of an option is not output.
1119 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1122 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1123 $sender_host_name and config options to manage this, and basic check
1126 PP/05 DSCP support for outbound connections and control modifier for inbound.
1128 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1129 (Only plugin which currently uses this is kerberos4, which nobody should
1130 be using, but we should make it available and other future plugins might
1131 conceivably use it, even though it would break NAT; stuff *should* be
1132 using channel bindings instead).
1134 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1135 name; added for Sendmail compatibility; requires admin caller.
1136 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1137 just ignore it); requires trusted caller.
1138 Also parse but ignore: -Ac -Am -X<logfile>
1141 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1143 TL/02 Add +smtp_confirmation as a default logging option.
1145 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1146 Patch by Magnus Holmgren from 2007-02-20.
1148 TL/04 Bugzilla 1281 - Spec typo.
1149 Bugzilla 1283 - Spec typo.
1150 Bugzilla 1290 - Spec grammar fixes.
1152 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1154 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1156 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1157 mailing list by by Dmitry Isaikin.
1159 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1161 JH/02 Support "G" suffix to numbers in ${if comparisons.
1163 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1165 NM/01 Bugzilla 1197 - Spec typo
1166 Bugzilla 1196 - Spec examples corrections
1168 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1170 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1171 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1174 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1175 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1177 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1178 "acl {{name}{arg}...}", and optional args on acl condition
1181 JH/05 Permit multiple router/transport headers_add/remove lines.
1183 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1185 JH/07 Avoid using a waiting database for a single-message-only transport.
1186 Performance patch from Paul Fisher. Bugzilla 1262.
1188 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1191 JH/09 Add $headers_added variable, with content from use of ACL modifier
1192 add_header (but not yet added to the message). Bugzilla 199.
1194 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1195 Pulled from Bugzilla 817 by Wolfgang Breyha.
1197 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1199 (nb: this is the same fix as in Exim 4.80.1)
1201 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1204 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1207 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1209 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1210 advertise SMTP AUTH mechanism to us, instead of a generic
1211 protocol violation error. Also, make Exim more robust to bad
1212 data from the Dovecot auth socket.
1214 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1216 When a queue runner is handling a message, Exim first routes the
1217 recipient addresses, during which it prunes them based on the retry
1218 hints database. After that it attempts to deliver the message to
1219 any remaining recipients. It then updates the hints database using
1222 So if a recipient address works intermittently, it can get repeatedly
1223 deferred at routing time. The retry hints record remains fresh so the
1224 address never reaches the final cutoff time.
1226 This is a fairly common occurrence when a user is bumping up against
1227 their storage quota. Exim had some logic in its local delivery code
1228 to deal with this. However it did not apply to per-recipient defers
1229 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1231 This change adds a proper retry rule check during routing so that the
1232 final cutoff time is checked against the message's age. We only do
1233 this check if there is an address retry record and there is not a
1234 domain retry record; this implies that previous attempts to handle
1235 the address had the retry_use_local_parts option turned on. We use
1236 this as an approximation for the destination being like a local
1237 delivery, as in LMTP.
1239 I suspect this new check makes the old local delivery cutoff check
1240 redundant, but I have not verified this so I left the code in place.
1242 TF/02 Correct gecos expansion when From: is a prefix of the username.
1244 Test 0254 submits a message to Exim with the header
1248 When I ran the test suite under the user fanf2, Exim expanded
1249 the header to contain my full name, whereas it should have added
1250 a Resent-Sender: header. It erroneously treats any prefix of the
1251 username as equal to the username.
1253 This change corrects that bug.
1255 GF/01 DCC debug and logging tidyup
1256 Error conditions log to paniclog rather than rejectlog.
1257 Debug lines prefixed by "DCC: " to remove any ambiguity.
1259 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1261 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1262 Bug spotted by Jeremy Harris; was flawed since initial commit.
1263 Would have resulted in OCSP responses post-SNI triggering an Exim
1264 NULL dereference and crash.
1266 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1268 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1269 Bug detection, analysis and fix by Samuel Thibault.
1270 Bugzilla 1331, Debian bug #698092.
1272 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1274 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1275 Server implementation by Todd Lyons, client by JH.
1276 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1277 config variable "prdr_enable" controls whether the server
1278 advertises the facility. If the client requests PRDR a new
1279 acl_data_smtp_prdr ACL is called once for each recipient, after
1280 the body content is received and before the acl_smtp_data ACL.
1281 The client is controlled by both of: a hosts_try_prdr option
1282 on the smtp transport, and the server advertisement.
1283 Default client logging of deliveries and rejections involving
1284 PRDR are flagged with the string "PRDR".
1286 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1287 fclose(). Diagnosis by Todd Lyons.
1289 PP/17 Update configure.default to handle IPv6 localhost better.
1290 Patch by Alain Williams (plus minor tweaks).
1293 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1294 This is now consistent with GnuTLS, and is now documented: the
1295 previous undocumented portable approach to treating the option as
1296 unset was to force an expansion failure. That still works, and
1297 an empty string is now equivalent.
1299 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1300 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1301 not performing validation itself.
1303 PP/20 Added force_command boolean option to pipe transport.
1304 Patch from Nick Koston, of cPanel Inc.
1306 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1309 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1311 PP/21 Fix eximon continuous updating with timestamped log-files.
1312 Broken in a format-string cleanup in 4.80, missed when I repaired the
1313 other false fix of the same issue.
1314 Report and fix from Heiko Schlichting.
1317 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1318 Report from Prashanth Katuri.
1320 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1321 It's SecureTransport, so affects any MacOS clients which use the
1322 system-integrated TLS libraries, including email clients.
1324 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1325 using a MIME ACL for non-SMTP local injection.
1326 Report and assistance in diagnosis by Warren Baker.
1328 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1330 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1332 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1333 last id that failed so it may be referenced in subsequent ACL's.
1335 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1338 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1339 ldap library initialization, allowing self-signed CA's to be
1340 used. Also properly sets require_cert option later in code by
1341 using NULL (global ldap config) instead of ldap handle (per
1342 session). Bug diagnosis and testing by alxgomz.
1344 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1345 the src/util/ subdirectory.
1347 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1348 renamed to Transport Post Delivery Action by Jeremy Harris, as
1351 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1352 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1353 redis_servers = needs to be configured which will be used by the redis
1354 lookup. Patch from Warren Baker, of The Packet Hub.
1356 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1358 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1359 hostname or reverse DNS when processing a host list. Used suggestions
1360 from multiple comments on this bug.
1362 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1364 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1365 Missed a few lines, added it to make the runtest require no keyboard
1368 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1369 contains upper case chars. Make router use caseful_local_part.
1371 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1372 support when GnuTLS has been built with p11-kit.
1378 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1380 This, or similar/improved, will also be change PP/11 of 4.82.
1386 PP/01 Handle short writes when writing local log-files.
1387 In practice, only affects FreeBSD (8 onwards).
1388 Bugzilla 1053, with thanks to Dmitry Isaikin.
1390 NM/01 Bugzilla 949 - Documentation tweak
1392 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1395 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1397 PP/02 Implemented gsasl authenticator.
1399 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1401 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1402 `pkg-config foo` for cflags/libs.
1404 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1405 with rest of GSASL and with heimdal_gssapi.
1407 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1408 `pkg-config foo` for cflags/libs for the TLS implementation.
1410 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1411 properties get this fed in as external SSF. A number of robustness
1412 and debugging improvements to the cyrus_sasl authenticator.
1414 PP/08 cyrus_sasl server now expands the server_realm option.
1416 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1417 Patch by Jeremy Harris.
1419 PP/10 Added dbmjz lookup type.
1421 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1423 PP/12 MAIL args handles TAB as well as SP, for better interop with
1424 non-compliant senders.
1425 Analysis and variant patch by Todd Lyons.
1427 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1428 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1429 Patch from Dirk Mueller <dmueller@suse.com>
1431 PP/13 tls_peerdn now print-escaped for spool files.
1432 Observed some $tls_peerdn in wild which contained \n, which resulted
1433 in spool file corruption.
1435 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1436 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1437 or write after TLS renegotiation, which otherwise led to messages
1440 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1441 as a tracking header (ie: a signed header comes before the signature).
1442 Patch from Wolfgang Breyha.
1444 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1445 comma-sep list; embedded commas doubled.
1447 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1449 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1451 Report and patch from Dmitry Banschikov.
1453 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1454 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1455 failures. We appear to now support TLS1.1+ with Exim.
1457 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1458 lets Exim select keys and certificates based upon TLS SNI from client.
1459 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1460 before an outbound SMTP session. New log_selector, +tls_sni.
1462 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1463 NULL dereference. Report and patch from Alun Jones.
1465 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1466 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1467 Not seeing resolver debug output on NetBSD, but suspect this is a
1468 resolver implementation change.
1470 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1471 Left warnings. Added "eximon gdb" invocation mode.
1473 PP/21 Defaulting "accept_8bitmime" to true, not false.
1475 PP/22 Added -bw for inetd wait mode support.
1477 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1478 locate the relevant includes and libraries. Made this the default.
1480 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1481 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1483 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1484 This may cause build issues on older platforms.
1486 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1487 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1488 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1489 Added SNI support via GnuTLS too.
1490 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1492 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1494 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1495 Applied second patch from Janne, implementing suggestion to default
1496 multiple-strings-in-record handling to match SPF spec.
1498 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1500 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1501 read-only, out of scope).
1502 Patch from Wolfgang Breyha, report from Stuart Northfield.
1504 PP/29 Fix three issues highlighted by clang analyser static analysis.
1505 Only crash-plausible issue would require the Cambridge-specific
1506 iplookup router and a misconfiguration.
1507 Report from Marcin Mirosław.
1509 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1511 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1512 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1513 As part of this, removing so much warning spew let me fix some minor
1514 real issues in debug logging.
1516 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1517 assignment on my part. Fixed.
1519 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1520 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1521 Janne Snabb (who went above and beyond: thank you).
1523 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1524 string otherwise requires a connection and a bunch more work and it's
1525 relatively easy to get wrong. Should also expose TLS library linkage
1528 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1529 64-bit ${eval} (JH/03).
1531 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1532 GNU libc to support some of the 64-bit stuff, should not lead to
1533 conflicts. Defined before os.h is pulled in, so if a given platform
1534 needs to override this, it can.
1536 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1537 protection layer was required, which is not implemented.
1538 Bugzilla 1254, patch from Wolfgang Breyha.
1540 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1541 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1542 tls_dhparam take prime identifiers. Also unbreak combination of
1543 OpenSSL+DH_params+TLSSNI.
1545 PP/39 Disable SSLv2 by default in OpenSSL support.
1551 PP/01 Solaris build fix for Oracle's LDAP libraries.
1552 Bugzilla 1109, patch from Stephen Usher.
1554 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1556 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1559 TF/02 Fix a couple more cases where we did not log the error message
1560 when unlink() failed. See also change 4.74-TF/03.
1562 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1563 lock up or crash if it happened to be inside a call to libc when it
1564 got a SIGUSR1 from exiwhat.
1566 The SIGUSR1 handler appends the current process status to the process
1567 log which is later printed by exiwhat. It used to use the general
1568 purpose logging code to do this, but several functions it calls are
1569 not safe for signals.
1571 The new output code in the SIGUSR1 handler is specific to the process
1572 log, and simple enough that it's easy to inspect for signal safety.
1573 Removing some special cases also simplifies the general logging code.
1574 Removing the spurious timestamps from the process log simplifies
1577 TF/04 Improved ratelimit ACL condition.
1579 The /noupdate option has been deprecated in favour of /readonly which
1580 has clearer semantics. The /leaky, /strict, and /readonly update modes
1581 are mutually exclusive. The update mode is no longer included in the
1582 database key; it just determines when the database is updated. (This
1583 means that when you upgrade Exim will forget old rate measurements.)
1585 Exim now checks that the per_* options are used with an update mode that
1586 makes sense for the current ACL. For example, when Exim is processing a
1587 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1588 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1589 must specify per_mail/readonly. If you omit the update mode it defaults to
1590 /leaky where that makes sense (as before) or /readonly where required.
1592 The /noupdate option is now undocumented but still supported for
1593 backwards compatibility. It is equivalent to /readonly except that in
1594 ACLs where /readonly is required you may specify /leaky/noupdate or
1595 /strict/noupdate which are treated the same as /readonly.
1597 A useful new feature is the /count= option. This is a generalization
1598 of the per_byte option, so that you can measure the throughput of other
1599 aggregate values. For example, the per_byte option is now equivalent
1600 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1602 The per_rcpt option has been generalized using the /count= mechanism
1603 (though it's more complicated than the per_byte equivalence). When it is
1604 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1605 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1606 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1607 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1608 Note that using per_rcpt with a non-readonly update mode in more than
1609 one ACL will cause the recipients to be double-counted. (The per_mail
1610 and per_byte options don't have this problem.)
1612 The handling of very low rates has changed slightly. If the computed rate
1613 is less than the event's count (usually one) then this event is the first
1614 after a long gap. In this case the rate is set to the same as this event's
1615 count, so that the first message of a spam run is counted properly.
1617 The major new feature is a mechanism for counting the rate of unique
1618 events. The new per_addr option counts the number of different
1619 recipients that someone has sent messages to in the last time period. It
1620 behaves like per_rcpt if all the recipient addresses are different, but
1621 duplicate recipient addresses do not increase the measured rate. Like
1622 the /count= option this is a general mechanism, so the per_addr option
1623 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1624 example, measure the rate that a client uses different sender addresses
1625 with the options per_mail/unique=$sender_address. There are further
1626 details in the main documentation.
1628 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1630 TF/06 Removed a few PCRE remnants.
1632 TF/07 Automatically extract Exim's version number from tags in the git
1633 repository when doing development or release builds.
1635 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1636 Bugzilla 879. Patch from Paul Fisher.
1638 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1639 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1642 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1644 PP/05 Variable $av_failed, true if the AV scanner deferred.
1645 Bugzilla 1078. Patch from John Horne.
1647 PP/06 Stop make process more reliably on build failure.
1648 Bugzilla 1087. Patch from Heiko Schlittermann.
1650 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1651 Bugzilla 1089. Patch from Heiko Schlittermann.
1653 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1654 Bugzilla 1131. Patch from Holger Weiß.
1656 PP/09 Handle IPv6 addresses with SPF.
1657 Bugzilla 860. Patch from Wolfgang Breyha.
1659 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1661 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1664 PP/11 match_* no longer expand right-hand-side by default.
1665 New compile-time build option, EXPAND_LISTMATCH_RHS.
1666 New expansion conditions, "inlist", "inlisti".
1668 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1670 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1672 PP/14 fix log_write() format string regression from TF/03.
1673 Bugzilla 1152. Patch from Dmitry Isaikin.
1679 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1681 PP/02 Harmonised TLS library version reporting; only show if debugging.
1682 Layout now matches that introduced for other libraries in 4.74 PP/03.
1684 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1686 PP/04 New "dns_use_edns0" global option.
1688 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1691 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1692 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1694 TK/01 Updated PolarSSL code to 0.14.2.
1695 Bugzilla 1097. Patch from Andreas Metzler.
1697 PP/07 Catch divide-by-zero in ${eval:...}.
1698 Fixes bugzilla 1102.
1700 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1703 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1704 format-string attack -- SECURITY: remote arbitrary code execution.
1706 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1707 time unintentionally subject to list matching rules, letting the header
1708 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1709 arbitrary string expansion). This allowed for information disclosure.
1711 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1712 INT_MIN/-1 -- value coerced to INT_MAX.
1718 NM/01 Workaround for PCRE version dependency in version reporting
1721 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1722 This fixes portability to compilers other than gcc, notably
1723 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1725 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1726 makefiles for portability to HP-UX and POSIX correctness.
1728 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1729 Also via indented variable definition in the Makefile.
1730 (Debugging by Oliver Heesakkers).
1732 PP/02 Restore caching of spamd results with expanded spamd_address.
1733 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1735 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1736 Improves build reliability. Fix from: Frank Elsner
1738 NM/02 Fix wide character breakage in the rfc2047 coding
1739 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1741 NM/03 Allow underscore in dnslist lookups
1742 Fixes bug 1026. Patch from Graeme Fowler
1744 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1745 Code patches from Adam Ciarcinski of NetBSD.
1747 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1750 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1751 is logged, to avoid truncation. Patch from John Horne.
1753 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1754 Patch from Jakob Hirsch.
1756 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1757 SQL string expansion failure details.
1758 Patch from Andrey Oktyabrski.
1760 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1761 Patch from Simon Arlott.
1763 PP/09 New lookups functionality failed to compile on old gcc which rejects
1764 extern declarations in function scope.
1765 Patch from Oliver Fleischmann
1767 PP/10 Use sig_atomic_t for flags set from signal handlers.
1768 Check getgroups() return and improve debugging.
1769 Fixed developed for diagnosis in bug 927 (which turned out to be
1772 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1773 Patch from Mark Zealey.
1775 PP/12 Bugzilla 1056: Improved spamd server selection.
1776 Patch from Mark Zealey.
1778 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1779 Based on patch from Heiko Schlittermann.
1781 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1782 Patch from Uwe Doering, sign-off by Michael Haardt.
1784 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1785 variable declaration deep within a block. Bug and patch from
1788 PP/15 lookups-Makefile IRIX compatibility coercion.
1790 PP/16 Make DISABLE_DKIM build knob functional.
1792 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1793 Patch by Simon Arlott
1795 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1796 variable argument macros. Our copy now differs from upstream.
1802 TF/01 Failure to get a lock on a hints database can have serious
1803 consequences so log it to the panic log.
1805 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1806 controlled using the smtp_confirmation log selector.
1808 TF/03 Include the error message when we fail to unlink a spool file.
1810 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1811 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1812 for maintaining out-of-tree patches for some time.
1814 PP/01 Bugzilla 139: Documentation and portability issues.
1815 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1816 Handle per-OS dynamic-module compilation flags.
1818 PP/02 Let /dev/null have normal permissions.
1819 The 4.73 fixes were a little too stringent and complained about the
1820 permissions on /dev/null. Exempt it from some checks.
1821 Reported by Andreas M. Kirchwitz.
1823 PP/03 Report version information for many libraries, including
1824 Exim version information for dynamically loaded libraries. Created
1825 version.h, now support a version extension string for distributors
1826 who patch heavily. Dynamic module ABI change.
1828 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1829 privilege escalation vulnerability whereby the Exim run-time user
1830 can cause root to append content of the attacker's choosing to
1833 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1836 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1837 If dropping privileges for untrusted macros, we disabled normal logging
1838 on the basis that it would fail; for the Exim run-time user, this is not
1839 the case, and it resulted in successful deliveries going unlogged.
1840 Fixed. Reported by Andreas Metzler.
1846 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1847 only prepend for the Resent-* case. Fixes regression introduced in
1848 Exim 4.70 by NM/22 for Bugzilla 607.
1850 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1851 increasing numbers of administrators be bitten by this.
1853 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1855 PP/03 Bugzilla 994: added openssl_options main configuration option.
1857 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1859 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1861 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1863 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1864 without a peer certificate, leading to a segfault because of an
1865 assumption that peers always have certificates. Be a little more
1866 paranoid. Problem reported by Martin Tscholak.
1868 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1869 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1870 NB: ClamAV planning to remove STREAM in "middle of 2010".
1871 CL also introduces -bmalware, various -d+acl logging additions and
1872 more caution in buffer sizes.
1874 PP/09 Implemented reverse_ip expansion operator.
1876 PP/10 Bugzilla 937: provide a "debug" ACL control.
1878 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1880 PP/12 Bugzilla 973: Implement --version.
1882 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1884 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1886 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1888 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1889 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1890 ignore trailing whitespace.
1892 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1894 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1897 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1898 Notification from Dr Andrew Aitchison.
1900 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1901 ExtendedDetectionInfo response format.
1902 Notification from John Horne.
1904 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1907 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1908 XSL and documented dependency on system catalogs, with examples of how
1911 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1914 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1915 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1916 configuration file which is writeable by the Exim user or group.
1918 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1919 of configuration files to cover files specified with the -C option if
1920 they are going to be used with root privileges, not just the default
1923 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1924 option (effectively making it always true).
1926 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1927 files to be used while preserving root privileges.
1929 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1930 that rogue child processes cannot use them.
1932 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1933 run-time user, instead of root.
1935 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1936 Exim run-time user without dropping privileges.
1938 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1939 result string, instead of calling string_vformat() twice with the same
1942 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1943 for other users. Others should always drop root privileges if they use
1944 -C on the command line, even for a whitelisted configure file.
1946 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1948 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1954 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1955 $data_path, and $header_path variables; fixed documentation bugs and
1958 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1959 exipick to access non-standard spools, including the "frozen" queue
1962 NM/01 Bugzilla 965: Support mysql stored procedures.
1963 Patch from Alain Williams
1965 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1967 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1968 Patch from Andreas Metzler
1970 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1971 Patch from Kirill Miazine
1973 NM/05 Bugzilla 671: Added umask to procmail example.
1975 JJ/03 installed exipick 20100323.0, fixing doc bug
1977 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1978 directory. Notification and patch from Dan Rosenberg.
1980 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1982 TK/02 Improve log output when DKIM signing operation fails.
1984 MH/01 Treat the transport option dkim_domain as a colon separated
1985 list, not as a single string, and sign the message with each element,
1986 omitting multiple occurences of the same signer.
1988 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1989 Bugzilla 985, 986. Patch by Simon Arlott
1991 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1992 Patch by Simon Arlott
1994 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1995 MBX locking. Notification from Dan Rosenberg.
2001 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2003 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2005 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2007 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2009 NM/04 Bugzilla 915: Flag broken perl installation during build.
2015 TK/01 Added patch by Johannes Berg that expands the main option
2016 "spamd_address" if it starts with a dollar sign.
2018 TK/02 Write list of recipients to X-Envelope-Sender header when building
2019 the mbox-format spool file for content scanning (suggested by Jakob
2022 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2023 (http://www.dcc-servers.net/) support via dccifd. Activated by
2024 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2026 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2027 by Mark Daniel Reidel <mr@df.eu>.
2029 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2030 When building exim an external PCRE library is now needed -
2031 PCRE is a system library on the majority of modern systems.
2032 See entry on PCRE_LIBS in EDITME file.
2034 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2035 conversation. Added nologin parameter to request.
2036 Patch contributed by Kirill Miazine.
2038 TF/01 Do not log submission mode rewrites if they do not change the address.
2040 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2042 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2043 log files in place. Contributed by Roberto Lima.
2045 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2047 TF/03 Bugzilla 615: When checking the local_parts router precondition
2048 after a local_part_suffix or local_part_prefix option, Exim now
2049 does not use the address's named list lookup cache, since this
2050 contains cached lookups for the whole local part.
2052 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2053 Robert Millan. Documentation is in experimental-spec.txt.
2055 TF/04 Bugzilla 668: Fix parallel build (make -j).
2057 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2059 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2060 Patch provided by Jan Srzednicki.
2062 TF/05 Leading white space used to be stripped from $spam_report which
2063 wrecked the formatting. Now it is preserved.
2065 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2066 that they are available at delivery time.
2068 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2070 TF/08 TLS error reporting now respects the incoming_interface and
2071 incoming_port log selectors.
2073 TF/09 Produce a more useful error message if an SMTP transport's hosts
2074 setting expands to an empty string.
2076 NM/06 Bugzilla 744: EXPN did not work under TLS.
2077 Patch provided by Phil Pennock.
2079 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2080 Patch provided by Richard Godbee.
2082 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2083 acl_smtp_notquit, added index entry.
2085 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2086 Patch provided by Eugene Bujak.
2088 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2089 accept(). Patch provided by Maxim Dounin.
2091 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2092 Patch provided by Phil Pennock.
2094 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2096 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2097 Patch provided by Brad "anomie" Jorsch.
2099 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2100 Patch provided by Dean Brooks.
2102 TK/05 Add native DKIM support (does not depend on external libraries).
2104 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2105 Patch provided by Graeme Fowler.
2107 NM/16 Bugzilla 851: Documentation example syntax fix.
2109 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2111 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2114 NM/19 Bugzilla 745: TLS version reporting.
2115 Patch provided by Phil Pennock.
2117 NM/20 Bugzilla 167: bool: condition support.
2118 Patch provided by Phil Pennock.
2120 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2121 clients. Patch provided by Phil Pennock.
2123 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2124 Patch provided by Brad "anomie" Jorsch.
2126 NM/23 Bugzilla 687: Fix misparses in eximstats.
2127 Patch provided by Heiko Schlittermann.
2129 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2130 Patch provided by Heiko Schlittermann.
2132 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2133 plus update to original patch.
2135 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2137 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2138 Patch provided by David Brownlee.
2140 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2142 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2144 NM/30 Bugzilla 888: TLS documentation bugfixes.
2146 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2148 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2149 Unlike the original bugzilla I have changed all shell scripts in src tree.
2151 NM/33 Bugzilla 898: Transport filter timeout fix.
2152 Patch by Todd Rinaldo.
2154 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2155 Patch by Serge Demonchaux.
2157 NM/35 Bugzilla 39: Base64 decode bug fixes.
2158 Patch by Jakob Hirsch.
2160 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2162 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2164 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2166 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2172 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2173 ALT-N's libdkim that I have put here:
2174 http://duncanthrax.net/exim-experimental/
2176 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2177 were called 'true' and it seems that C99 defines that as a reserved
2178 keyword to be used with 'bool' variable types. That means you could
2179 not include C99-style headers which use bools without triggering
2180 build errors in sieve.c.
2182 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2183 as mailq or other aliases. Changed the --help handling significantly
2184 to do whats expected. exim_usage() emits usage/help information.
2186 SC/01 Added the -bylocaldomain option to eximstats.
2188 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2190 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2192 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2198 PH/01 Another patch from the Sieve maintainer.
2200 PH/02 When an IPv6 address is converted to a string for single-key lookup
2201 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2202 dots are used instead of colons so that keys in lsearch files need not
2203 contain colons. This was done some time before quoting was made available
2204 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2205 (notated using the quote facility) so as to distinguish them from IPv4
2206 keys. This meant that lookups for IP addresses in host lists did not work
2207 for iplsearch lookups.
2209 This has been fixed by arranging for IPv6 addresses to be expressed with
2210 colons if the lookup type is iplsearch. This is not incompatible, because
2211 previously such lookups could never work.
2213 The situation is now rather anomalous, since one *can* have colons in
2214 ordinary lsearch keys. However, making the change in all cases is
2215 incompatible and would probably break a number of configurations.
2217 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2220 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2221 conversion specification without a maximum field width, thereby enabling
2222 a rogue spamd server to cause a buffer overflow. While nobody in their
2223 right mind would setup Exim to query an untrusted spamd server, an
2224 attacker that gains access to a server running spamd could potentially
2225 exploit this vulnerability to run arbitrary code as the Exim user.
2227 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2228 $primary_hostname instead of what libspf2 thinks the hosts name is.
2230 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2231 a directory entry by the name of the lookup key. Previously, if a
2232 symlink pointed to a non-existing file or a file in a directory that
2233 Exim lacked permissions to read, a lookup for a key matching that
2234 symlink would fail. Now it is enough that a matching directory entry
2235 exists, symlink or not. (Bugzilla 503.)
2237 PH/03 The body_linecount and body_zerocount variables are now exported in the
2240 PH/04 Added the $dnslist_matched variable.
2242 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2243 This means they are set thereafter only if the connection becomes
2246 PH/06 Added the client_condition to authenticators so that some can be skipped
2247 by clients under certain conditions.
2249 PH/07 The error message for a badly-placed control=no_multiline_responses left
2250 "_responses" off the end of the name.
2252 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2254 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2255 (without spaces) instead of just copying the configuration text.
2257 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2259 PH/11 Added $max_received_linelength.
2261 PH/12 Added +ignore_defer and +include_defer to host lists.
2263 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2264 way in which PCRE > 7.0 is built.
2266 PH/14 Implemented queue_only_load_latch.
2268 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2269 MAIL command. The effect was to mangle the value on 64-bit systems.
2271 PH/16 Another patch from the Sieve maintainer.
2273 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2275 PH/18 If a system quota error occurred while trying to create the file for
2276 a maildir delivery, the message "Mailbox is full" was not appended to the
2277 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2278 only to a quota excession during the actual writing of the file.
2280 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2281 characters?) which causes problems in log lines. The DN values are now
2282 passed through string_printing() before being added to log lines.
2284 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2285 and InterBase are left for another time.)
2287 PH/21 Added message_body_newlines option.
2289 PH/22 Guard against possible overflow in moan_check_errorcopy().
2291 PH/23 POSIX allows open() to be a macro; guard against that.
2293 PH/24 If the recipient of an error message contained an @ in the local part
2294 (suitably quoted, of course), incorrect values were put in $domain and
2295 $local_part during the evaluation of errors_copy.
2301 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2302 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2305 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2306 issue a MAIL command.
2308 PH/02 In an ACL statement such as
2310 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2312 if a client was not listed at all, or was listed with a value other than
2313 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2314 the condition was not true (as it should be), so access was not denied.
2315 The bug was that the ! inversion was incorrectly passed on to the second
2316 item. This has been fixed.
2318 PH/03 Added additional dnslists conditions == and =& which are different from
2319 = and & when the dns lookup returns more than one IP address.
2321 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2322 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2324 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2325 FSYNC, which compiles an option called disable_fsync that allows for
2326 bypassing fsync(). The documentation is heavily laced with warnings.
2328 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2330 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2331 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2332 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2333 including adding "make clean"; (3) Added -fPIC when compiling the test
2334 dynamically loaded module, to get rid of a warning.
2336 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2337 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2338 = 0s. The bug is otherwise harmless.
2340 PH/07 There was a bug in the dovecot authenticator such that the value of
2341 $auth1 could be overwritten, and so not correctly preserved, after a
2342 successful authentication. This usually meant that the value preserved by
2343 the server_setid option was incorrect.
2345 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2347 PH/09 Installed PCRE release 7.0.
2349 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2350 run for batched SMTP input. It is now run at the start of every message
2351 in the batch. While fixing this I discovered that the process information
2352 (output by running exiwhat) was not always getting set for -bs and -bS
2353 input. This is fixed, and it now also says "batched" for BSMTP.
2355 PH/11 Added control=no_pipelining.
2357 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2358 patch, slightly modified), and move the expansion of helo_data till after
2359 the connection is made in the smtp transport (so it can use these
2362 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2364 PH/14 Added log_selector = +pid.
2366 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2368 PH/16 Add ${if forany and ${if forall.
2370 PH/17 Added dsn_from option to vary the From: line in DSNs.
2372 PH/18 Flush SMTP output before performing a callout, unless control =
2373 no_callout_flush is set.
2375 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2376 was true (the default) a successful delivery failed to delete the retry
2377 item, thus causing premature timeout of the address. The bug is now
2380 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2382 PH/21 Long custom messages for fakedefer and fakereject are now split up
2383 into multiline responses in the same way that messages for "deny" and
2384 other ACL rejections are.
2386 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2387 with slight modification.
2389 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2390 draft, changing the syntax and factoring some duplicate code".
2392 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2393 for deliveries of the second and subsequent messages over the same SMTP
2396 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2397 ${reduce, with only minor "tidies".
2399 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2401 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2402 expansion side effects.
2404 PH/27 When a message times out after an over-quota error from an Exim-imposed
2405 quota, the bounce message says "mailbox is full". This message was not
2406 being given when it was a system quota that was exceeded. It now should
2409 MH/03 Made $recipients available in local_scan(). local_scan() already has
2410 better access to the recipient list through recipients_list[], but
2411 $recipients can be useful in postmaster-provided expansion strings.
2413 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2414 in the case of a MAIL command with additional options following the
2415 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2416 were accidentally chopped off.
2418 PH/29 SMTP synchronization checks are implemented when a command is read -
2419 there is a check that no more input is waiting when there shouldn't be
2420 any. However, for some commands, a delay in an ACL can mean that it is
2421 some time before the response is written. In this time, more input might
2422 arrive, invalidly. So now there are extra checks after an ACL has run for
2423 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2424 pipelining has not been advertised.
2426 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2428 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2429 returned in the SMTP response when smtp_return_error_details was false.
2430 This has been fixed.
2432 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2433 instead of the C I/O that was originally supplied, because problems were
2434 reported on Solaris.
2436 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2437 Exim which did not show up earlier: it was assuming that a call to
2438 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2439 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2440 was a macro that became an assignment, so it seemed to work. This has
2441 changed to a proper function call with a void return, hence the compile
2442 error. Exim's code has been fixed.
2444 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2447 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2449 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2451 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2452 because it turns out to be too big or there is a timeout) there is no
2453 "Completed" line in the log. When some messages of this type were
2454 selected by exigrep, they were listed as "not completed". Others were
2455 picked up by some special patterns. I have improved the selection
2456 criteria to be more general.
2458 PH/38 The host_find_failed option in the manualroute router can now be set
2459 to "ignore", to completely ignore a host whose IP address cannot be
2460 found. If all hosts are ignored, the behaviour is controlled by the new
2461 host_all_ignored option.
2463 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2464 homing or because of multiple MX records with /mx) generated more than
2465 one IP address, and the following item turned out to be the local host,
2466 all the secondary addresses of the first item were incorrectly removed
2467 from the list, along with the local host and any following hosts (which
2468 is what is supposed to happen).
2470 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2471 whoever called Exim into the -H file. In the case of the daemon it was
2472 behaving confusingly. When first started, it used values for whoever
2473 started the daemon, but after a SIGHUP it used the Exim user (because it
2474 calls itself on a restart). I have changed the code so that it now always
2477 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2478 message are rejected with the same error (e.g. no authentication or bad
2479 sender address), and a DATA command is nevertheless sent (as can happen
2480 with PIPELINING or a stupid MUA), the error message that was given to the
2481 RCPT commands is included in the rejection of the DATA command. This is
2482 intended to be helpful for MUAs that show only the final error to their
2485 PH/42 Another patch from the Sieve maintainer.
2487 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2488 Eximstats - Fixed some broken HTML links and added missing column headers
2490 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2491 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2493 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2495 PH/43 Yet another patch from the Sieve maintainer.
2497 PH/44 I found a way to check for a TCP/IP connection going away before sending
2498 the response to the final '.' that terminates a message, but only in the
2499 case where the client has not sent further data following the '.'
2500 (unfortunately, this is allowed). However, in many cases there won't be
2501 any further data because there won't be any more messages to send. A call
2502 to select() can be used: if it shows that the input is "ready", there is
2503 either input waiting, or the socket has been closed. An attempt to read
2504 the next input character can distinguish the two cases. Previously, Exim
2505 would have sent an OK response which the client would never have see.
2506 This could lead to message repetition. This fix should cure that, at
2507 least in a lot of common cases.
2509 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2510 advertised in response to EHLO.
2516 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2517 fixed by 4.65/MH/01 (is this a record?) are fixed:
2519 (i) An empty string was always treated as zero by the numeric comparison
2520 operators. This behaviour has been restored.
2522 (ii) It is documented that the numeric comparison operators always treat
2523 their arguments as decimal numbers. This was broken in that numbers
2524 starting with 0 were being interpreted as octal.
2526 While fixing these problems I realized that there was another issue that
2527 hadn't been noticed. Values of message_size_limit (both the global option
2528 and the transport option) were treated as octal if they started with 0.
2529 The documentation was vague. These values are now always treated as
2530 decimal, and I will make that clear in the documentation.
2536 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2537 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2540 MH/01 Don't check that the operands of numeric comparison operators are
2541 integers when their expansion is in "skipping" mode (fixes bug
2542 introduced by 4.64-PH/07).
2544 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2545 child addresses, Exim now panics and dies. Previously, because the count
2546 is held in a short int, deliveries were likely to be lost. As such a
2547 large number of recipients for a single message is ridiculous
2548 (performance will be very, very poor), I have chosen to impose a limit
2549 rather than extend the field.
2555 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2556 leftover -K file (the existence of which was triggered by #402).
2557 While we were at it, introduced process PID as part of the -K
2558 filename. This should rule out race conditions when creating
2561 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2562 processing considerably. Previous code took too long for large mails,
2563 triggering a timeout which in turn triggers #401.
2565 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2566 in the DK code in transports.c. sendfile() is not really portable,
2567 hence the _LINUX specificness.
2569 TF/01 In the add_headers option to the mail command in an Exim filter,
2570 there was a bug that Exim would claim a syntax error in any
2571 header after the first one which had an odd number of characters
2574 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2575 callout verification, Exim cached a "reject" for the entire domain. This
2576 is correct for most verifications, but it is not correct for a recipient
2577 verification with use_sender or use_postmaster set, because in that case
2578 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2579 case of MAIL FROM:<> rejection from other early rejections (e.g.
2580 rejection of HELO). When verifying a recipient using a non-null MAIL
2581 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2582 Whatever the result of the callout, the value of the domain cache is
2583 left unchanged (for any other kind of callout, getting as far as trying
2584 RCPT means that the domain itself is ok).
2586 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2589 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2590 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2591 session, the daemon ignores ECONNECT errors and logs others; it now
2592 ignores EPIPE as well.
2594 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2595 (quoted-printable decoding).
2597 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2598 later the small subsequent patch to fix an introduced bug.
2600 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2602 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2604 PH/08 An error is now given if message_size_limit is specified negative.
2606 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2607 to be given (somewhat) arbitrary names.
2609 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2612 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2613 miscellaneous code fixes
2615 PH/10 Added the log_reject_target ACL modifier to specify where to log
2618 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2619 hostname. This is wrong, because it relates to the incoming message (and
2620 probably the interface on which it is arriving) and not to the outgoing
2621 callout (which could be using a different interface). This has been
2622 changed to use the value of the helo_data option from the smtp transport
2623 instead - this is what is used when a message is actually being sent. If
2624 there is no remote transport (possible with a router that sets up host
2625 addresses), $smtp_active_hostname is used.
2627 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2628 tweaks were necessary in order to get it to work (see also 21 below):
2629 (a) The code assumed that strncpy() returns a negative number on buffer
2630 overflow, which isn't the case. Replaced with Exim's string_format()
2632 (b) There were several signed/unsigned issues. I just did the minimum
2633 hacking in of casts. There is scope for a larger refactoring.
2634 (c) The code used strcasecmp() which is not a standard C function.
2635 Replaced with Exim's strcmpic() function.
2636 (d) The code set only $1; it now sets $auth1 as well.
2637 (e) A simple test gave the error "authentication client didn't specify
2638 service in request". It would seem that Dovecot has changed its
2639 interface. Fortunately there's a specification; I followed it and
2640 changed what the client sends and it appears to be working now.
2642 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2645 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2646 address A is aliased to B and C, where B exists and C does not. Without
2647 -v the output is "A verified" because verification stops after a
2648 successful redirection if more than one address is generated. However,
2649 with -v the child addresses are also verified. Exim was outputting "A
2650 failed to verify" and then showing the successful verification for C,
2651 with its parentage. It now outputs "B failed to verify", showing B's
2652 parentage before showing the successful verification of C.
2654 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2655 look up a TXT record in a specific list after matching in a combined
2658 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2659 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2660 they consult the DNS. I had assumed they would set it the way they
2661 wanted; and indeed my experiments on Linux seem to show that in some
2662 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2663 To be on the safe side, however, I have now made the interface to
2664 host_find_byname() similar to host_find_bydns(), with an argument
2665 containing the DNS resolver options. The host_find_byname() function now
2666 sets these options at its start, just as host_find_bydns() does. The smtp
2667 transport options dns_qualify_single and dns_search_parents are passed to
2668 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2669 of host_find_byname() use the default settings of RES_DEFNAMES
2670 (qualify_single) but not RES_DNSRCH (search_parents).
2672 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2673 spool_read_header() do less string testing, by means of a preliminary
2674 switch on the second character of optional "-foo" lines. (This is
2675 overdue, caused by the large number of possibilities that now exist.
2676 Originally there were few.) While I was there, I also converted the
2677 str(n)cmp tests so they don't re-test the leading "-" and the first
2678 character, in the hope this might squeeze out yet more improvement.
2680 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2681 flag allowing group syntax was set by the header_syntax check but not
2682 turned off, possible causing trouble later; (2) The flag was not being
2683 set at all for the header_verify test, causing "group"-style headers to
2684 be rejected. I have now set it in this case, and also caused header_
2685 verify to ignore an empty address taken from a group. While doing this, I
2686 came across some other cases where the code for allowing group syntax
2687 while scanning a header line wasn't quite right (mostly, not resetting
2688 the flag correctly in the right place). These bugs could have caused
2689 trouble for malformed header lines. I hope it is now all correct.
2691 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2692 with the "reply" argument non-NULL. The code, however (which originally
2693 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2694 but it didn't always do it. This confused somebody who was copying the
2695 code for some other use. I have removed all the tests.
2697 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2698 feature that was used to support insecure browsers during the U.S. crypto
2699 embargo. It requires special client support, and Exim is probably the
2700 only MTA that supported it -- and would never use it because real RSA is
2701 always available. This code has been removed, because it had the bad
2702 effect of slowing Exim down by computing (never used) parameters for the
2703 RSA_EXPORT functionality.
2705 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2706 authenticator to fail if there's a tab character in the incoming data
2707 (there should never be unless someone is messing about, as it's supposed
2708 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2709 if the connection is using TLS or if the remote IP is the same as the
2710 local IP, and the "valid-client-cert option" if a client certificate has
2713 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2714 authenticators. This can be used for authorization after authentication
2715 succeeds. (In the case of plaintext, it servers for both authentication
2718 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2719 if any retry times were supplied.
2721 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2722 connection, where there is no client IP address to check. In this
2723 situation, the verify now always succeeds.
2725 PH/25 Applied John Jetmore's -Mset patch.
2727 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2729 PH/27 In a string expansion for a processed (not raw) header when multiple
2730 headers of the same name were present, leading whitespace was being
2731 removed from all of them, but trailing whitespace was being removed only
2732 from the last one. Now trailing whitespace is removed from each header
2733 before concatenation. Completely empty headers in a concatenation (as
2734 before) are ignored.
2736 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2737 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2739 PH/29 [Removed. This was a change that I later backed out, and forgot to
2740 correct the ChangeLog entry (that I had efficiently created) before
2741 committing the later change.]
2743 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2744 address errors (4xx response to RCPT) over the same connection as other
2745 messages routed to the same hosts. Such deliveries are always "forced",
2746 so retry times are not inspected. This resulted in far too many retries
2747 for the affected addresses. The effect occurred only when there were more
2748 hosts than the hosts_max_try setting in the smtp transport when it had
2749 the 4xx errors. Those hosts that it had tried were not added to the list
2750 of hosts for which the message was waiting, so if all were tried, there
2751 was no problem. Two fixes have been applied:
2753 (i) If there are any address or message errors in an SMTP delivery, none
2754 of the hosts (tried or untried) are now added to the list of hosts
2755 for which the message is waiting, so the message should not be a
2756 candidate for sending over the same connection that was used for a
2757 successful delivery of some other message. This seems entirely
2758 reasonable: after all the message is NOT "waiting for some host".
2759 This is so "obvious" that I'm not sure why it wasn't done
2760 previously. Hope I haven't missed anything, but it can't do any
2761 harm, as the worst effect is to miss an optimization.
2763 (ii) If, despite (i), such a delivery is accidentally attempted, the
2764 routing retry time is respected, so at least it doesn't keep
2765 hammering the server.
2767 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2768 in ${readsocket because some servers need this prod.
2770 PH/32 Added some extra debug output when updating a wait-xxx database.
2772 PH/33 The hint "could be header name not terminated by colon", which has been
2773 given for certain expansion errors for a long time, was not being given
2774 for the ${if def:h_colon_omitted{... case.
2776 PH/34 The spec says: "With one important exception, whenever a domain list is
2777 being scanned, $domain contains the subject domain." There was at least
2778 one case where this was not true.
2780 PH/35 The error "getsockname() failed: connection reset by peer" was being
2781 written to the panic log as well as the main log, but it isn't really
2782 panic-worthy as it just means the connection died rather early on. I have
2783 removed the panic log writing for the ECONNRESET error when getsockname()
2786 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2787 runs only) independently of the message's sender address. This meant
2788 that, if the 4xx error was in fact related to the sender, a different
2789 message to the same recipient with a different sender could confuse
2790 things. In particular, this can happen when sending to a greylisting
2791 server, but other circumstances could also provoke similar problems.
2792 I have changed the default so that the retry time for these errors is now
2793 based a combination of the sender and recipient addresses. This change
2794 can be overridden by setting address_retry_include_sender=false in the
2797 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2798 remote server are returned as part of bounce messages. This was not
2799 happening for LMTP over a pipe (the lmtp transport), but now it is the
2800 same for both kinds of LMTP.
2802 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2803 in header lines that were in fact CNAMEs. This is no longer the case.
2805 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2806 and queue runs started by the daemon processed all messages. This has
2807 been fixed so that -R and -S can now usefully be given with -q<time>.
2809 PH/40 Import PCRE release 6.7 (fixes some bugs).
2811 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2813 PH/42 Give an error if -q is specified more than once.
2815 PH/43 Renamed the variables $interface_address and $interface_port as
2816 $received_ip_address and $received_port, to make it clear that these
2817 values apply to message reception, and not to the outgoing interface when
2818 a message is delivered. (The old names remain recognized, of course.)
2820 PH/44 There was no timeout on the connect() call when using a Unix domain
2821 socket in the ${readsocket expansion. There now is.
2823 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2824 be meaningful with "accept".
2826 SC/01 Eximstats V1.43
2827 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2829 SC/02 Eximstats V1.44
2830 Use a glob alias rather than an array ref in the generated
2831 parser. This improves both readability and performance.
2833 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2834 Collect SpamAssassin and rejection statistics.
2835 Don't display local sender or destination tables unless
2836 there is data to show.
2837 Added average volumes into the top table text output.
2839 SC/04 Eximstats V1.46
2840 Collect data on the number of addresses (recipients)
2841 as well as the number of messages.
2843 SC/05 Eximstats V1.47
2844 Added 'Message too big' to the list of mail rejection
2845 reasons (thanks to Marco Gaiarin).
2847 SC/06 Eximstats V1.48
2848 Mainlog lines which have GMT offsets and are too short to
2849 have a flag are now skipped.
2851 SC/07 Eximstats V1.49 (Alain Williams)
2852 Added the -emptyok flag.
2854 SC/08 Eximstats V1.50
2855 Fixes for obtaining the IP address from reject messages.
2857 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2858 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2859 whitespace changes from 4.64-PH/27
2861 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2864 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2865 are found, allow negative numbers in numeric criteria)
2867 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2869 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2870 to match changes made in 4.64-PH/43
2872 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2874 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2875 "domain" is actually an IP address, and force a failure. This locks out
2876 those revolvers/nameservers that support "A-for-A" lookups, in
2877 contravention of the specifications.
2879 PH/48 When a host name was looked up from an IP address, and the subsequent
2880 forward lookup of the name timed out, the host name was left in
2881 $sender_host_name, contrary to the specification.
2883 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2884 restricted to single-key lookups, Exim was not diagnosing an error if
2885 * or *@ was used with a query-style lookup.
2887 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2889 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2890 long ago, but noone interested enough thought of it. Let's just say that
2891 the "1.1" means that there are some new functions that weren't there at
2892 some point in the past.
2894 PH/51 Error processing for expansion failure of helo_data from an smtp
2895 transport during callout processing was broken.
2897 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2898 tested/used via the -bh/-bhc/-bs options.
2900 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2901 bug, fixed in subsequent PCRE releases).
2903 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2904 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2906 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2912 SC/01 Use a glob alias rather than an array ref in eximstats generated
2913 parser. This improves both readability and performance.
2915 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2916 Don't display local sender or destination tables in eximstats unless
2917 there is data to show.
2918 Added average volumes into the eximstats top table text output.
2920 SC/03 Collect data on the number of addresses (recipients) as well
2921 as the number of messages in eximstats.
2923 TF/01 Correct an error in the documentation for the redirect router. Exim
2924 does (usually) call initgroups() when daemonizing.
2926 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2927 with consistent privilege compared to when running as a daemon.
2929 TF/03 Note in the spec that $authenticated_id is not set for local
2930 submissions from trusted users.
2932 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2933 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2935 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2936 by adding some example configuration directives to the default
2937 configuration file. A little bit of work is required to uncomment the
2938 directives and define how usernames and passwords are checked, but
2939 there is now a framework to start from.
2941 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2942 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2943 without this. I don't know how relevant this is to other LDAP libraries.
2945 PH/02 Add the verb name to the "unknown ACL verb" error.
2947 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2949 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2951 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2952 directory not expanded when it should be if an expanded home directory
2953 was set for the address (which is overridden by the transport).
2955 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2958 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2959 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2960 because it is too late at that time, and has no effect.
2962 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2963 security issue with \' (bugzilla #107). I could not use the
2964 PQescapeStringConn() function, because it needs a PGconn value as one of
2967 PH/08 When testing addresses using -bt, indicate those final addresses that
2968 are duplicates that would not cause an additional delivery. At least one
2969 person was confused, thinking that -bt output corresponded to deliveries.
2970 (Suppressing duplicates isn't a good idea as you lose the information
2971 about possibly different redirections that led to the duplicates.)
2973 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2974 systems where poll() doesn't work, in particular OS X.
2976 PH/10 Added more information to debugging output for retry time not reached.
2978 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2979 operations in malware.c.
2981 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2984 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2985 syslog_duplication set false, log lines that would normally be written
2986 both the the main log and to the reject log were not written to syslog at
2989 PH/14 In the default configuration, change the use of "message" in ACL warn
2990 statements to "add_header".
2992 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2993 not followed by a command (e.g. "seen endif").
2995 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2996 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2999 PH/17 Added extra conditions to the default value of delay_warning_condition
3003 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3004 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3005 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3008 The Auto-Submitted: and various List- headers are standardised, whereas I
3009 don't think Precedence: ever was.
3011 PH/18 Refactored debugging code in route_finduser() to show more information,
3012 in particular, the error code if getpwnam() issues one.
3014 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3015 This is apparently needed in addition to the PH/07 change above to avoid
3016 any possible encoding problems.
3018 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3019 but not after initializing Perl.
3021 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3022 output them only if debugging. By default they are written stderr,
3023 apparently, which is not desirable.
3025 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3028 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3031 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3033 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3034 authenticated or an ident call has been made. Suppress the default
3035 values for $authenticated_id and $authenticated_sender (but permit -oMai
3036 and -oMas) when testing with -bh.
3038 PH/24 Re-jigged the order of the tests in the default configuration so that the
3039 tests for valid domains and recipients precede the DNS black list and CSA
3040 tests, on the grounds that those ones are more expensive.
3042 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3043 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3044 This bug exists in every version of Exim that I still have, right back to
3047 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3048 However, an attempt to turn on case-sensitivity in a regex key by
3049 including (?-i) didn't work because the subject string was already
3050 lowercased, and the effects were non-intuitive. It turns out that a
3051 one-line patch can be used to allow (?-i) to work as expected.
3057 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3058 other effects) broke the use of negated acl sub-conditions.
3060 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3063 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3064 "Deny" causes Exim to reject the incoming connection with a 554 error.
3065 Unfortunately, if there is a major crisis, such as a disk failure,
3066 tcp-wrappers gives "deny", whereas what one would like would be some
3067 kind of temporary error. A kludge has been added to help with this.
3068 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3069 554 error is used if errno is still zero or contains ENOENT (which occurs
3070 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3073 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3075 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3076 errors. Otherwise a message that provokes a temporary error (when other
3077 messages do not) can cause a whole host to time out.
3079 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3080 addresses were routed directly to files or pipes from a redirect router.
3081 File deliveries just didn't batch; pipe deliveries might have suffered
3084 PH/06 A failure to get a lock for a hints database would erroneously always say
3085 "Failed to get write lock", even when it was really a read lock.
3087 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3088 of 0600. This has been changed to use the value of the lockfile_mode
3089 option (which defaults to 0600).
3091 PH/08 Applied small patch from the Sieve maintainer.
3093 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3094 folder from quota calculations, a direct delivery into this folder messed
3095 up the contents of the maildirsize file. This was because the regex was
3096 used only to exclude .Trash (or whatever) when the size of the mailbox
3097 was calculated. There was no check that a delivery was happening into an
3098 excluded directory. This bug has been fixed by ignoring all quota
3099 processing for deliveries into excluded directories.
3101 PH/10 Added the maildirfolder_create_regex option to appendfile.
3107 PH/01 The code for finding all the local interface addresses on a FreeBSD
3108 system running IPv6 was broken. This may well have applied to all BSD
3109 systems, as well as to others that have similar system calls. The broken
3110 code found IPv4 interfaces correctly, but gave incorrect values for the
3111 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3112 that it would not match correctly against @[] and not recognize the IPv6
3115 PH/02 The ipliteral router was not recognizing addresses of the form user@
3116 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3118 PH/03 Added disable_ipv6.
3120 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3121 decoded form, because it is most often used to construct To: headers
3122 lines in autoreplies, and the decoded form may well be syntactically
3123 invalid. However, $reply_address has leading white space removed, and all
3124 newlines turned into spaces so that the autoreply transport does not
3127 PH/05 If group was specified without a user on a router, and no group or user
3128 was specified on a transport, the group from the router was ignored.
3130 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3131 for visible compile-time settings that can be used to change these
3132 numbers, for those that want even more. Backwards compatibility with old
3133 spool files has been maintained. However, going back to a previous Exim
3134 release will lost any variables that are in spool files.
3136 PH/07 Two small changes when running in the test harness: increase delay when
3137 passing a TCP/IP connection to a new process, in case the original
3138 process has to generate a bounce, and remove special handling of
3139 127.0.0.2 (sic), which is no longer necessary.
3141 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3142 be the same on different OS.
3144 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3147 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3148 whether --show-vars was specified or not
3150 JJ/02 exipick: Added support for new ACL variable spool format introduced
3153 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3154 syntactically invalid From: or Reply-to: line, and a filter used this to
3155 generate an autoreply, and therefore failed to obtain an address for the
3156 autoreply, Exim could try to deliver to a non-existent relative file
3157 name, causing unrelated and misleading errors. What now happens is that
3158 it logs this as a hard delivery error, but does not attempt to create a
3161 PH/11 The exinext utility has a -C option for testing purposes, but although
3162 the given file was scanned by exinext itself; it wasn't being passed on
3163 when Exim was called.
3165 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3166 an end-of-file indication when reading a command response.
3168 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3169 compiled. In many other places in Exim, IPv6 addresses are always
3170 recognized, so I have changed this. It also means that IPv4 domain
3171 literals of the form [IPV4:n.n.n.n] are now always recognized.
3173 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3174 used if the router is not running as root, for example, when verifying at
3175 ACL time, or when using -bh. The debugging output from this situation was
3176 non-existent - all you got was a failure to exec. I have made two
3179 (a) Failures to set uid/gid, the current directory, or a process leader
3180 in a subprocess such as that created by queryprogram now generate
3181 suitable debugging output when -d is set.
3183 (b) The queryprogram router detects when it is not running as root,
3184 outputs suitable debugging information if -d is set, and then runs
3185 the subprocess without attempting to change uid/gid.
3187 PH/15 Minor change to Makefile for building test_host (undocumented testing
3190 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3191 additional section of a DNS packet that returns MX or SRV records.
3192 Instead, it always explicitly searches for A/AAAA records. This avoids
3193 major problems that occur when a DNS server includes only records of one
3194 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3195 fixed another bug: if SRV records were looked up and the corresponding
3196 address records were *not* found in the additional section, the port
3197 values from the SRV records were lost.
3199 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3200 using the correct key (the original address) when searching the retry
3201 rules in order to find which one to use for generating the retry hint.
3203 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3204 adding the default one. Similarly, if it contains a Reply-To: header, the
3205 errors_reply_to option, if set, is not used.
3207 PH/19 When calculating a retry time, Exim used to measure the "time since
3208 failure" by looking at the "first failed" field in the retry record. Now
3209 it does not use this if it is later than than the arrival time of the
3210 message. Instead it uses the arrival time. This makes for better
3211 behaviour in cases where some deliveries succeed, thus re-setting the
3212 "first failed" field. An example is a quota failure for a huge message
3213 when small messages continue to be delivered. Without this change, the
3214 "time since failure" will always be short, possible causing more frequent
3215 delivery attempts for the huge message than are intended.
3216 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3218 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3219 $1, $2, $3) because the numerical variables can be reset during some
3220 expansion items (e.g. "match"), thereby losing the authentication data.
3222 PH/21 Make -bV show the size of off_t variables so that the test suite can
3223 decide whether to run tests for quotas > 2G.
3225 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3226 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3227 filecount value is greater than 2G or if a quota value is greater than 2G
3228 on a system where the size of off_t is not greater than 4, a panic error
3231 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3232 never match. The debug and -bh output now contains an explicit error
3233 message indicating a malformed IPv4 address or mask.
3235 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3236 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3237 PH/23 above applies.
3239 PH/25 Do not write to syslog when running in the test harness. The only
3240 occasion when this arises is a failure to open the main or panic logs
3241 (for which there is an explicit test).
3243 PH/26 Added the /no_tell option to "control=freeze".
3245 PH/27 If a host name lookup failed very early in a connection, for example, if
3246 the IP address matched host_lookup and the reverse lookup yielded a name
3247 that did not have a forward lookup, an error message of the form "no IP
3248 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3249 could be logged. Now it outputs the IP address instead of "NULL".
3251 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3252 allows the sender and the authenticated sender to be set when
3253 submitting a message from within Exim. Since child_open_exim() is
3254 documented for local_scan(), the new function should be too.
3256 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3257 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3258 results in an empty string is now treated as unset.
3260 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3262 PH/31 Added sender_verify_fail logging option.
3264 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3265 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3266 tidied the source and removed it altogether.
3268 PH/33 When a queue run was abandoned because the load average was too high, a
3269 log line was always written; now it is written only if the queue_run log
3270 selector is set. In addition, the log line for abandonment now contains
3271 information about the queue run such as the pid. This is always present
3272 in "start" and "stop" lines but was omitted from the "abandon" line.
3274 PH/34 Omit spaces between a header name and the colon in the error message that
3275 is given when verify = headers_syntax fails (if there are lots of them,
3276 the message gets confusing).
3278 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3279 names, as there are now some PTR records that contain slashes. This check
3280 is only to protect against broken name servers that fall over on strange
3281 characters, so the fact that it applies to all lookups doesn't matter.
3283 PH/36 Now that the new test suite is complete, we can remove some of the
3284 special code in Exim that was needed for the old test suite. For example,
3285 sorting DNS records because real resolvers return them in an arbitrary
3286 order. The new test suite's fake resolver always returns records in the
3289 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3290 bounces) except when queue_only is set, to avoid logging races between
3291 the different processes.
3293 PH/38 Panic-die if .include specifies a non-absolute path.
3295 PH/39 A tweak to the "H" retry rule from its user.
3297 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3298 a label. They prevented compilation on older perls.
3300 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3301 a warning to be raised on newish perls.
3303 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3304 on queue. Changes to match documented behaviour of showing count of
3305 messages matching specified criteria.
3307 PH/40 Changed the default ident timeout from 30s to 5s.
3309 PH/41 Added support for the use of login_cap features, on those BSD systems
3310 that have them, for controlling the resources used by pipe deliveries.
3312 PH/42 The content-scanning code uses fopen() to create files in which to put
3313 message data. Previously it was not paying any attention to the mode of
3314 the files. Exim runs with umask(0) because the rest of the code creates
3315 files with open(), and sets the required mode explicitly. Thus, these
3316 files were ending up world-writeable. This was not a big issue, because,
3317 being within the spool directory, they were not world-accessible. I have
3318 created a function called modefopen, which takes an additional mode
3319 argument. It sets umask(777), creates the file, chmods it to the required
3320 mode, then resets the umask. All the relevant calls to fopen() in the
3321 content scanning code have been changed to use this function.
3323 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3324 to 24 hours. This avoids potential overflow problems when processing G
3325 and H retry rules. I suspect nobody ever tinkers with this value.
3327 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3329 PH/45 When the plaintext authenticator is running as a client, the server's
3330 challenges are checked to ensure they are valid base64 strings. By
3331 default, the authentication attempt is cancelled if an invalid string is
3332 received. Setting client_ignore_invalid_base64 true ignores these errors.
3333 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3334 they are received. Thus, the responses can be made to depend on the
3335 challenges. If an invalid string is ignored, an empty string is placed in
3338 PH/46 Messages that are created by the autoreply transport now contains a
3339 References: header, in accordance with RFCs 2822 and 3834.
3341 PH/47 Added authenticated_sender_force to the smtp transport.
3343 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3345 PH/49 Installed latest patch from the Sieve maintainer.
3347 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3348 was also set, the appendfile transport was unnecessarily scanning a
3349 directory of message files (e.g. for maildir delivery) to find the count
3350 of files (along with the size), even though it did not need this
3351 information. It now does the scan only if it needs to find either the
3352 size of the count of files.
3354 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3356 PH/52 Two bugs concerned with error handling when the smtp transport is
3359 (i) Exim was not creating retry information for temporary errors given
3360 for individual recipients after the DATA command when the smtp transport
3361 was used in LMTP mode. This meant that they could be retried too
3362 frequently, and not timed out correctly.
3364 (ii) Exim was setting the flag that allows error details to be returned
3365 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3366 recipients that were returned after the DATA command.
3368 PH/53 This is related to PH/52, but is more general: for any failing address,
3369 when detailed error information was permitted to be returned to the
3370 sender, but the error was temporary, then after the final timeout, only
3371 "retry timeout exceeded" was returned. Now it returns the full error as
3372 well as "retry timeout exceeded".
3374 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3375 do this, and (what is worse) MTAs that accept it.
3377 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3378 will now be deprecated.
3380 PH/56 New os.c-cygwin from the Cygwin maintainer.
3382 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3383 formats (previously only available in exim formats via -bpr, -bpru,
3384 and -bpra. Now also available in native and exiqgrep formats)
3386 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3387 with very large, slow to parse queues
3389 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3391 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3393 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3394 responses to authentication challenges, though it was showing the
3395 challenges; (ii) I've removed the CR characters from the debug output for
3398 PH/58 Allow for the insertion of a newline as well as a space when a string
3399 is turned into more than one encoded-word during RFC 2047 encoding. The
3400 Sieve code now uses this.
3402 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3403 data_4xx, lost_connection, tls_required.
3405 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3406 message was being sent as an SMTP response.
3408 PH/61 Add -l and -k options to exicyclog.
3410 PH/62 When verifying, if an address was redirected to one new address, so that
3411 verification continued, and the new address failed or deferred after
3412 having set something in $address_data, the value of $address_data was not
3413 passed back to the ACL. This was different to the case when no
3414 redirection occurred. The value is now passed back in both cases.
3416 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3417 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3418 use login_cap.h, so on its own it isn't the distinguishing feature. The
3419 new name refers directly to the setclassresources() function.
3421 PH/65 Added configuration files for NetBSD3.
3423 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3425 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3426 is preferred over IPv4.
3428 PH/68 The bounce_return_message and bounce_return_body options were not being
3429 honoured for bounces generated during the reception of non-SMTP messages.
3430 In particular, this applied to messages rejected by the ACL. This bug has
3431 been fixed. However, if bounce_return_message is true and bounce_return_
3432 body is false, the headers that are returned for a non-SMTP message
3433 include only those that have been read before the error was detected.
3434 (In the case of an ACL rejection, they have all been read.)
3436 PH/69 The HTML version of the specification is now built in a directory called
3437 spec_html instead of spec.html, because the latter looks like a path with
3438 a MIME-type, and this confuses some software.
3440 PH/70 Catch two compiler warnings in sieve.c.
3442 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3443 function verify_get_ident() calls ip_connect() to connect a socket, but
3444 if the "connect()" function timed out, ip_connect() used to close the
3445 socket. However, verify_get_ident() also closes the socket later, and in
3446 between Exim writes to the log, which may get opened at this point. When
3447 the socket was closed in ip_connect(), the log could get the same file
3448 descriptor number as the socket. This naturally causes chaos. The fix is
3449 not to close the socket in ip_connect(); the socket should be closed by
3450 the function that creates it. There was only one place in the code where
3451 this was missing, in the iplookup router, which I don't think anybody now
3452 uses, but I've fixed it anyway.
3454 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3455 well as to direct DNS lookups. Otherwise the handling of names in host
3456 lists is inconsistent and therefore confusing.
3462 PH/01 Two changes to the default runtime configuration:
3464 (1) Move the checks for relay_from_hosts and authenticated clients from
3465 after to before the (commented out) DNS black list checks.
3467 (2) Add control=submission to the relay_from_hosts and authenticated
3468 clients checks, on the grounds that messages accepted by these
3469 statements are most likely to be submissions.
3471 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3473 (1) Generate an error if the third argument for the ${prvs expansion is
3476 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3479 (3) Reset the variables that are obtained from the first argument of
3480 ${prvscheck and used in the second argument before leaving the code,
3481 because their memory is reclaimed, so using them afterwards may do
3484 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3485 one (it's much easier than Tom thought :-).
3487 (5) Because of (4), we can now allow for the use of $prvscheck_result
3488 inside the third argument.
3490 PH/03 For some reason, the default setting of PATH when running a command from
3491 a pipe transport was just "/usr/bin". I have changed it to
3494 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3495 anything to be listed in the output from -bV.
3497 PH/05 When a filter generated an autoreply, the entire To: header line was
3498 quoted in the delivery log line, like this:
3500 => >A.N.Other <ano@some.domain> <original@ddress> ...
3502 This has been changed so that it extracts the operative address. There
3503 may be more than one such address. If so, they are comma-separated, like
3506 => >ano@some.domain,ona@other.domain <original@ddress> ...
3508 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3509 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3510 not being looked up in the rDNS to get a host name, Exim was showing the
3511 IP address twice in Received: lines, even though the IP addresses were
3512 identical. For example:
3514 Received: from [1.2.3.4] (helo=[1.2.3.4])
3516 However, if the real host name was known, it was omitting the HELO data
3517 if it matched the actual IP address. This has been tidied up so that it
3518 doesn't show the same IP address twice.
3520 PH/07 When both +timestamp and +memory debugging was on, the value given by
3521 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3522 called by the debug printing, thereby overwriting the timestamp buffer.
3523 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3525 PH/08 When the original message was included in an autoreply transport, it
3526 always said "this is a copy of the message, including all the headers",
3527 even if body_only or headers_only was set. It now gives an appropriate
3530 PH/09 Applied a patch from the Sieve maintainer which:
3532 o fixes some comments
3533 o adds the (disabled) notify extension core
3534 o adds some debug output for the result of if/elsif tests
3535 o points to the current vacation draft in the documentation
3536 and documents the missing references header update
3540 o fixes a bug in processing the envelope test (when testing
3541 multiple envelope elements, the last element determined the
3544 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3545 Electronic Mail") by including:
3547 Auto-submitted: auto-generated
3549 in the messages that it generates (bounce messages and others, such as
3550 warnings). In the case of bounce messages for non-SMTP messages, there was
3551 also a typo: it was using "Auto_submitted" (underscore instead of
3552 hyphen). Since every message generated by Exim is necessarily in response
3553 to another message, thes have all been changed to:
3555 Auto-Submitted: auto-replied
3557 in accordance with these statements in the RFC:
3559 The auto-replied keyword:
3561 - SHOULD be used on messages sent in direct response to another
3562 message by an automatic process,
3564 - MUST NOT be used on manually-generated messages,
3566 - MAY be used on Delivery Status Notifications (DSNs) and Message
3567 Disposition Notifications (MDNs),
3569 - MUST NOT be used on messages generated by automatic or periodic
3570 processes, except for messages which are automatic responses to
3573 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3574 to the default Received: header definition.
3576 PH/12 Added log selector acl_warn_skipped (default on).
3578 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3579 variables because (a) they are in the wrong storage pool and (b) even if
3580 they were copied, it wouldn't work properly because of the caching.
3582 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3583 checking when decoding. Apparently there are clients that generate
3584 overlong encoded strings. Why am I not surprised?
3586 PH/15 If the first argument of "${if match_address" was not empty, but did not
3587 contain an "@" character, Exim crashed. Now it writes a panic log message
3588 and treats the condition as false.
3590 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3592 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3593 extension "envelope-auth". The code is finished and in agreement with
3594 other implementations, but there is no documentation so far and in fact,
3595 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3596 not changing the active code.
3598 Print executed "if" and "elsif" statements when debugging is used. This
3599 helps a great deal to understand what a filter does.
3601 Document more things not specified clearly in RFC3028. I had all this
3602 sorted out, when out of a sudden new issues came to my mind. Oops."
3604 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3607 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3608 canonical form (as documented). However, after a host name lookup from
3609 the IP address, check_host() was doing a simple string comparison with
3610 addresses acquired from the DNS when checking that the found name did
3611 have the original IP as one of its addresses. Since any found IPv6
3612 addresses are likely to be in abbreviated form, the comparison could
3613 fail. Luckily, there already exists a function for doing the comparison
3614 by converting both addresses to binary, so now that is used instead of
3615 the text comparison.
3617 PH/20 There was another similar case to PH/19, when a complete host name was
3618 given in a host list; looking up its IP address could give an abbreviated
3619 form, whereas the current host's name might or might not be abbreviated.
3620 The same fix has been applied.
3626 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3627 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3630 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3631 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3633 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3635 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3636 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3637 Also, for "Auto-Submitted", treat anything other than "no" as
3638 non-personal, in accordance with RFC 3834. (Previously it treated
3639 anything starting "auto-" as non-personal.)
3641 TF/01 The control=submission/name=... option had a problem with syntax
3642 errors if the name included a slash character. The /name= option
3643 now slurps the rest of the string, so it can include any characters
3644 but it must come last in the list of options (after /sender_retain
3647 PH/05 Some modifications to the interface to the fake nameserver for the new
3655 TK/01 Added the "success_on_redirect" address verification option. See
3656 NewStuff for rationale and an example.
3658 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3660 PH/02 Patch to exigrep to allow it to work on syslog lines.
3662 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3663 fread() to skip over the body file's header line, because in Cygwin the
3664 header line is locked and is inaccessible.
3666 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3667 co-exist for some time) to make it clear that it is the Exim ID that is
3668 referenced, not the Message-ID: header line.
3670 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3671 string_format() function, because snprintf() does not exist on all
3674 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3675 ${stat: expansion item.
3677 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3678 protocol synchronization error", to keep the pedants happy.
3680 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3681 well as for IRIX systems, when gcc is being used. See the host.c source
3684 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3686 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3689 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3690 transport and to the smtp transport in LMTP mode.
3692 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3694 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3695 run a filter in a subprocess. This could lead to confusion in subsequent
3696 lookups in the parent process. There should also be a search_tidyup() at
3697 the end of the subprocess.
3699 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3700 only if the host matched helo_try_verify_hosts, which caused the
3701 verification to occur when the EHLO/HELO command was issued. The ACL just
3702 tested the remembered result. Now, if a previous verification attempt has
3703 not happened, "verify = helo" does it there and then.
3705 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3707 TK/03 Fix log output including CR from clamd.
3709 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3710 exist provoked a memory error which could cause a segfault.
3712 PH/15 Installed PCRE 6.2
3714 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3716 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3717 of the problem. Specifically, suggested +O2 rather than +O1 for the
3720 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3722 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3723 record, the port was not correctly logged when the outgoing_port log
3724 selector was set (it logged the transort's default port).
3726 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3727 fallback_hosts, and "hosts" in the smtp transport.
3729 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3730 host errors such as "Connection refused".
3732 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3733 authentication with radiusclient 0.4.9:
3735 - Error returned from rc_read_config was caught wrongly
3736 - Username/password not passed on to radius server due to wrong length.
3738 The presumption is that some radiusclient API changes for 4.51/PH/17
3739 were not taken care of correctly. The code is still untested by me (my
3740 Linux distribution still has 0.3.2 of radiusclient), but it was
3741 contributed by a Radius user.
3743 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3744 expanding the "port" option of the smtp transport.
3746 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3747 while reading a MIME header. Thanks to Tom Hughes for a patch.
3749 PH/24 Include config.h inside local_scan.h so that configuration settings are
3752 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3753 that in an ACL for RCPT (for example), you can examine exactly what was
3756 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3757 commands, but it was not correctly comparing the address with the actual
3758 client host address. Thus, it would show the EHLO address in Received:
3759 header lines when this was not necessary.
3761 PH/27 Added the % operator to ${eval:}.
3763 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3764 it should be ignoring failures (because with -C, for example, it has lost
3765 privilege). It wasn't ignoring creation failures other than "already
3768 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3771 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3772 because an input line was too long, either on its own, or by virtue of
3773 too many continuations, the temporary file was not being removed, and the
3774 return code was incorrect.
3776 PH/31 Missing "BOOL" in function definition in filtertest.c.
3778 PH/32 Applied Sieve patches from the maintainer.
3780 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3782 PH/33 Added "verify = not_blind".
3784 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3785 Local/Makefile (with some defaults set). These are used in built scripts
3786 such as exicyclog, but they have never been used in the exim_install
3787 script (though there are many overriding facilities there). I have
3788 arranged that the exim_install script now takes note of these two
3791 PH/35 Installed configuration files for Dragonfly.
3793 PH/36 When a locally submitted message by a trusted user did not contain a
3794 From: header, and the sender address was obtained from -f or from an SMTP
3795 MAIL command, and the trusted user did not use -F to supply a sender
3796 name, $originator_name was incorrectly used when constructing a From:
3797 header. Furthermore, $originator_name was used for submission mode
3798 messages from external hosts without From: headers in a similar way,
3799 which is clearly wrong.
3801 PH/37 Added control=suppress_local_fixups.
3803 PH/38 When log_selector = +received_sender was set, and the addition of the
3804 sender made the log line's construction buffer exactly full, or one byte
3805 less than full, an overflow happened when the terminating "\n" was
3808 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3809 when the result of a list match is failure because a DNS lookup failed.
3811 PH/40 RM_COMMAND is now used in the building process.
3813 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3814 the "build-* directories that it finds.
3816 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3817 address in a domain literal was a prefix of an interface address.
3819 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3820 when verifying a sender address, unless rewrite_headers is false.
3822 PH/44 Wrote a long comment about why errors_to addresses are verified as
3823 recipients, not senders.
3825 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3826 the ratelimit ACL was added.
3828 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3830 PH/46 Added extra information about PostgreSQL errors to the error string.
3832 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3833 avoiding the need to install special zones in a real server. This is
3834 backwards compatible; if it can't find the fake resolver, it drops back.
3835 Thus, both old and new test suites can be run.
3837 TF/02 Added util/ratelimit.pl
3839 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3840 clock is set back in time.
3842 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3843 Candler <B.Candler@pobox.com>.
3845 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3846 OK for addresses that are the result of redirections.
3848 PH/48 A number of further additions for the benefit of the new test suite,
3849 including a fake gethostbyname() that interfaces to the fake DNS resolver
3852 TF/06 The fix for widen_domains has also been applied to qualify_single and
3853 search_parents which are the other dnslookup options that can cause
3856 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3859 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3861 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3862 service" if it encountered an MX record with an empty target hostname.
3863 The message is now "an MX or SRV record indicated no SMTP service".
3865 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3866 if the DNS of the sending site is misconfigured. This is quite a
3867 common situation. This change restores the behaviour of treating a
3868 helo verification defer as a failure.
3870 PH/51 If self=fail was set on a router, the bounce message did not include the
3871 actual error message.
3877 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3879 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3880 command itself ended in error, the underlying message about the transport
3881 filter timeout was being overwritten with the pipe command error. Now the
3882 underlying error message should be appended to the second error message.
3884 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3886 PH/02 Reduce the amount of output that "make" produces by default. Full output
3887 can still be requested.
3889 PH/03 The warning log line about a condition test deferring for a "warn" verb
3890 was being output only once per connection, rather than after each
3891 occurrence (because it was using the same function as for successful
3892 "warn" verbs). This seems wrong, so I have changed it.
3894 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3895 it should not have, which might have caused a crash in the right
3896 circumstances, but probably never did.
3898 PH/04 Installed a modified version of Tony Finch's patch to make submission
3899 mode fix the return path as well as the Sender: header line, and to
3900 add a /name= option so that you can make the user's friendly name appear
3903 TF/03 Added the control = fakedefer ACL modifier.
3905 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3906 Mark Lowes for thorough testing.
3908 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3910 TK/03 Merged latest SRS patch from Miles Wilton.
3912 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3913 with the definition in sysexits.h (which is #included earlier).
3914 Fortunately, Exim does not actually use EX_OK. The code used to try to
3915 preserve the sysexits.h value, by assuming that macro definitions were
3916 scanned for macro replacements. I have been disabused of this notion,
3917 so now the code just undefines EX_OK before #including unistd.h.
3919 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3920 in the smtp transport. When a block could not be written in a single
3921 write() function, the timeout was being re-applied to each part-write.
3922 This seems wrong - if the receiver was accepting one byte at a time it
3923 would take for ever. The timeout is now adjusted when this happens. It
3924 doesn't have to be particularly precise.
3926 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3927 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3929 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3930 without a domain if the check to <postmaster@domain> fails.
3932 SC/01 Eximstats: added -xls and the ability to specify output files
3933 (patch written by Frank Heydlauf).
3935 SC/02 Eximstats: use FileHandles for outputting results.
3937 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3939 SC/04 Eximstats: fixed display of large numbers with -nvr option
3941 SC/05 Eximstats: fixed merging of reports with empty tables.
3943 SC/06 Eximstats: added the -include_original_destination flag
3945 SC/07 Eximstats: removed tabs and trailing whitespace.
3947 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3949 TK/06 MBOX spool code: Add real "From " MBOX separator line
3950 so the .eml file is really in mbox format (even though
3951 most programs do not really care). Patch from Alex Miller.
3953 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3954 The latter is generated from $received_to and is only set if the
3955 message has one envelope recipient. SA can use these headers,
3956 obviously out-of-the-box. Patch from Alex Miller.
3958 PH/08 The ${def test on a variable was returning false if the variable's
3959 value was "0", contrary to what the specification has always said!
3960 The result should be true unless the variable is empty.
3962 PH/09 The syntax error of a character other than { following "${if
3963 def:variable_name" (after optional whitespace) was not being diagnosed.
3964 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3965 accidental colon was present, for example, could give incorrect results.
3967 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3968 result is used (not including appendfile, where other changes are about
3971 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3972 This involved changing a lot of size variables from int to off_t. It
3973 should work with maildirs and everything.
3975 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3976 spamd dying while we are connected to it.
3978 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3981 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3982 function for BDB 4.3.
3984 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3986 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3987 the sources for PCRE have been re-arranged and more files are now
3990 PH/15 The code I had for printing potentially long long variables in PH/11
3991 above was not the best (it lost precision). The length of off_t variables
3992 is now inspected at build time, and an appropriate printing format (%ld
3993 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3994 to be "long long int" or "long int". This is needed for the internal
3995 formatting function string_vformat().
3997 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3998 the configuration file to be ":syslog", then the script "guesses" where
3999 the logs files are, rather than using the compiled in default. In our
4000 case the guess is not the same as the compiled default, so the script
4001 suddenly stopped working when I started to use syslog. The patch checks
4002 to see if log_file_path is "". If so, it attempts to read it from exim
4003 with no configuration file to get the compiled in version, before it
4004 falls back to the previous guessing code."
4006 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4007 implementing BATV in an Exim configuration. See NewStuff for the gory
4010 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4011 Makefile that are specific to HP-UX.
4013 PH/18 If the "use_postmaster" option was set for a recipient callout together
4014 with the "random" option, the postmaster address was used as the MAIL
4015 FROM address for the random test, but not for the subsequent recipient
4016 test. It is now used for both.
4018 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4019 patch removes a few documentation additions to RFC 3028, because the
4020 latest draft now contains them. It adds the new en;ascii-case comparator
4021 and a new error check for 8bit text in MIME parts. Comparator and
4022 require names are now matched exactly. I enabled the subaddress
4023 extension, but it is not well tested yet (read: it works for me)."
4025 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4026 rework some of the code of TK/09 above to avoid the hardwired use of
4027 "%lld" and "long long". Replaced the call to snprintf() with a call to
4030 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4031 records point to non-existent hosts", "retry timeout exceeded", and
4032 "retry time not reached for any host after a long failure period".
4034 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4035 experimental DomainKeys support:
4037 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4038 (2) On an error such as an illegally used "control", the wrong name for
4039 the control was given.
4041 These problems did NOT occur unless DomainKeys support was compiled.
4043 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4045 PH/24 Added ${if match_ip condition.
4047 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4048 databases so that it will be absolutely obvious if a crash occurs in the
4049 DB library. This is a regular occurrence (often caused by mis-matched
4052 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4053 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4054 picked up on a user's system that detects such things. There doesn't seem
4055 to be a gcc warning option for this - only an attribute that has to be
4056 put on the function's prototype. It seems that in Fedora Core 4 they have
4057 set this on a number of new functions. No doubt there will be more in due
4060 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4061 specify a transport. However, if an address that was verified by such a
4062 router was the subject of a callout, Exim crashed because it tried to
4063 read the rcpt_include_affixes from the non-existent transport. Now it
4064 just assumes that the setting of that option is false. This bug was
4065 introduced by 4.51/PH/31.
4067 PH/28 Changed -d+all to exclude +memory, because that information is very
4068 rarely of interest, but it makes the output a lot bigger. People tend to
4069 do -d+all out of habit.
4071 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4072 code in os-type was giving problems when libc.so lives in lib64, like on
4075 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4076 aren't the modern standard, and it seems that some systems' include files
4077 don't always have them. Exim was already checking for some of the newer
4078 ones like T_AAAA, and defining it itself. I've added checks for all the
4079 record types that Exim uses.
4081 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4082 not automatically generating a new one, as it is supposed to. This
4083 prevented TLS from working. If the file did exist, but contained invalid
4084 data, a new version was generated, as expected. It was only the case of a
4085 non-existent file that was broken.
4087 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4088 with a change in libdomainkeys > 0.64.
4090 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4091 from DNS. If the selector record carries the flag, it now has
4092 precedence over the domain-wide flag.
4094 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4096 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4097 the use of a port name that isn't defined in /etc/services) occurred, the
4098 message was deferred as in a normal delivery, and thus remained on the
4099 spool, instead of being failed because of the mua_wrapper setting. This
4100 is now fixed, and I tidied up some of the mua_wrapper messages at the
4103 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4104 the messages in a hash of arrays rather than using individual hashes.
4105 This is a bit cleaner and results in dramatic memory savings, albeit
4106 at a slight CPU cost.
4108 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4109 as requested by Marc Sherman.
4111 SC/10 Eximstats: added histograms for user specified patterns as requested
4114 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4116 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4117 fopen() in the content-scanning modules that did not already have it.
4123 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4124 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4126 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4128 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4130 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4131 to test Sieve filters that use "vacation".
4133 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4134 that changes the way the GnuTLS parameters are stored in the cache file.
4135 The new format can be generated externally. For backward compatibility,
4136 if the data in the cache doesn't make sense, Exim assumes it has read an
4137 old-format file, and it generates new data and writes a new file. This
4138 means that you can't go back to an older release without removing the
4141 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4142 work if there are any delivery delays because "one_time" forces the
4143 parent to be marked "delivered", so its unseen clone is never tried
4144 again. For this reason, Exim now forbids the simultaneous setting of
4147 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4148 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4149 entry if you want to know the details. The fix, however, neglected to
4150 consider the case when local delivery batching is involved. The test for
4151 "previously delivered" was not happening when checking to see if an
4152 address could be batched with a previous (undelivered) one; under
4153 certain circumstances this could lead to multiple deliveries to the same
4156 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4157 in its include files, and this causes problems building Exim.
4159 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4160 header_syntax) but Exim was just ignoring anything given after a slash.
4161 In particular, this caused confusion with an attempt to use "verify =
4162 reverse_host_lookup/defer_ok". An error is now given when options are
4163 supplied for verify items that do not have them. (Maybe reverse_host_
4164 lookup should have a defer_ok option, but that's a different point.)
4166 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4167 defined by RFC 821) to 2048, because there were problems with some AUTH
4168 commands, and RFC 1869 says the size should be increased for extended
4169 SMTP commands that take arguments.
4171 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4174 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4175 "unknown" error; now it says that the functionality isn't in the binary.
4177 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4178 an address' error message when a string expansion fails (syntax or
4179 whatever). Otherwise the password may appear in the log. Following change
4180 PH/42 below, there is no longer a chance of it appearing in a bounce
4183 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4185 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4186 of its addresses was ever tried. (Bugzilla bug #2.)
4188 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4189 the result incorrectly in the debug output. (It correctly added a newline
4190 to what was transported.)
4192 TF/01 Added $received_time.
4194 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4195 commented out examples of how to interface to a virus scanner and to
4196 SpamAssassin. Also added commented examples of av_scanner and
4197 spamd_address settings.
4199 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4200 and controls are allowed in which ACLs. There were a couple of minor
4201 errors. Some of the entries in the conditions table (which is a table of
4202 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4203 negation of where the condition IS allowed.
4205 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4207 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4208 header file does not have a version number, so I've had to invent a new
4209 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4210 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4211 radiusclient), but it was contributed by a Radius user.
4213 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4214 files or directories, for OpenSSL.
4216 PH/19 When an Exim process that is running as root has to create an Exim log
4217 file, it does so in a subprocess that runs as exim:exim so as to get the
4218 ownership right at creation (otherwise, other Exim processes might see
4219 the file with the wrong ownership). There was no test for failure of this
4220 fork() call, which would lead to the process getting stuck as it waited
4221 for a non-existent subprocess. Forks do occasionally fail when resources
4222 run out. I reviewed all the other calls to fork(); they all seem to check
4225 PH/20 When checking for unexpected SMTP input at connect time (before writing
4226 the banner), Exim was not dealing correctly with a non-positive return
4227 from the read() function. If the client had disconnected by this time,
4228 the result was a log entry for a synchronization error with an empty
4229 string after "input=" when read() returned zero. If read() returned -1
4230 (an event I could not check), uninitialized data bytes were printed.
4231 There were reports of junk text (parts of files, etc) appearing after
4234 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4236 PH/22 Added support for macro redefinition, and (re)definition in between
4237 driver and ACL definitions.
4239 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4240 forgetting to use the resulting value; it was using the unexpanded value.
4242 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4243 hadn't been configured. The fix is from Juergen Kreileder, who
4244 understands it better than I do:
4246 "Here's what I see happening with three configured cyrus_sasl
4247 authenticators configured (plain, login, cram-md5):
4249 On startup auth_cyrus_sasl_init() gets called for each of these.
4250 This means three calls to sasl_listmech() without a specified mech_list.
4251 => SASL tests which mechs of all available mechs actually work
4252 => three warnings about OTP not working
4253 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4255 With the patch, sasl_listmech() also gets called three times. But now
4256 SASL's mech_list option is set to the server_mech specified in the the
4257 authenticator. Or in other words, the answer from sasl_listmech()
4258 gets limited to just the mech you're testing for (which is different
4260 => the return list contains just 'plain' or 'login', 'cram-md5' or
4261 nothing depending on the value of ob->server_mech.
4263 I've just tested the patch: Authentication still works fine,
4264 unavailable mechs specified in the exim configuration are still
4265 caught, and the auth.log warnings about OTP are gone."
4267 PH/25 When debugging is enabled, the contents of the command line are added
4268 to the debugging output, even when log_selector=+arguments is not
4271 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4272 answer is "GNU", and only if the return is "GNU/something" is the answer
4275 PH/27 $acl_verify_message is now set immediately after the failure of a
4276 verification in an ACL, and so is available in subsequent modifiers. In
4277 particular, the message can be preserved by coding like this:
4279 warn !verify = sender
4280 set acl_m0 = $acl_verify_message
4282 Previously, $acl_verify_message was set only while expanding "message"
4283 and "log_message" when a very denied access.
4285 PH/28 Modified OS/os.c-Linux with
4287 -#ifndef OS_LOAD_AVERAGE
4288 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4290 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4291 nomenclature these days.)
4293 PH/29 Installed patch from the Sieve maintainer that adds the options
4294 sieve_useraddress and sieve_subaddress to the redirect router.
4296 PH/30 In these circumstances:
4297 . Two addresses routed to the same list of hosts;
4298 . First host does not offer TLS;
4299 . First host accepts first address;
4300 . First host gives temporary error to second address;
4301 . Second host offers TLS and a TLS session is established;
4302 . Second host accepts second address.
4303 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4304 and peerdn, if requested) that were in fact used only for the second
4307 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4308 attention to any local part prefix or suffix that was matched by the
4309 router that accepted the address. It now behaves in the same way as it
4310 does for delivery: the affixes are removed from the local part unless
4311 rcpt_include_affixes is set on the transport.
4313 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4314 timeout during the DATA phase of an incoming message.
4316 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4317 applied a patch sanctioned by the Sieve maintainer.
4319 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4320 the uid or gid is negative. A case of a negative gid caused this to be
4321 noticed. The fix allows for either to be negative.
4323 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4324 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4327 PH/36 The change PH/12 above was broken. Fixed it.
4329 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4330 the grounds that routing the same address twice would always produce the
4331 same answer. This might have been true once, but it is certainly no
4332 longer true now. Routing a child address may depend on the previous
4333 routing that produced that child. Some complicated redirection strategies
4334 went wrong when messages had multiple recipients, and made Exim's
4335 behaviour dependent on the order in which the addresses were given.
4337 I have moved the duplicate checking until after the routing is complete.
4338 Exim scans the addresses that are assigned to local and remote
4339 transports, and removes any duplicates. This means that more work will be
4340 done, as duplicates will always all be routed, but duplicates are
4341 presumably rare, so I don't expect this is of any significance.
4343 For deliveries to pipes, files, and autoreplies, the duplicate checking
4344 still happens during the routing process, since they are not going to be
4347 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4348 It corrects a timeout issue with spamd. This is Ian's comment: "The
4349 background is that sometimes spamd either never reads data from a
4350 connection it has accepted, or it never writes response data. The exiscan
4351 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4352 blindly assumes that writes won't block so it may never time out."
4354 PH/39 Allow G after quota size as well as K and M.
4356 PH/40 The value set for $authenticated_id in an authenticator may not contain
4357 binary zeroes or newlines because the value is written to log lines and
4358 to spool files. There was no check on this. Now the value is run through
4359 the string_printing() function so that such characters are converted to
4360 printable escape sequences.
4362 PH/41 $message_linecount is a new variable that contains the total number of
4363 lines in the message. Compare $body_linecount, which is the count for the
4366 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4367 bounce and delay warning messages, except in certain special cases, which
4370 (a) An SMTP error message from a remote host;
4371 (b) A message specified in a :fail: redirection;
4372 (c) A message specified in a "fail" command in a system filter;
4373 (d) A message specified in a FAIL return from the queryprogram router;
4374 (e) A message specified by the cannot_route_message router option.
4376 In these cases only, Exim does include the error details in bounce and
4377 warning messages. There are also a few cases where bland messages such
4378 as "unrouteable address" or "local delivery error" are given.
4380 PH/43 $value is now also set for the "else" part of a ${run expansion.
4382 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4383 being worked on, but at least Exim now implements the latest version to
4386 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4387 process to complete was treated as a delivery failure, a timeout while
4388 writing the message to the pipe was logged, but erroneously treated as a
4389 successful delivery. Such timeouts include transport filter timeouts. For
4390 consistency with the overall process timeout, these timeouts are now
4391 treated as errors, giving rise to delivery failures by default. However,
4392 there is now a new Boolean option for the pipe transport called
4393 timeout_defer, which, if set TRUE, converts the failures into defers for
4394 both kinds of timeout. A transport filter timeout is now identified in
4397 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4398 systems where "make" and "gmake" are different, calling "gmake" at top
4399 level broke things. I've arranged for the value of $(MAKE) to be passed
4400 from the Makefile to this script so that it can call the same version of
4404 A note about Exim versions 4.44 and 4.50
4405 ----------------------------------------
4407 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4408 changes of various kinds. As a consequence, a big documentation update was
4409 needed. This delayed the release for rather longer than seemed good, especially
4410 in the light of a couple of (minor) security issues. Therefore, the changes
4411 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4412 release. So 4.44 and 4.50 are in effect two different branches that both start
4415 I have left the 4.50 change log unchanged; it contains all the changes since
4416 4.43. The change log for 4.44 is below; many of its items are identical to
4417 those for 4.50. This seems to be the most sensible way to preserve the
4418 historical information.
4424 1. Minor wording change to the doc/README.SIEVE file.
4426 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4427 computation of the current number of files was incorrect.
4429 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4430 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4433 4. Give more explanation in the error message when the command for a transport
4434 filter fails to execute.
4436 5. There are several places where Exim runs a non-Exim command in a
4437 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4438 was being done only for the command run by the queryprogram router. It is
4439 now done for all such subprocesses. The other cases are: ${run, transport
4440 filters, and the commands run by the lmtp and pipe transports.
4442 6. Added CONFIGURE_GROUP build-time option.
4444 7. Some older OS have a limit of 256 on the maximum number of file
4445 descriptors. Exim was using setrlimit() to set 1000 as a large value
4446 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4447 systems. I've change it so that if it can't get 1000, it tries for 256.
4449 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4450 was an oversight, and furthermore, ever since the addition of extra
4451 controls (e.g. 4.43/32), the checks on when to allow different forms of
4452 "control" were broken. There should now be diagnostics for all cases when a
4453 control that does not make sense is encountered.
4455 9. Added the /retain_sender option to "control=submission".
4457 10. $recipients is now available in the predata ACL (oversight).
4459 11. Tidy the search cache before the fork to do a delivery from a message
4460 received from the command line. Otherwise the child will trigger a lookup
4461 failure and thereby defer the delivery if it tries to use (for example) a
4462 cached ldap connection that the parent has called unbind on.
4464 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4465 of $address_data from the recipient verification was clobbered by the
4466 sender verification.
4468 13. The value of address_data from a sender verification is now available in
4469 $sender_address_data in subsequent conditions in the ACL statement.
4471 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4473 15. Added a new option "connect=<time>" to callout options, to set a different
4476 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4477 was its contents. (It was OK if the option was not defined at all.)
4479 17. A "Completed" log line is now written for messages that are removed from
4480 the spool by the -Mrm option.
4482 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4483 information about exactly what failed.
4485 19. Added -dd to debug only the daemon process.
4487 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4488 handles timeouts, both on the server side and network timeouts. Renamed the
4489 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4491 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4492 It is now set to "smtps".
4494 22. $host_address is now set to the target address during the checking of
4495 ignore_target_hosts.
4497 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4498 being passed; this would have caused $sender_host_name to have been used if
4499 matching the list had actually called for a host name (not very likely,
4500 since this list is usually IP addresses). A host name is now passed as
4503 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4504 code that specifies a non-restarting handler (typically sigaction() in
4505 modern systems) in an attempt to fix a rare and obscure crash bug.
4507 25. Narrowed the window for a race in the daemon that could cause it to ignore
4508 SIGCHLD signals. This is not a major problem, because they are used only to
4509 wake it up if nothing else does.
4511 26. A malformed maildirsize file could cause Exim to calculate negative values
4512 for the mailbox size or file count. Odd effects could occur as a result.
4513 The maildirsize information is now recalculated if the size or filecount
4516 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4517 support for a long time. Removed HAVE_SYS_VFS_H.
4519 28. Installed the latest version of exipick from John Jetmore.
4521 29. In an address list, if the pattern was not a regular expression, an empty
4522 subject address (from a bounce message) matched only if the pattern was an
4523 empty string. Non-empty patterns were not even tested. This was the wrong
4524 because it is perfectly reasonable to use an empty address as part of a
4525 database query. An empty address is now tested by patterns that are
4526 lookups. However, all the other forms of pattern expect the subject to
4527 contain a local part and a domain, and therefore, for them, an empty
4528 address still always fails if the pattern is not itself empty.
4530 30. Exim went into a mad DNS loop when attempting to do a callout where the
4531 host was specified on an smtp transport, and looking it up yielded more
4532 than one IP address.
4534 31. Re-factored the code for checking spool and log partition space into a
4535 function that finds that data and another that does the check. The former
4536 is then used to implement four new variables: $spool_space, $log_space,
4537 $spool_inodes, and $log_inodes.
4539 32. The RFC2047 encoding function was originally intended for short strings
4540 such as real names; it was not keeping to the 75-character limit for
4541 encoded words that the RFC imposes. It now respects the limit, and
4542 generates multiple encoded words if necessary. To be on the safe side, I
4543 have increased the buffer size for the ${rfc2047: expansion operator from
4546 33. It is now permitted to omit both strings after an "if" condition; if the
4547 condition is true, the result is "true". As before, when the second string
4548 is omitted, a false condition yields an empty string. This makes it less
4549 cumbersome to write custom ACL and router conditions.
4551 34. Failure to deliver a bounce message always caused it to be frozen, even if
4552 there was an errors_to setting on the router. The errors_to setting is now
4555 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4556 canonical form (fully expanded) before being placed in
4557 $sender_host_address.
4559 36. The table in the code that translates DNS record types into text (T_A to
4560 "A" for instance) was missing entries for NS and CNAME. It is just possible
4561 that this could have caused confusion if both these types were looked up
4562 for the same domain, because the text type is used as part of Exim's
4563 per-process caching. But the chance of anyone hitting this buglet seems
4566 37. The dnsdb lookup has been extended in a number of ways.
4568 (1) There is a new type, "zns", which walks up the domain tree until it
4569 finds some nameserver records. It should be used with care.
4571 (2) There is a new type, "mxh", which is like "mx" except that it returns
4572 just the host names, not the priorities.
4574 (3) It is now possible to give a list of domains (or IP addresses) to be
4575 looked up. The behaviour when one of the lookups defers can be
4576 controlled by a keyword.
4578 (4) It is now possible to specify the separator character for use when
4579 multiple records are returned.
4581 38. The dnslists ACL condition has been extended: it is now possible to supply
4582 a list of IP addresses and/or domains to be looked up in a particular DNS
4585 39. Added log_selector=+queue_time_overall.
4587 40. When running the queue in the test harness, wait just a tad after forking a
4588 delivery process, to get repeatability of debugging output.
4590 41. Include certificate and key file names in error message when GnuTLS fails
4591 to set them up, because the GnuTLS error message doesn't include the name
4592 of the failing file when there is a problem reading it.
4594 42. Allow both -bf and -bF in the same test run.
4596 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4598 44. The "Exiscan patch" is now merged into the mainline Exim source.
4600 45. Sometimes the final signoff response after QUIT could fail to get
4601 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4602 < 0 before doing a fflush(). This bug looks as though it goes back to the
4603 introduction of TLS in release 3.20, but "sometimes" must have been rare
4604 because the tests only now provoked it.
4606 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4607 (this can affect the format of dates).
4609 47. exim_tidydb, when checking for the continued existence of a message for
4610 which it has found a message-specific retry record, was not finding
4611 messages that were in split spool directories. Consequently, it was
4612 deleting retry records that should have stayed in existence.
4614 48. Steve fixed some bugs in eximstats.
4616 49. The SPA authentication driver was not abandoning authentication and moving
4617 on to the next authenticator when an expansion was forced to fail,
4618 contradicting the general specification for all authenticators. Instead it
4619 was generating a temporary error. It now behaves as specified.
4621 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4622 (the order specifies the preference for clients). The order is now AES256,
4623 AES128, 3DES, ARCFOUR128.
4625 51. Small patch to Sieve code - explicitly set From: when generating an
4628 52. Exim crashed if a remote delivery caused a very long error message to be
4629 recorded - for instance if somebody sent an entire SpamAssassin report back
4630 as a large number of 550 error lines. This bug was coincidentally fixed by
4631 increasing the size of one of Exim's internal buffers (big_buffer) that
4632 happened as part of the Exiscan merge. However, to be on the safe side, I
4633 have made the code more robust (and fixed the comments that describe what
4636 53. Now that there can be additional text after "Completed" in log lines (if
4637 the queue_time_overall log selector is set), a one-byte patch to exigrep
4638 was needed to allow it to recognize "Completed" as not the last thing in
4641 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4642 patch that reportedly fixes this has been added. I am not expert enough to
4643 create a test for it. This is what the patch creator wrote:
4645 "I found a little strange behaviour of ldap code when working with
4646 Windows 2003 AD Domain, where users was placed in more than one
4647 Organization Units. When I tried to give exim partial DN, the exit code
4648 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4649 But simultaneously result of request was absolutely normal ldap result,
4650 so I produce this patch..."
4652 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4653 REFERENCE, so I have modified the code to exclude the patch when that macro
4656 55. Some experimental protocols are using DNS PTR records for new purposes. The
4657 keys for these records are domain names, not reversed IP addresses. The
4658 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4659 leaves it alone. Component reversal etc. now happens only for IP addresses.
4662 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4664 57. Double the size of the debug message buffer (to 2048) so that more of very
4665 long debug lines gets shown.
4667 58. The exicyclog utility now does better if the number of log files to keep
4668 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4670 59. Two changes related to the smtp_active_hostname option:
4672 (1) $smtp_active_hostname is now available as a variable.
4673 (2) The default for smtp_banner uses $smtp_active_hostname instead
4674 of $primary_hostname.
4676 60. The host_aton() function is supposed to be passed a string that is known
4677 to be a valid IP address. However, in the case of IPv6 addresses, it was
4678 not checking this. This is a hostage to fortune. Exim now panics and dies
4679 if the condition is not met. A case was found where this could be provoked
4680 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4681 components; fortuitously, this particular loophole had already been fixed
4682 by change 4.50/55 above.
4684 If there are any other similar loopholes, the new check in host_aton()
4685 itself should stop them being exploited. The report I received stated that
4686 data on the command line could provoke the exploit when Exim was running as
4687 exim, but did not say which command line option was involved. All I could
4688 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4689 running as the user.
4692 61. There was a buffer overflow vulnerability in the SPA authentication code
4693 (which came originally from the Samba project). I have added a test to the
4694 spa_base64_to_bits() function which I hope fixes it.
4697 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4698 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4700 63. The daemon start-up calls getloadavg() while still root for those OS that
4701 need the first call to be done as root, but it missed one case: when
4702 deliver_queue_load_max is set with deliver_drop_privilege. This is
4703 necessary for the benefit of the queue runner, because there is no re-exec
4704 when deliver_drop_privilege is set.
4706 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4707 This has been fixed.
4709 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4710 was in use, was not putting the data itself into the right store pool;
4711 consequently, it could be overwritten for a subsequent message in the same
4712 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4715 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4717 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4718 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4719 the function were treating the return as a boolean value, which happened to
4720 work because 0=false and not-0=true, but is not correct code.
4722 68. The host_aton() function was not handling scoped IPv6 addresses (those
4723 with, for example, "%eth0" on the end) correctly.
4725 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4726 negated items (that is, ~something) in unsigned ints. Some compilers
4727 apparently mutter when there is no cast.
4729 70. If an address verification called from an ACL failed, and did not produce a
4730 user-specific message (i.e. there was only a "system" message), nothing was
4731 put in $acl_verify_message. In this situation, it now puts the system
4734 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4735 session; change 4.31/43 added the unwanted input to the log line - except
4736 that it did not do this in the start of session case. It now does.
4738 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4739 This is wrong and can cause the other end to generate a synchronization
4740 error if it is another Exim or anything else that does the synchronization
4741 check. A QUIT command is no longer sent after a timeout.
4743 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4744 during host lookups.
4746 74. The defer_ok option of callout verification was not working if it was used
4747 when verifying addresses in header lines, that is, for this case:
4749 verify = header_sender/callout=defer_ok
4751 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4752 those file descriptors could be used for SMTP connections. If anything
4753 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4754 could be sent to the SMTP client, causing chaos. The daemon now opens
4755 stdin, stdout, and stderr to /dev/null when it puts itself into the
4758 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4759 log by default. The user can override this with suitable Perl magic.
4761 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4762 the log message when discard triggers, was not working for the DATA ACL or
4763 for the non-SMTP ACL.
4765 78. Error message wording change in sieve.c.
4767 79. If smtp_accept_max_per_host was set, the number of connections could be
4768 restricted to fewer than expected, because the daemon was trying to set up
4769 a new connection before checking whether the processes handling previous
4770 connections had finished. The check for completed processes is now done
4771 earlier. On busy systems, this bug wouldn't be noticed because something
4772 else would have woken the daemon, and it would have reaped the completed
4775 80. If a message was submitted locally by a user whose login name contained one
4776 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4777 It caused a spool format error. I have fixed the spool reading code. A
4778 related problem was that the "from" clause in the Received: line became
4779 illegal because of the space(s). It is now covered by ${quote_local_part.
4781 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4784 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4786 83. Patch from Sieve maintainer to fix unterminated string problem in
4787 "vacation" handling.
4789 84. Some minor changes to the Linux configuration files to help with other
4790 OS variants using glibc.
4792 85. One more patch for Sieve to update vacation handling to latest spec.
4795 ----------------------------------------------------
4796 See the note above about the 4.44 and 4.50 releases.
4797 ----------------------------------------------------
4803 1. Change 4.43/35 introduced a bug that caused file counts to be
4804 incorrectly computed when quota_filecount was set in an appendfile
4807 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4808 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4811 3. Give more explanation in the error message when the command for a transport
4812 filter fails to execute.
4814 4. There are several places where Exim runs a non-Exim command in a
4815 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4816 was being done only for the command run by the queryprogram router. It is
4817 now done for all such subprocesses. The other cases are: ${run, transport
4818 filters, and the commands run by the lmtp and pipe transports.
4820 5. Some older OS have a limit of 256 on the maximum number of file
4821 descriptors. Exim was using setrlimit() to set 1000 as a large value
4822 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4823 systems. I've change it so that if it can't get 1000, it tries for 256.
4825 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4826 was an oversight, and furthermore, ever since the addition of extra
4827 controls (e.g. 4.43/32), the checks on when to allow different forms of
4828 "control" were broken. There should now be diagnostics for all cases when a
4829 control that does not make sense is encountered.
4831 7. $recipients is now available in the predata ACL (oversight).
4833 8. Tidy the search cache before the fork to do a delivery from a message
4834 received from the command line. Otherwise the child will trigger a lookup
4835 failure and thereby defer the delivery if it tries to use (for example) a
4836 cached ldap connection that the parent has called unbind on.
4838 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4839 of $address_data from the recipient verification was clobbered by the
4840 sender verification.
4842 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4843 was its contents. (It was OK if the option was not defined at all.)
4845 11. A "Completed" log line is now written for messages that are removed from
4846 the spool by the -Mrm option.
4848 12. $host_address is now set to the target address during the checking of
4849 ignore_target_hosts.
4851 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4852 being passed; this would have caused $sender_host_name to have been used if
4853 matching the list had actually called for a host name (not very likely,
4854 since this list is usually IP addresses). A host name is now passed as
4857 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4858 code that specifies a non-restarting handler (typically sigaction() in
4859 modern systems) in an attempt to fix a rare and obscure crash bug.
4861 15. Narrowed the window for a race in the daemon that could cause it to ignore
4862 SIGCHLD signals. This is not a major problem, because they are used only to
4863 wake it up if nothing else does.
4865 16. A malformed maildirsize file could cause Exim to calculate negative values
4866 for the mailbox size or file count. Odd effects could occur as a result.
4867 The maildirsize information is now recalculated if the size or filecount
4870 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4871 support for a long time. Removed HAVE_SYS_VFS_H.
4873 18. Updated exipick to current release from John Jetmore.
4875 19. Allow an empty sender to be matched against a lookup in an address list.
4876 Previously the only cases considered were a regular expression, or an
4879 20. Exim went into a mad DNS lookup loop when doing a callout where the
4880 host was specified on the transport, if the DNS lookup yielded more than
4883 21. The RFC2047 encoding function was originally intended for short strings
4884 such as real names; it was not keeping to the 75-character limit for
4885 encoded words that the RFC imposes. It now respects the limit, and
4886 generates multiple encoded words if necessary. To be on the safe side, I
4887 have increased the buffer size for the ${rfc2047: expansion operator from
4890 22. Failure to deliver a bounce message always caused it to be frozen, even if
4891 there was an errors_to setting on the router. The errors_to setting is now
4894 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4895 canonical form (fully expanded) before being placed in
4896 $sender_host_address.
4898 24. Updated eximstats to version 1.33
4900 25. Include certificate and key file names in error message when GnuTLS fails
4901 to set them up, because the GnuTLS error message doesn't include the name
4902 of the failing file when there is a problem reading it.
4904 26. Expand error message when OpenSSL has problems setting up cert/key files.
4907 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4908 (this can affect the format of dates).
4910 28. exim_tidydb, when checking for the continued existence of a message for
4911 which it has found a message-specific retry record, was not finding
4912 messages that were in split spool directories. Consequently, it was
4913 deleting retry records that should have stayed in existence.
4915 29. eximstats updated to version 1.35
4916 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4917 1.35 - bugfix such that pie charts by volume are generated correctly
4919 30. The SPA authentication driver was not abandoning authentication and moving
4920 on to the next authenticator when an expansion was forced to fail,
4921 contradicting the general specification for all authenticators. Instead it
4922 was generating a temporary error. It now behaves as specified.
4924 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4925 (the order specifies the preference for clients). The order is now AES256,
4926 AES128, 3DES, ARCFOUR128.
4928 31. Small patch to Sieve code - explicitly set From: when generating an
4931 32. Exim crashed if a remote delivery caused a very long error message to be
4932 recorded - for instance if somebody sent an entire SpamAssassin report back
4933 as a large number of 550 error lines. This bug was coincidentally fixed by
4934 increasing the size of one of Exim's internal buffers (big_buffer) that
4935 happened as part of the Exiscan merge. However, to be on the safe side, I
4936 have made the code more robust (and fixed the comments that describe what
4939 33. Some experimental protocols are using DNS PTR records for new purposes. The
4940 keys for these records are domain names, not reversed IP addresses. The
4941 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4942 leaves it alone. Component reversal etc. now happens only for IP addresses.
4945 34. The host_aton() function is supposed to be passed a string that is known
4946 to be a valid IP address. However, in the case of IPv6 addresses, it was
4947 not checking this. This is a hostage to fortune. Exim now panics and dies
4948 if the condition is not met. A case was found where this could be provoked
4949 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4950 components; fortuitously, this particular loophole had already been fixed
4951 by change 4.50/55 or 4.44/33 above.
4953 If there are any other similar loopholes, the new check in host_aton()
4954 itself should stop them being exploited. The report I received stated that
4955 data on the command line could provoke the exploit when Exim was running as
4956 exim, but did not say which command line option was involved. All I could
4957 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4958 running as the user.
4961 35. There was a buffer overflow vulnerability in the SPA authentication code
4962 (which came originally from the Samba project). I have added a test to the
4963 spa_base64_to_bits() function which I hope fixes it.
4966 36. The daemon start-up calls getloadavg() while still root for those OS that
4967 need the first call to be done as root, but it missed one case: when
4968 deliver_queue_load_max is set with deliver_drop_privilege. This is
4969 necessary for the benefit of the queue runner, because there is no re-exec
4970 when deliver_drop_privilege is set.
4972 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4973 was in use, was not putting the data itself into the right store pool;
4974 consequently, it could be overwritten for a subsequent message in the same
4975 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4978 38. Sometimes the final signoff response after QUIT could fail to get
4979 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4980 < 0 before doing a fflush(). This bug looks as though it goes back to the
4981 introduction of TLS in release 3.20, but "sometimes" must have been rare
4982 because the tests only now provoked it.
4988 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4989 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4990 Now it returns an integer. A number of calls were still expecting a T/F
4991 return. Fortuitously, in all cases, the tests worked in OK situations,
4992 which is the norm. However, things would have gone wrong on any write
4993 failures on the smtp file descriptor. This function is used when sending
4994 messages over SMTP and also when doing verify callouts.
4996 2. When Exim is called to do synchronous delivery of a locally submitted
4997 message (the -odf or -odi options), it no longer closes stderr before doing
5000 3. Implemented the mua_wrapper option.
5002 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5004 5. Implemented the functions header_remove(), header_testname(),
5005 header_add_at_position(), and receive_remove_recipient(), and exported them
5008 6. If an ACL "warn" statement specified the addition of headers, Exim already
5009 inserted X-ACL-Warn: at the start if there was no header name. However, it
5010 was not making this test for the second and subsequent header lines if
5011 there were newlines in the string. This meant that an invalid header could
5012 be inserted if Exim was badly configured.
5014 7. Allow an ACL "warn" statement to add header lines at the start or after all
5015 the Received: headers, as well as at the end.
5017 8. Added the rcpt_4xx retry error code.
5019 9. Added postmaster_mailfrom=xxx to callout verification option.
5021 10. Added mailfrom=xxxx to the callout verification option, for verify=
5024 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5025 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5027 12. Inserted some casts to stop certain compilers warning when using pointer
5028 differences as field lengths or precisions in printf-type calls (mostly
5029 affecting debugging statements).
5031 13. Added optional readline() support for -be (dynamically loaded).
5033 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5034 same clock tick as a message's arrival, so that its received time was the
5035 same as the "first fail" time on the retry record, and that message
5036 remained on the queue past the ultimate address timeout, every queue runner
5037 would try a delivery (because it was past the ultimate address timeout) but
5038 after another failure, the ultimate address timeout, which should have then
5039 bounced the address, did not kick in. This was a "< instead of <=" error;
5040 in most cases the first failure would have been in the next clock tick
5041 after the received time, and all would be well.
5043 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5044 being recognized when the domain list was tested by the match_domain
5045 condition in an expansion string.
5047 16. Added the ${str2b64: operator.
5049 17. Exim was always calling setrlimit() to set a large limit for the number of
5050 processes, without checking whether the existing limit was already
5051 adequate. (It did check for the limit on file descriptors.) Furthermore,
5052 errors from getrlimit() and setrlimit() were being ignored. Now they are
5053 logged to the main and panic logs, but Exim does carry on, to try to do its
5054 job under whatever limits there are.
5056 18. Imported PCRE 5.0.
5058 19. Trivial typo in log message " temporarily refused connection" (the leading
5061 20. If the log selector return_path_on_delivery was set and an address was
5062 redirected to /dev/null, the delivery process crashed because it assumed
5063 that a return path would always be set for a "successful" delivery. In this
5064 case, the whole delivery is bypassed as an optimization, and therefore no
5067 21. Internal re-arrangement: the function for sending a challenge and reading
5068 a response while authentication was assuming a zero-terminated challenge
5069 string. It's now changed to take a pointer and a length, to allow for
5070 binary data in such strings.
5072 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5074 23. Exim was not respecting finduser_retries when seeking the login of the
5075 uid under which it was called; it was always trying 10 times. (The default
5076 setting of finduser_retries is zero.) Also, it was sleeping after the final
5077 failure, which is pointless.
5079 24. Implemented tls_on_connect_ports.
5081 25. Implemented acl_smtp_predata.
5083 26. If the domain in control=submission is set empty, Exim assumes that the
5084 authenticated id is a complete email address when it generates From: or
5085 Sender: header lines.
5087 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5088 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5089 chown and chgrp in /bin and hostname in /usr/bin.
5091 28. Exim was keeping the "process log" file open after each use, just as it
5092 does for the main log. This opens the possibility of it remaining open for
5093 long periods when the USR1 signal hits a daemon. Occasional processlog
5094 errors were reported, that could have been caused by this. Anyway, it seems
5095 much more sensible not to leave this file open at all, so that is what now
5098 29. The long-running daemon process does not normally write to the log once it
5099 has entered its main loop, and it closes the log before doing so. This is
5100 so that log files can straightforwardly be renamed and moved. However,
5101 there are a couple of unusual error situations where the daemon does write
5102 log entries, and I had neglected to close the log afterwards.
5104 30. The text of an SMTP error response that was received during a remote
5105 delivery was being truncated at 512 bytes. This is too short for some of
5106 the long messages that one sometimes sees. I've increased the limit to
5109 31. It is now possible to make retry rules that apply only when a message has a
5110 specific sender, in particular, an empty sender.
5112 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5113 it possible to be selective about when SMTP synchronization is enforced.
5115 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5117 32. Implemented hosts_connection_nolog.
5119 33. Added an ACL for QUIT.
5121 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5124 35. Added mailbox_size and mailbox_filecount to appendfile.
5126 36. Added control = no_multiline_responses to ACLs.
5128 37. There was a bug in the logic of the code that waits for the clock to tick
5129 in the case where the clock went backwards by a substantial amount such
5130 that the microsecond fraction of "now" was more than the microsecond
5131 fraction of "then" (but the whole seconds number was less).
5133 38. Added support for the libradius Radius client library this is found on
5134 FreeBSD (previously only the radiusclient library was supported).
5140 1. When certain lookups returned multiple values in the form name=value, the
5141 quoting of the values was not always being done properly. Specifically:
5142 (a) If the value started with a double quote, but contained no whitespace,
5144 (b) If the value contained whitespace other than a space character (i.e.
5145 tabs or newlines or carriage returns) it was not quoted.
5146 This fix has been applied to the mysql and pgsql lookups by writing a
5147 separate quoting function and calling it from the lookup code. The fix
5148 should probably also be applied to nisplus, ibase and oracle lookups, but
5149 since I cannot test any of those, I have not disturbed their existing code.
5151 2. A hit in the callout cache for a specific address caused a log line with no
5152 reason for rejecting RCPT. Now it says "Previous (cached) callout
5153 verification failure".
5155 3. There was an off-by-one bug in the queryprogram router. An over-long
5156 return line was truncated at 256 instead of 255 characters, thereby
5157 overflowing its buffer with the terminating zero. As well as fixing this, I
5158 have increased the buffer size to 1024 (and made a note to document this).
5160 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5161 when Exim is waiting for an SMTP response from a remote server, Exim
5162 restarts its select() call on the socket, thereby resetting its timeout.
5163 This is not a problem when such interrupts are rare. Somebody set up a cron
5164 job to run exiwhat every 2 minutes, which is less than the normal select()
5165 timeout (5 or 10 minutes). This meant that the select() timeout never
5166 kicked in because it was always reset. I have fixed this by comparing the
5167 time when an interrupt arrives with the time at the start of the first call
5168 to select(). If more time than the timeout has elapsed, the interrupt is
5169 treated as a timeout.
5171 5. Some internal re-factoring in preparation for the addition of Sieve
5172 extensions (by MH). In particular, the "personal" test is moved to a
5173 separate function, and given an option for scanning Cc: and Bcc: (which is
5174 not set for Exim filters).
5176 6. When Exim created an email address using the login of the caller as the
5177 local part (e.g. when creating a From: or Sender: header line), it was not
5178 quoting the local part when it contained special characters such as @.
5180 7. Installed new OpenBSD configuration files.
5182 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5183 try to make them clearer.
5185 9. Callout options, other than the timeout value, were being ignored when
5186 verifying sender addresses in header lines. For example, when using
5188 verify = header_sender/callout=no_cache
5190 the cache was (incorrectly) being used.
5192 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5193 only the Cygwin environment.
5195 11. When return_path_on_delivery was set as a log selector, if different remote
5196 addresses in the same message used different return paths and parallel
5197 remote delivery occurred, the wrong values would sometimes be logged.
5198 (Whenever a remote delivery process finished, the return path value from
5199 the most recently started remote delivery process was logged.)
5201 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5202 lines when AUTH and/or TLS are in use. This is the "received protocol"
5203 field. Exim used to use "asmtp" for authenticated SMTP, without any
5204 indication (in the protocol name) for TLS use. Now it follows the RFC and
5205 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5206 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5207 names appear in log lines as well as in Received: header lines.
5209 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5210 extensions, and comparison tests, and to fix some bugs.
5212 14. Changes to the "personal" filter test:
5214 (1) The test was buggy in that it was just doing the equivalent of
5215 "contains" tests on header lines. For example, if a user's address was
5216 anne@some.where, the "personal" test would incorrectly be true for
5218 To: susanne@some.where
5220 This test is now done by extracting each address from the header in turn,
5221 and checking the entire address. Other tests that are part of "personal"
5222 are now done using regular expressions (for example, to check local parts
5223 of addresses in From: header lines).
5225 (2) The list of non-personal local parts in From: addresses has been
5226 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5227 taken from the Sieve specification recommendations.
5229 (3) If the message contains any header line starting with "List-" it is
5230 treated as non-personal.
5232 (4) The test for "circular" in the Subject: header line has been removed
5233 because it now seems ill-conceived.
5235 15. Minor typos in src/EDITME comments corrected.
5237 16. Installed latest exipick from John Jetmore.
5239 17. If headers_add on a router specified a text string that was too long for
5240 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5241 of string_sprintf() is now avoided.
5243 18. $message_body_size was not set (it was always zero) when running the DATA
5244 ACL and the local_scan() function.
5246 19. For the "mail" command in an Exim filter, no default was being set for
5247 the once_repeat time, causing a random time value to be used if "once" was
5248 specified. (If the value happened to be <= 0, no repeat happened.) The
5249 default is now 0s, meaning "never repeat". The "vacation" command was OK
5250 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5251 (I found it when inspecting the code).
5253 20. There is now an overall timeout for performing a callout verification. It
5254 defaults to 4 times the callout timeout, which applies to individual SMTP
5255 commands during the callout. The overall timeout applies when there is more
5256 than one host that can be tried. The timeout is checked before trying the
5257 next host. This prevents very long delays if there are a large number of
5258 hosts and all are timing out (e.g. when the network connections are timing
5259 out). The value of the overall timeout can be changed by specifying an
5260 additional sub-option for "callout", called "maxwait". For example:
5262 verify = sender/callout=5s,maxwait=20s
5264 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5265 to the end before writing, but this should make it even safer).
5267 22. Exim was forgetting that it had advertised PIPELINING for the second and
5268 subsequent messages on an SMTP connection. It was also not resetting its
5269 memory on STARTTLS and an internal HELO.
5271 23. When Exim logs an SMTP synchronization error within a session, it now
5272 records whether PIPELINING has been advertised or not.
5274 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5275 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5276 rather than long int.
5278 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5280 26. Added the never_mail option to autoreply.
5286 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5287 crash if the getsockname() call failed; this can happen if a connection is
5288 closed very soon after it is established. The problem was simply in the
5289 order in which certain operations were done, causing Exim to try to write
5290 to the SMTP stream before it had set up the file descriptor. The bug has
5291 been fixed by making things happen in the correct order.
5297 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5298 before the connection was closed, thus losing the rejection response.
5300 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5301 some early Solaris releases, but causes trouble in current releases where
5302 socklen_t is defined.
5304 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5307 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5310 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5311 that releasing the top of it at the end releases what was used for sub-
5312 expansions (unless the block got too big). However, discard this block if
5313 the first thing is a variable or header, so that we can use its block when
5314 it is dynamic (useful for very large $message_headers, for example).
5316 6. Lookups now cache *every* query, not just the most recent. A new, separate
5317 store pool is used for this. It can be recovered when all lookup caches are
5318 flushed. Lookups now release memory at the end of their result strings.
5319 This has involved some general refactoring of the lookup sources.
5321 7. Some code has been added to the store_xxx() functions to reduce the amount
5322 of flapping under certain conditions.
5324 8. log_incoming_interface used to affect only the <= reception log lines. Now
5325 it causes the local interface and port to be added to several more SMTP log
5326 lines, for example "SMTP connection from", and rejection lines.
5328 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5330 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5332 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5333 could be overwritten at the end of the current message (or the start of a
5334 new message if it was set in a HELO ACL). The value is now preserved for
5335 the duration of the SMTP connection.
5337 12. If a transport had a headers_rewrite setting, and a matching header line
5338 contained an unqualified address, that address was qualified, even if it
5339 did not match any rewriting rules. The underlying bug was that the values
5340 of the flags that permit the existence of unqualified sender and recipient
5341 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5342 non-local messages, and by -bnq for local messages) were not being
5343 preserved with the message after it was received.
5345 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5346 "next input=" as part of the text comprising the host identity instead of
5347 the correct text. The code was using the same buffer for two different
5348 strings. However, depending on which order the printing function evaluated
5349 its arguments, the bug did not always show up. Under Linux, for example, my
5350 test suite worked just fine.
5352 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5353 doesn't work with some older versions of Perl. It has been changed to "my",
5354 which in any case is probably the better facility to use.
5356 15. A really picky compiler found some instances of statements for creating
5357 error messages that either had too many or two few arguments for the format
5360 16. The size of the buffer for calls to the DNS resolver has been increased
5361 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5362 for addresses that have a lot of PTR records. This alleviates a problem; it
5363 does not fully solve it.
5365 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5366 buffer now truncates the list and logs the incident, which is the same
5367 action as happens when Exim is looking up a host name and its aliases.
5368 Previously in this situation something unpredictable would happen;
5369 sometimes it was "internal error: store_reset failed".
5371 18. If a server dropped the connection unexpectedly when an Exim client was
5372 using GnuTLS and trying to read a response, the client delivery process
5373 crashed while trying to generate an error log message.
5375 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5376 string, for example:
5378 warn message = H1: something\nH2: something
5380 the text was added as a single header line from Exim's point of view
5381 though it ended up OK in the delivered message. However, searching for the
5382 second and subsequent header lines using $h_h2: did not work. This has been
5383 fixed. Similarly, if a system filter added multiple headers in this way,
5384 the routers could not see them.
5386 20. Expanded the error message when iplsearch is called with an invalid key to
5387 suggest using net-iplsearch in a host list.
5389 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5390 ACLs are no longer actually imposed (and a message to that effect is
5393 22. If a "gecos" field in a passwd entry contained escaped characters, in
5394 particular, if it contained a \" sequence, Exim got it wrong when building
5395 a From: or a Sender: header from that name. A second bug also caused
5396 incorrect handling when an unquoted " was present following a character
5397 that needed quoting.
5399 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5400 was not being matched caselessly.
5402 24. Arranged for all hyphens in the exim.8 source to be escaped with
5405 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5406 properly. Recipient callout cache records were still being keyed to include
5407 the sender, even when use_sender was set false. This led to far more
5408 callouts that were necessary. The sender is no longer included in the key
5409 when use_sender is false.
5411 26. Added "control = submission" modifier to ACLs.
5413 27. Added the ${base62d: operator to decode base 62 numbers.
5415 28. dnsdb lookups can now access SRV records.
5417 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5418 the configuration file.
5420 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5421 (-v) mode. This makes the output for a verbose queue run more intelligible.
5423 31. Added a use_postmaster feature to recipient callouts.
5425 32. Added the $body_zerocount variable, containing the number of binary zero
5426 bytes in the message body.
5428 33. The time of last modification of the "new" subdirectory is now used as the
5429 "mailbox time last read" when there is a quota error for a maildir
5432 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5434 35. Added +ignore_unknown as a special item in host lists.
5436 36. Code for decoding IPv6 addresses in host lists is now included, even if
5437 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5438 address was recognized as an IP address, but was then not correctly decoded
5439 into binary, causing unexpected and incorrect effects when compared with
5446 1. Very minor rewording of debugging text in manualroute to say "list of
5447 hosts" instead of "hostlist".
5449 2. If verify=header_syntax was set, and a header line with an unqualified
5450 address (no domain) and a large number of spaces between the end of the
5451 name and the colon was received, the reception process suffered a buffer
5452 overflow, and (when I tested it) crashed. This was caused by some obsolete
5453 code that should have been removed. The fix is to remove it!
5455 3. When running in the test harness, delay a bit after writing a bounce
5456 message to get a bit more predictability in the log output.
5458 4. Added a call to search_tidyup() just before forking a reception process. In
5459 theory, someone could use a lookup in the expansion of smtp_accept_max_
5460 per_host which, without the tidyup, could leave open a database connection.
5462 5. Added the variables $recipient_data and $sender_data which get set from a
5463 lookup success in an ACL "recipients" or "senders" condition, or a router
5464 "senders" option, similar to $domain_data and $local_part_data.
5466 6. Moved the writing of debug_print from before to after the "senders" test
5469 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5470 problems for message scanning, either using a data ACL, or using
5471 local_scan() because the Received: header was not generated till after they
5472 were called (in order to set the time as the time of reception completion).
5473 I have revised the way this works. The header is now generated after the
5474 body is received, but before the ACL or local_scan() are called. After they
5475 are run, the timestamp in the header is updated.
5481 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5482 before starting a queue runner without re-exec. This happened only when
5483 deliver_drop_privilege was set or when the Exim user was set to root. The
5484 effect of the bug was that timeouts during subsequent deliveries caused
5485 crashes instead of being properly handled. The handler is now left at its
5486 default (and expected) setting.
5488 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5489 message, again when deliver_drop_privilege is set or Exim is run as root.
5490 The bug described in (1) was not present in this case, but the tidying up
5491 of the other signals was missing. I have made the two cases consistent.
5493 3. The ignore_target_hosts setting on a manualroute router was being ignored
5494 for hosts that were looked up using the /MX notation.
5496 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5499 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5500 operated on the sender address. After changing the $sender_address to <>
5501 for the sender address verify, Exim was re-instated it as the original
5502 (before rewriting) address, but remembering that it had rewritten it, so it
5503 wasn't rewriting it again. This bug also had the effect of breaking the
5504 sender address verification caching when the sender address was rewritten.
5506 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5507 This has been changed so that if the ip literal address matches
5508 ignore_target_hosts, the router declines.
5510 7. Added expansion conditions match_domain, match_address, and match_local_
5511 part (NOT match_host).
5513 8. The placeholder for the Received: header didn't have a length field set.
5515 9. Added code to Exim itself and to exim_lock to test for a specific race
5516 condition that could lead to file corruption when using MBX delivery. The
5517 issue is with the lockfile that is created in /tmp. If this file is removed
5518 after a process has opened it but before that process has acquired a lock,
5519 there is the potential for a second process to recreate the file and also
5520 acquire a lock. This could lead to two Exim processes writing to the file
5521 at the same time. The added code performs the same test as UW imapd; it
5522 checks after acquiring the lock that its file descriptor still refers to
5523 the same named file.
5525 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5526 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5529 11. Added the smtp_active_hostname option. If used, this will typically be made
5530 to depend on the incoming interface address. Because $interface_address is
5531 not set up until the daemon has forked a reception process, error responses
5532 that can happen earlier (such as "too many connections") no longer contain
5535 12. If an expansion in a condition on a "warn" statement fails because a lookup
5536 defers, the "warn" statement is abandoned, and the next ACL statement is
5537 processed. Previously this caused the whole ACL to be aborted.
5539 13. Added the iplsearch lookup type.
5541 14. Added ident_timeout as a log selector.
5543 15. Added tls_certificate_verified as a log selector.
5545 16. Added a global option tls_require_ciphers (compare the smtp transport
5546 option of the same name). This controls incoming TLS connections.
5548 17. I finally figured out how to make tls_require_ciphers do a similar thing
5549 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5550 before starting the TLS session.
5552 18. Tabs are now shown as \t in -bP output.
5554 19. If the log selector return_path_on_delivery was set, Exim crashed when
5555 bouncing a message because it had too many Received: header lines.
5557 20. If two routers both had headers_remove settings, and the first one included
5558 a superfluous trailing colon, the final name in the first list and the
5559 first name in the second list were incorrectly joined into one item (with a
5560 colon in the middle).
5566 1. Added -C and -D options to the exinext utility, mainly to make it easier
5567 to include in the automated testing, but these could be helpful when
5568 multiple configurations are in use.
5570 2. The exinext utility was not formatting the output nicely when there was
5571 an alternate port involved in the retry record key, nor when there was a
5572 message id as well (for retries that were specific to a specific message
5573 and a specific host). It was also confused by IPv6 addresses, because of
5574 the additional colons they contain. I have fixed the IPv4 problem, and
5575 patched it up to do a reasonable job for IPv6.
5577 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5578 delivery, the log line now contains "pipelined" if PIPELINING was used.
5580 4. An SMTP transport process used to panic and die if the bind() call to set
5581 an explicit outgoing interface failed. This has been changed; it is now
5582 treated in the same way as a connect() failure.
5584 5. A reference to $sender_host_name in the part of a conditional expansion
5585 that was being skipped was still causing a DNS lookup. This no longer
5588 6. The def: expansion condition was not recognizing references to header lines
5589 that used bh_ and bheader_.
5591 7. Added the _cache feature to named lists.
5593 8. The code for checking quota_filecount in the appendfile transport was
5594 allowing one more file than it should have been.
5596 9. For compatibility with Sendmail, the command line option
5604 and sets the incoming protocol and host name (for trusted callers). The
5605 host name and its colon can be omitted when only the protocol is to be set.
5606 Note the Exim already has two private options, -pd and -ps, that refer to
5607 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5608 "s", but I don't think that's a major issue.
5610 10. A number of refactoring changes to the code, none of which should affect
5613 (a) The number of logging options was getting close to filling up the
5614 32-bit word that was used as a bit map. I have split them into two classes:
5615 those that are passed in the argument to log_write(), and those that are
5616 only ever tested independently outside of that function. These are now in
5617 separate 32-bit words, so there is plenty of room for expansion again.
5618 There is no change in the user interface or the logging behaviour.
5620 (b) When building, for example, log lines, the code previously used a
5621 macro that called string_cat() twice, in order to add two strings. This is
5622 not really sufficiently general. Furthermore, there was one instance where
5623 it was actually wrong because one of the argument was used twice, and in
5624 one call a function was used. (As it happened, calling the function twice
5625 did not affect the overall behaviour.) The macro has been replaced by a
5626 function that can join an arbitrary number of extra strings onto a growing
5629 (c) The code for expansion conditions now uses a table and a binary chop
5630 instead of a serial search (which was left over from when there were very
5631 few conditions). Also, it now recognizes conditions like "pam" even when
5632 the relevant support is not compiled in: a suitably worded error message is
5633 given if an attempt is made to use such a condition.
5635 11. Added ${time_interval:xxxxx}.
5637 12. A bug was causing one of the ddress fields not to be passed back correctly
5638 from remote delivery subprocesses. The field in question was not being
5639 subsequently used, so this caused to problems in practice.
5641 13. Added new log selectors queue_time and deliver_time.
5643 14. Might have fixed a bug in maildirsizefile handling that threw up
5644 "unexpected character" debug warnings, and recalculated the data
5645 unnecessarily. In any case, I expanded the warning message to give more
5648 15. Added the message "Restricted characters in address" to the statements in
5649 the default ACL that block characters like @ and % in local parts.
5651 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5652 Three changes have been made:
5654 (a) There was a serious bug; a negative response to MAIL caused the whole
5655 recipient domain to be cached as invalid, thereby blocking all messages
5656 to all local parts at the same domain, from all senders. This bug has
5657 been fixed. The domain is no longer cached after a negative response to
5658 MAIL if the sender used is not empty.
5660 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5663 (c) A new callout option, "use_sender" has been added for people who want
5664 the modified behaviour.
5670 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5673 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5674 indeed breaks things for older releases.
5676 3. Added additional logging to the case where there is a problem reading data
5677 from a filter that is running in a subprocess using a pipe, in order to
5678 try to track down a specific problem.
5680 4. Testing facility fudge: when running in the test harness and attempting
5681 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5682 getting "No route to host". Convert this to a timeout.
5684 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5687 6. Some OS don't have socklen_t but use size_t instead. This affects the
5688 fifth argument of getsockopt() amongst other things. This is now
5689 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5690 can be set for individual OS. I have set it for SunOS5, OSF1, and
5691 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5692 some earlier ones do not.
5694 7. Change 4.30/15 was not doing the test caselessly.
5696 8. The standard form for an IPv6 address literal was being rejected by address
5697 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5698 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5699 this, as well as the form without the "IPv6" on the front (but only when
5700 address literals are enabled, of course).
5702 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5704 10. Exim crashed if a message with an empty sender address specified by -f
5705 encountered a router with an errors_to setting. This could be provoked only
5706 by a command such as
5710 where an empty string was supplied; "<>" did not hit this bug.
5712 11. Installed PCRE release 4.5.
5714 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5715 remained set. It is now erased.
5717 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5718 times from message ids (which are base 36 rather than the normal 62).
5720 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5721 were being counted as actual protocol errors, and logged if the log
5722 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5723 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5724 and DATA following a set of rejected RCPTs do not count as protocol errors.
5725 In other words, Exim assumes they were pipelined, though this may not
5726 actually be the case. Of course, in all cases the client gets an
5727 appropriate error code.
5729 15. If a lookup fails in an ACL condition, a message about the failure may
5730 be available; it is used if testing the ACL cannot continue, because most
5731 such messages specify what the cause of the deferral is. However, some
5732 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5733 that caused an old message to be retained and used if a later statement
5734 caused a defer, replacing the real cause of the deferral.
5736 16. If an IP address had so many PTR records that the DNS lookup buffer
5737 was not large enough to hold them, Exim could crash while trying to process
5738 the truncated data. It now detects and logs this case.
5740 17. Further to 4.21/58, another change has been made: if (and only if) the
5741 first line of a message (the first header line) ends with CRLF, a bare LF
5742 in a subsequent header line has a space inserted after it, so as not to
5743 terminate the header.
5745 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5746 unnecessarily, used atoi() instead of strtol(), and didn't check the
5747 termination when getting file sizes from file names by regex.
5749 19. Completely re-implemented the support for maildirsize files, in the light
5750 of a number of problems with the previous contributed implementation
5751 (4.30/29). In particular:
5753 . If the quota is zero, the maildirsize file is maintained, but no quota is
5756 . If the maildir directory does not exist, it is created before any attempt
5757 to write a maildirsize file.
5759 . The quota value in the file is just a cache; if the quota is changed in
5760 the transport, the new value overrides.
5762 . A regular expression is available for excluding directories from the
5765 20. The autoreply transport checks the characters in options that define the
5766 message's headers; it allows continued headers, but it was checking with
5767 isspace() after an embedded newline instead of explicitly looking for a
5770 21. If all the "regular" hosts to which an address was routed had passed their
5771 expiry times, and had not reached their retry times, the address was
5772 bounced, even if fallback hosts were defined. Now Exim should go on to try
5775 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5776 equivalent code in the SMTP transport. Some hosts send humungous responses
5777 to HELO/EHLO, more than 1024 it seems.
5779 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5780 gives compiler warnings in some environments. I've now done it "properly",
5783 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5784 (because of problems with the built-in one) was declared to return uschar *
5785 instead of char *, causing compiler failure.
5787 25. Fixed a file descriptor leak when processing alias/forward files.
5789 26. Fixed a minor format string issue in dbfn.c.
5791 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5793 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5794 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5797 29. When a sender address is verified, it is cached, to save repeating the test
5798 when there is more than one recipient in a message. However, when the
5799 verification involves a callout, it is possible for different callout
5800 options to be set for different recipients. It is too complicated to keep
5801 track of this in the cache, so now Exim always runs a verification when a
5802 callout is required, relying on the callout cache for the optimization.
5803 The overhead is duplication of the address routing, but this should not be
5806 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5807 to be verified with callout=postmaster, and the main callout worked but the
5808 postmaster check failed, the verification correctly failed. However, if a
5809 subsequent RCPT command asked for sender verification *without* the
5810 postmaster check, incorrect caching caused this verification also to fail,
5813 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5814 it was not caching the DNS options (qualify_single, search_parents) that
5815 were used when the lookup failed. A subsequent lookup with different
5816 options therefore always gave the same answer, though there were cases
5817 where it should not have. (Example: a "domains = !$mx_any" option on a
5818 dnslookup router: the "domains" option is always processed without any
5819 widening, but the router might have qualify_single set.) Now Exim uses the
5820 cached value only when the same options are set.
5822 32. Added John Jetmore's "exipick" utility to the distribution.
5824 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5825 than a timeout (e.g. a certificate is required, and is not provided), an
5826 Exim server now closes the connection immediately. Previously it waited for
5827 the client to close - but if the client is SSL, it seems that they each
5828 wait for each other, leading to a delay before one of them times out.
5830 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5831 maintained 0.8.x compatibility because I don't think many are using it, and
5832 it is clearly obsolete.
5834 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5837 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5838 client certificate was expired. A simple patch fixes this, though I don't
5839 understand the full logic of why the verify callback is called multiple
5842 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5843 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5844 which causes problems with some clients (such as the Certicom SSL Plus
5845 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5846 disables the coutermeasure allowing Eudora to connect."
5848 38. Exim was not checking that a write() to a log file succeeded. This could
5849 lead to Bad Things if a log got too big, in particular if it hit a file
5850 size limit. Exim now panics and dies if it cannot write to a log file, just
5851 as it does if it cannot open a log file.
5853 39. Modified OS/Makefile-Linux so that it now contains
5855 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5857 The two -D definitions ensure that Exim is compiled with large file
5858 support, which makes it possible to handle log files that are bigger than
5861 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5862 instance) a domain was checked against a named list that involved a lookup,
5863 causing $domain_data to be set, then another domain was checked against the
5864 same list, then the first domain was re-checked, the value of $domain_data
5865 after the final check could be wrong. In particular, if the second check
5866 failed, it could be set empty. This bug probably also applied to
5869 41. The strip_trailing_dot option was not being applied to the address given
5870 with the -f command-line option.
5872 42. The code for reading a message's header from the spool was incrementing
5873 $received_count, but never initializing it. This meant that the value was
5874 incorrect (doubled) while delivering a message in the same process in which
5875 it was received. In the most common configuration of Exim, this never
5876 happens - a fresh exec is done - but it can happen when
5877 deliver_drop_privilege is set.
5879 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5880 it now includes up to 150 characters of the unexpected data in the log
5883 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5884 and building data strings. The size of both of these buffers was 10 000
5885 bytes - far larger than anybody would *ever* want, thought I. Needless to
5886 say, somebody hit the limit. I have increased the maximum line length to
5887 20 000 and the maximum data length of concatenated lines to 100 000. I have
5888 also fixed two bugs, because there was no checking on these buffers. Tsk,
5889 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5890 buffer is too small.
5892 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5893 lsearch lookups. Now it does.
5895 46. When parsing a route_list item in a manualroute router, a fixed-length
5896 buffer was used for the list of hosts. I made this 1024 bytes long,
5897 thinking that nobody would ever have a list of hosts that long. Wrong.
5898 Somebody had a whole pile of complicated expansion conditions, and the
5899 string was silently truncated, leading to an expansion error. It turns out
5900 that it is easier to change to an unlimited length (owing to other changes
5901 that have happened since this code was originally written) than to build
5902 structure for giving a limitation error. The length of the item that
5903 expands into the list of hosts is now unlimited.
5905 47. The lsearch lookup could not handle data where the length of text line was
5906 more than 4095 characters. Such lines were truncated, leading to shortened
5907 data being returned. It should now handle lines of any length.
5909 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5910 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5913 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5914 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5915 no longer happens in comment lines. A list of replacements is now placed
5916 at the head of all of the source files, except those whose only change is
5917 to replace PERL_COMMAND in the very first #! line.
5919 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5920 messages on the queue, with a bottom-up merge sort, using code contributed
5921 by Michael Haardt. This should make operations like -bp somewhat faster on
5922 large queues. It won't affect queue runners, except when queue_run_in_order
5925 51. Installed eximstats 1.31 in the distribution.
5927 52. Added support for SRV lookups to the dnslookup router.
5929 53. If an ACL referred to $message_body or $message_body_end, the value was not
5930 reset for any messages that followed in the same SMTP session.
5932 54. The store-handling optimization for building very long strings was not
5933 differentiating between the different store pools. I don't think this
5934 actually made any difference in practice, but I've tidied it.
5936 55. While running the routers to verify a sender address, $sender_address
5937 was still set to the sender address. This is wrong, because when routing to
5938 send a bounce to the sender, it would be empty. Therefore, I have changed
5939 it so that, while verifying a sender address, $sender_address is set to <>.
5940 (There is no change to what happens when verifying a recipient address.)
5942 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5943 target A or AAAA records (if not already returned) without resetting the
5944 qualify_single or search_parents options of the DNS resolver. These are
5945 inappropriate in this case because the targets of MX and SRV records must
5946 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5947 target that, when qualified, matched something in the local domain. These
5948 two options are now turned off when doing these lookups.
5950 57. It seems that at least some releases of Reiserfs (which does not have the
5951 concept of a fixed number of inodes) returns zero and not -1 for the
5952 number of available inodes. This interacted badly with check_spool_inodes,
5953 which assumed that -1 was the "no such thing" setting. What I have done is
5954 to check that the total number of inodes is greater than zero before doing
5955 the test of how many are available.
5957 58. When a "warn" ACL statement has a log_message modifier, the message is
5958 remembered, and not repeated. This is to avoid a lot of repetition when a
5959 message has many recipients that cause the same warning to be written.
5960 However, Exim was preserving the list of already written lines for an
5961 entire SMTP session, which doesn't seem right. The memory is now reset if a
5962 new message is started.
5964 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5965 debugging output unless log_rewrite was also set.
5967 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5968 of (int)(handle) when we know that handle contains (void *)(-1).
5970 61. The Exim daemon panic-logs an error return when it closes the incoming
5971 connection. However "connection reset by peer" seems to be common, and
5972 isn't really an error worthy of noting specially, so that particular error
5975 62. When Exim is trying to find all the local interfaces, it used to panic and
5976 die if the ioctl to get the interface flags failed. However, it seems that
5977 on at least one OS (Solaris 9) it is possible to have an interface that is
5978 included in the list of interfaces, but for which you get a failure error
5979 for this call. This happens when the interface is not "plumbed" into a
5980 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5981 failure of the "get flags" call assumes that the interface is down.
5983 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5984 makes life easier for people who are doing arithmetic on fields extracted
5985 from dates, where you often get leading zeros that should not be
5986 interpreted as octal.
5988 64. Added qualify_domain to the redirect router, to override the global
5991 65. If a pathologically long header line contained very many addresses (the
5992 report of this problem mentioned 10 000) and each of them was rewritten,
5993 Exim could use up a very large amount of memory. (It kept on making new
5994 copies of the header line as it rewrote, and never released the old ones.)
5995 At the expense of a bit more processing, the header rewriting function has
5996 been changed so that it no longer eats memory in this way.
5998 66. The generation of the Received: header has been moved from the time that a
5999 message starts to be received, to the time that it finishes. The timestamp
6000 in the Received: header should now be very close to that of the <= log
6001 line. There are two side-effects of this change:
6003 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6004 logged header lines no longer include the local Received: line, because
6005 it has not yet been created. The same applies to a copy of the message
6006 that is returned to a non-SMTP sender when a message is rejected.
6008 (b) When a filter file is tested using -bf, no additional Received: header
6009 is added to the test message. After some thought, I decided that this
6012 This change does not affect the value of $received_for. It is still set
6013 after address rewriting, but before local_scan() is called.
6015 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6017 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6018 gave an unhelpful panic error message, and a defer error. I have managed to
6019 change this behaviour so that it now rejects any supplied certificate,
6020 which seems right, as the list of acceptable certificates is empty.
6022 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6023 gave an unhelpful defer error. I have not managed to make this reject any
6024 supplied certificates, but the error message it gives is "no certificate
6025 supplied", which is not helpful.
6027 70. exigrep's output now also includes lines that are not associated with any
6028 message, but which match the given pattern. Implemented by a patch from
6029 Martin Sluka, which also tidied up the Perl a bit.
6031 71. Recipient callout verification, like sender verification, was using <> in
6032 the MAIL FROM command. This isn't really the right thing, since the actual
6033 sender may affect whether the remote host accepts the recipient or not. I
6034 have changed it to use the actual sender in the callout; this means that
6035 the cache record is now keyed on a recipient/sender pair, not just the
6036 recipient address. There doesn't seem to be a real danger of callout loops,
6037 since a callout by the remote host to check the sender would use <>.
6038 [SEE ABOVE: changed after hitting problems.]
6040 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6041 temporary errors. However, in the case of such a code being given after
6042 the end of a data transmission (i.e. after ".") Exim was failing to write
6043 a retry record for the message. (Yes, there was some broken host that was
6044 actually sending 8xx at this point.)
6046 73. An unknown lookup type in a host list could cause Exim to panic-die when
6047 the list was checked. (An example that provoked this was putting <; in the
6048 middle of a list instead of at the start.) If this happened during a DATA
6049 ACL check, a -D file could be left lying around. This kind of configuration
6050 error no longer causes Exim to die; instead it causes a defer error. The
6051 incident is still logged to the main and panic logs.
6053 74. Buglet left over from Exim 3 conversion. The message "too many messages
6054 in one connection" was written to the rejectlog but not the mainlog, except
6055 when address rewriting (yes!) was being logged.
6057 75. Added write_rejectlog option.
6059 76. When a system filter was run not as root (that is, when system_filter_user
6060 was set), the values of the $n variables were not being returned to the
6061 main process; thus, they were not subsequently available in the $sn
6064 77. Added +return_path_on_delivery log selector.
6066 78. A connection timeout was being treated differently from recipients deferred
6067 when testing hosts_max_try with a message that was older than the host's
6068 retry timeout. (The host should not be counted, thus allowing all hosts to
6069 be tried at least once before bouncing.) This may have been the cause of an
6070 occasionally reported bug whereby a message would remain on the queue
6071 longer than the retry timeout, but would be bounced if a delivery was
6072 forced. I say "may" because I never totally pinned down the problem;
6073 setting up timeout/retry tests is difficult. See also the next item.
6075 79. The ultimate address timeout was not being applied to errors that involved
6076 a combination of host plus message (for example, a timeout on a MAIL
6077 command). When an address resolved to a number of possible hosts, and they
6078 were not all tried for each delivery (e.g. because of hosts_max_try), a
6079 message could remain on the queue longer than the retry timeout.
6081 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6084 81. Fixed an obscure SMTP outgoing bug which required at least the following
6085 conditions: (a) there was another message waiting for the same server;
6086 (b) the server returned 5xx to all RCPT commands in the first message so
6087 that the message was not completed; (c) the server dropped the connection
6088 or gave a negative response to the RSET that Exim sends to abort the
6089 transaction. The observed case was a dropped connection after DATA that had
6090 been sent in pipelining mode. That is, the server had advertised PIPELINING
6091 but was not implementing it correctly. The effect of the bug was incorrect
6092 behaviour, such as trying another host, and this could lead to a crash.
6098 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6099 and daemon.c were passed as pointers to ints; they should have been
6100 pointers to socklen_t variables (which are typically unsigned ints).
6102 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6105 3. Fixed a really odd bug that affected only the testing scheme; patching a
6106 certain fixed string in the binary changed the value of another string that
6107 happened to be identical to the end of the original first string.
6109 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6110 name", it returns that address as the IP address. On some operating
6111 systems (e.g. Solaris), it also passes back the IP address string as the
6112 "host name". However, on others (e.g. Linux), it passes back an empty
6113 string. Exim wasn't checking for this, and was changing the host name to an
6114 empty string, assuming it had been canonicalized.
6116 5. Although rare, it is permitted to have more than one PTR record for a given
6117 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6118 all the names associated with an address, because they do in Solaris.
6119 However, it seems that they do not in Linux for data that comes from the
6120 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6121 I found this out when I moved to a new Linux workstation and tried to run
6122 the Exim test suite.
6124 To get round this problem I have changed the code so that it now does its
6125 own call to the DNS to look up PTR records when searching for a host name.
6126 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6127 addresses that are only in /etc/hosts are still found.
6129 This behaviour is, however, controlled by an option called host_lookup_
6130 order, which defaults to "bydns:byaddr". If people want to use the other
6131 order, or indeed, just use one or the other means of lookup, they can
6132 specify it in this variable.
6134 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6135 some operating systems, this comes back from gethostbyaddr() as an empty
6136 string, and this is what Exim used to test for. However, it seems that in
6137 other systems, "." is yielded. Exim now tests for this case too.
6139 7. The values of check_spool_space and check_log_space are now held internally
6140 as a number of kilobytes instead of an absolute number of bytes. If a
6141 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6142 kilobyte. This means that much larger values can be stored.
6144 8. Exim monitor: an attempt to get the action menu when not actually pointing
6145 at a message produces an empty menu entitled "No message selected". This
6146 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6147 no entries in it ("Shell widget menu has zero width and/or height"). So I
6148 have added a single, blank menu entry in this case.
6150 9. Added ${quote_local_part.
6152 10. MIME decoding is now applied to the contents of Subject: header lines when
6155 11. Now that a reference to $sender_host_address automatically causes a reverse
6156 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6157 host lookup before query-style lookups in lists that might use this
6158 variable. This has therefore been abolished, and the "net-" prefix is no
6159 longer necessary for query-style lookups.
6161 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6162 have been a typo for LFLAGS, so it has been changed.
6164 13. The install script calls Exim with "-C /dev/null" in order to find the
6165 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6166 to be output. However, since Exim outputs its version number before the
6167 error, it didn't break the script. It just looked ugly. I fixed this by
6168 always allowing "-C /dev/null" if the caller is root.
6170 14. Ignore overlarge ACL variable number when reading spool file - insurance
6171 against a later release with more variables having written the file.
6173 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6174 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6175 this, as well as the form without the "IPv6" on the front.
6177 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6178 OS/Makefile-Darwin file.
6180 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6181 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6183 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6184 variable. However, some OS use ints for inodes. I've added cast to long int
6185 to get rid of the compiler warning.
6187 19. I had forgotten to lock out "/../" in configuration file names when
6188 ALT_CONFIG_PREFIX was set.
6190 20. Routers used for verification do not need to specify transports. However,
6191 if such a router generated a host list, and callout was configured, Exim
6192 crashed, because it could not find a port number from the (non-existent)
6193 transport. It now assumes port 25 in this circumstance.
6195 21. Added the -t option to exigrep.
6197 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6198 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6199 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6200 them is set, LOOKUP_LSEARCH is forced.
6202 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6204 24. Added sender and host information to the "rejected by local_scan()" log
6205 line; previously there was no indication of these.
6207 25. Added .include_if_exists.
6209 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6210 Linux. It turns out that not all file systems support this. Apparently some
6211 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6212 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6213 not supported on the file descriptor, is now ignored when Exim is trying to
6214 sync a directory. This applies only to Linux.
6216 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6218 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6219 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6222 29. Added support for maildirsize files from supplied patch (modified a bit).
6224 30. The use of :fail: followed by an empty string could lead Exim to respond to
6225 sender verification failures with (e.g.):
6227 550 Verification failed for <xxx>
6228 550 Sender verify failed
6230 where the first response line was missing the '-' that indicates it is not
6231 the final line of the response.
6233 31. The loop for finding the name of the user that called Exim had a hardwired
6234 limit of 10; it now uses the value of finduser_retries, which is used for
6235 all other user lookups.
6237 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6240 33. Exim was neglecting to zero errno before one call of strtol() when
6241 expanding a string and expecting an integer value. On some systems this
6242 resulted in spurious "integer overflow" errors. Also, it was casting the
6243 result into an int without checking.
6245 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6246 did not work. The code looks as if it has *never* worked, though it appears
6247 to have been documented since at least release 1.62. I have made it work.
6249 35. The "timeout_DNS" error in retry rules, also documented since at least
6250 1.62, also never worked. As it isn't clear exactly what this means, and
6251 clearly it isn't a major issue, I have abolished the feature by treating it
6252 as "timeout", and writing a warning to the main and panic logs.
6254 36. The display of retry rules for -brt wasn't always showing the error code
6257 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6258 timeout_connect_A, timeout_connect_MX.
6260 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6261 to the empty sender.
6263 39. The daemon was not analysing the content of -oX till after it had closed
6264 stderr and disconnected from the controlling terminal. This meant that any
6265 syntax errors were only noted on the panic log, and the return code from
6266 the command was 0. By re-arranging the code a little, I've made the
6267 decoding happen first, so such errors now appear on stderr, and the return
6268 code is 1. However, the actual setting up of the sockets still happens in
6269 the disconnected process, so errors there are still only recorded on the
6272 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6273 connections (as happens on some IP stacks) was logged at start up time as
6274 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6275 it from "IPv6 and IPv4", which means that two separate sockets are being
6278 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6279 says whether AF_INET or AF_INET6 was passed as an argument.
6281 42. Exiwhat output was messed up when time zones were included in log
6284 43. Exiwhat now gives more information about the daemon's listening ports,
6285 and whether -tls-on-connect was used.
6287 44. The "port" option of the smtp transport is now expanded.
6289 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6290 silently ignored. Now an error message is written to the main and panic
6293 46. There's a new ACL modifier called "logwrite" which writes to a log file
6294 as soon as it is encountered.
6296 47. Added $local_user_uid and $local_user_gid at routing time.
6298 48. Exim crashed when trying to verify a sender address that was being
6301 49. Exim was recognizing only a space character after ".include". It now also
6302 recognizes a tab character.
6304 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6305 extracting the relevant information from the specification. The man page no
6306 longer contains scrambled data for the -d option, and I've added a section
6307 at the front about calling Exim under different names.
6309 51. Added "extra_headers" argument to the "mail" command in filter files.
6311 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6314 53. Installed eximstats 1.29.
6316 54. Added transport_filter_timeout as a generic transport option.
6318 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6319 Cc: header lines. This was required by RFC 822, but it not required by RFC
6322 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6323 incoming messages that did not have them. Now it does so only if the
6324 message originates locally, that is, if there is no associated remote host
6325 address. When Resent- header lines are present, this applies to the Resent-
6326 lines rather than the non-Resent- lines.
6328 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6329 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6331 58. Messages for configuration errors now include the name of the main
6332 configuration file - useful now that there may be more than one file in a
6333 list (.included file names were always shown).
6335 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6336 for those rare installations that do not start the daemon as root or run it
6337 setuid root. I've cut out the call to initgroups() if the daemon is not
6340 60. The Exim user and group can now be bound into the binary as text strings
6341 that are looked up at the start of Exim's processing.
6343 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6345 62. Added $mailstore_basename variable.
6347 63. Installed patch to sieve.c from Michael Haardt.
6349 64. When Exim failed to open the panic log after failing to open the main log,
6350 the original message it was trying to log was written to stderr and debug
6351 output, but if they were not available (the usual case in production), it
6352 was lost. Now it is written to syslog before the two lines that record the
6353 failures to open the logs.
6355 65. Users' Exim filters run in subprocesses under the user's uid. It is
6356 possible for a "deliver" command or an alias in a "personal" command to
6357 provoke an address rewrite. If logging of address rewriting is configured,
6358 this fails because the process is not running as root or exim. There may be
6359 a better way of dealing with this, but for the moment (because 4.30 needs
6360 to be released), I have disabled address rewrite logging when running a
6361 filter in a non-root, non-exim process.
6367 1. The buildconfig auxiliary program wasn't quoting the value set for
6368 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6369 not defined. This bug was masked in 4.22 by the effect that was fixed in
6372 2. Some messages that were rejected after a message id was allocated were
6373 shown as "incomplete" by exigrep. It no longer does this for messages that
6374 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6376 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6377 have allow_domain_literals set, the ID did not get logged in the <= line.
6378 Domain literals are now always recognized in Message-ID: header lines.
6380 4. The first argument for a ${extract expansion item is the key name or field
6381 number. Leading and trailing spaces in this item were not being ignored,
6382 causing some misleading effects.
6384 5. When deliver_drop_privilege was set, single queue runner processes started
6385 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6386 same command in the process it spins off) were not dropping privilege.
6388 6. When the daemon running as "exim" started a queue runner, it always
6389 re-executed Exim in the spun-off process. This is a waste of effort when
6390 deliver_drop_privilege is set. The new process now just calls the
6391 queue-runner function directly.
6397 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6400 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6401 ignored. Though the use of -oP was forcing the writing of a pid file, it
6402 was always written to the default place.
6404 3. If the message "no IP address found for host xxxx" is generated during
6405 incoming verification, it is now followed by identification of the incoming
6406 connection (so you can more easily find what provoked it).
6408 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6410 5. Added some features to "harden" Exim a bit more against certain attacks:
6412 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6413 be put in Local/Makefile. This is like the never_users runtime option,
6414 but it cannot be overridden. The default setting is "root".
6416 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6417 prefix string with which any file named in a -C command line option
6420 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6421 is retained for -C and -D only if the caller of Exim is root. Without
6422 it, the exim user may also use -C and -D and retain privilege.
6424 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6425 command line option is disabled.
6427 6. Macro names set by the -D option must start with an upper case letter, just
6428 like macro names defined in the configuration file.
6430 7. Added "dereference=" facility to LDAP.
6432 8. Two instances of the typo "uknown" in the source files are fixed.
6434 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6435 the Configure-Makefile script screwed up while processing it.
6437 10. Incorporated PCRE 4.4.
6439 11. The SMTP synchronization check was not operating right at the start of an
6440 SMTP session. For example, it could not catch a HELO sent before the client
6441 waited for the greeting. There is now a check for outstanding input at the
6442 point when the greeting is written. Because of the duplex, asynchronous
6443 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6444 way, but not yet received, when the check is performed.
6446 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6447 on TCP/IP sockets, because this apparently causes some broken clients to
6450 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6451 unchanged) from the Cygwin maintainer.
6453 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6454 was supported instead of testing for mbx. Effectively a typo.
6456 15. The spa authenticator server code was not checking that the input it
6457 received was valid base64.
6459 16. The debug output line for the "set" modifier in ACLs was not showing the
6460 name of the variable that was being set.
6462 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6464 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6465 lookup on its own. Something else was needed to trigger the lookup. For
6466 example, a match in host_lookup or the need for a host name in a host list.
6467 Now, if $sender_host_name is referenced and the host name has not yet been
6468 looked up, a lookup is performed. If the lookup fails, the variable remains
6469 empty, and $host_lookup_failed is set to "1".
6471 19. Added "eqi" as a case-independent comparison operator.
6473 20. The saslauthd authentication condition could segfault if neither service
6474 nor realm was specified.
6476 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6477 error message that was logged was misleading, and incoming SMTP
6478 connections were dropped. The message is now more accurate, and temporary
6479 errors are given to SMTP connections.
6481 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6482 (except RSET) with a 421 error, until QUIT is received. However, it was
6483 failing to send a response to QUIT.
6485 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6486 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6487 ACL is now run after the helo_try_verify_hosts code.
6489 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6490 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6491 case-independent, so other case variants are also recognized). Apparently
6492 some systems use these upper case variants.
6494 25. If more than two messages were waiting for the same host, and a transport
6495 filter was specified for the transport, Exim sent two messages over the
6496 same TCP/IP connection, and then failed with "socket operation on non-
6497 socket" when it tried to send the third.
6499 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6501 27. The extern definition of crypt16() in expand.c was not being excluded when
6502 the OS had its own crypt16() function.
6504 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6505 as a preferred synonym for return_size_limit, both as an option and as an
6508 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6510 30. Changed the default configuration ACL to relax the local part checking rule
6511 for addresses that are not in any local domains. For these addresses,
6512 slashes and pipe symbols are allowed within local parts, but the sequence
6513 /../ is explicitly forbidden.
6515 31. SPA server authentication was not clearing the challenge buffer before
6518 32. log_message in a "warn" ACL statement was writing to the reject log as
6519 well as to the main log, which contradicts the documentation and doesn't
6520 seem right (because no rejection is happening). So I have stopped it.
6522 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6523 However, I am unable to do any testing of this.
6525 34. Fixed an infelicity in the appendfile transport. When checking directories
6526 for a mailbox, to see if any needed to be created, it was accidentally
6527 using path names with one or more superfluous leading slashes; tracing
6528 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6530 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6531 contents are added to the log line that is written for every discarded
6532 recipient. (Previously a log_message setting was ignored.)
6534 36. The ${quote: operator now quotes the string if it is empty.
6536 37. The install script runs exim in order to find its version number. If for
6537 some reason other than non-existence or emptiness, which it checks, it
6538 could not run './exim', it was installing it with an empty version number,
6539 i.e. as "exim-". This error state is now caught, and the installation is
6542 38. An argument was missing from the function that creates an error message
6543 when Exim fails to connect to the socket for saslauthd authentication.
6544 This could cause Exim to crash, or give a corrupted message.
6546 39. Added isip, isip4, and isip6 to ${if conditions.
6548 40. The ACL variables $acl_xx are now saved with the message, and can be
6549 accessed later in routers, transports, and filters.
6551 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6552 strings in the file are not string-expanded.
6554 42. If a MAIL command specified a SIZE value that was too large to fit into an
6555 int variable, the check against message_size_limit failed. Such values are
6556 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6557 day this will have to be increased, but I don't think I want to be around
6558 when emails are that large.
6565 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6566 iconv() is not standard in FreeBSD.
6568 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6569 IPv6 enabled. The observed symptom was a segmentation fault on return from
6570 the function os_common_find_running_interfaces() in src/os.c.
6572 3. In the check_special_case() function in daemon.c I had used "errno" as an
6573 argument name, which causes warnings on some systems. This was basically a
6574 typo, since it was named "eno" in the comments!
6576 4. The code that waits for the clock to tick (at a resolution of some fraction
6577 of a second) so as to ensure message-id uniqueness was always waiting for
6578 at least one whole tick, when it could have waited for less. [This is
6579 almost certainly not relevant at current processor speeds, where it is
6580 unlikely to ever wait at all. But we try to future-proof.]
6582 5. The function that sleeps for a time interval that includes fractions of a
6583 second contained a race. It did not block SIGALRM between setting the
6584 timer, and suspending (a couple of lines later). If the interval was short
6585 and the sigsuspend() was delayed until after it had expired, the suspension
6586 never ended. On busy systems this could lead to processes getting stuck for
6589 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6590 process, before it forks any delivery processes. The open lookup caching
6591 mechanism meant that the open file or database connection was passed into
6592 the delivery process. The problem was that delivery processes always tidy
6593 up cached lookup data. This could cause a problem for the next delivery
6594 process started by the queue runner, because the external queue runner
6595 process does not know about the closure. So the next delivery process
6596 still has data in the lookup cache. In the case of a file lookup, there was
6597 no problem because closing a file descriptor in a subprocess doesn't affect
6598 the parent. However, if the lookup was caching a connection to a database,
6599 the connection was closed, and the second delivery process was likely to
6600 see errors such as "PGSQL: query failed: server closed the connection
6601 unexpectedly". The problem has been fixed by closing all cached lookups
6602 in a queue runner before running a delivery process.
6604 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6605 seem to have the "const" qualifier which it has on other OS. I've
6608 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6609 *of the same type* (client or server) with the same public name that an
6610 error should be diagnosed.
6612 9. When Exim looked up a host name for an IP address, but failed to find the
6613 original IP address when looking up the host name (a safety check), it
6614 output the message "<ip address> does not match any IP for NULL", which was
6615 confusing, to say the least. The bug was that the host name should have
6616 appeared instead of "NULL".
6618 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6619 user that is built into the binary, and the -C or -D options is used, root
6620 privilege is dropped before the configuration file is read. In addition,
6621 logging is switched to stderr instead of the normal log files. If the
6622 configuration then re-defines the Exim user, the unprivileged environment
6623 is probably not what is expected, so Exim logs a panic warning message (but
6626 However, if deliver_drop_privilege is set, the unprivileged state may well
6627 be exactly what is intended, so the warning has been cut out in that case,
6628 and Exim is allowed to try to write to its normal log files.
6634 1. smtp_return_error_details was not giving details for temporary sender
6635 or receiver verification errors.
6637 2. Diagnose a configuration error if two authenticators have the same public
6640 3. Exim used not to create the message log file for a message until the first
6641 delivery attempt. This could be confusing when incoming messages were held
6642 for policy or load reasons. The message log file is now created at the time
6643 the message is received, and an initial "Received" line is written to it.
6645 4. The automatically generated man page for command line options had a minor
6646 bug that caused no ill effects; however, a more serious problem was that
6647 the procedure for building the man page automatically didn't always
6648 operate. Consequently, release 4.20 contains an out-of-date version. This
6649 shouldn't happen again.
6651 5. When building Exim with embedded Perl support, the script that builds the
6652 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6653 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6655 6. The freeze_tell option was not being used for messages that were frozen on
6656 arrival, either by an ACL or by local_scan().
6658 7. Added the smtp_incomplete_transaction log selector.
6660 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6661 was accepting AUTH without a new EHLO.
6663 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6664 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6667 10. Logging of TCP/IP connections (when configured) now happens in the main
6668 daemon process instead of the child process, so that the TCP/IP connection
6669 count is more accurate (but it can never be perfect).
6671 11. The use of "drop" in a nested ACL was not being handled correctly in the
6672 outer ACL. Now, if condition failure induced by the nested "drop" causes
6673 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6674 or "require"), the connection is dropped.
6676 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6677 that yield "discard" can now be used with an "accept" or a "discard" verb,
6678 but an error is generated for any others (because I can't see a useful way
6679 to define what should happen).
6681 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6682 are now processed in the same way as lines in the Exim configuration file.
6683 In particular, continuation lines are supported.
6685 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6687 15. Added -ti meaning -t -i.
6689 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6690 domains, and warn by logging if others are found.
6692 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6693 returned from GIFCONF when Exim is trying to find the list of interfaces on
6694 a host. The code in os.c has been modified to copy each ifreq to an aligned
6695 structure in all cases.
6697 Also, in some cases, the returned ifreq's were being copied to a 'struct
6698 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6699 means the last couple of bytes of an IPv6 address could be chopped if the
6700 ifreq contained only a normal sockaddr (14 bytes storage).
6702 18. Named domain lists were not supported in the hosts_treat_as_local option.
6703 An entry such as +xxxx was not recognized, and was treated as a literal
6706 19. Ensure that header lines added by a DATA ACL are included in the reject log
6707 if the ACL subsequently rejects the message.
6709 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6710 MD5 (which is deprecated).
6712 21. When testing a filter file using -bf, Exim was writing a message when it
6713 took the sender from a "From " line in the message, but it was not doing so
6714 when it took $return_path from a Return-Path: header line. It now does.
6716 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6717 with a valid header line field name (a series of printing characters
6718 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6720 23. Changed "disc" in the source to "disk" to conform to the documentation and
6721 the book and for uniformity.
6723 24. Ignore Sendmail's -Ooption=value command line item.
6725 25. When execve() failed while trying to run a command in a pipe transport,
6726 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6727 could be confused with a return value of 69 from the command itself. This
6728 has been changed to 127, the value the shell returns if it is asked to run
6729 a non-existent command. The wording for the related log line suggests a
6730 non-existent command as the problem.
6732 26. If received_header_text expands to an empty string, do not add a Received:
6733 header line to the message. (Well, it adds a token one on the spool, but
6734 marks it "old" so that it doesn't get used or transmitted.)
6736 27. Installed eximstats 1.28 (addition of -nt option).
6738 28. There was no check for failure on the call to getsockname() in the daemon
6739 code. This can fail if there is a shortage of resources on the system, with
6740 ENOMEM, for example. A temporary error is now given on failure.
6742 29. Contrary to the C standard, it seems that in some environments, the
6743 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6744 program. Exim now does this explicitly; it affects the formatting of
6745 timestamps using strftime().
6747 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6748 complaints. I've now initialized all the variables, to avoid this.
6750 32. Header lines added by a system filter were not being "seen" during
6751 transport-time rewrites.
6753 33. The info_callback() function passed to OpenSSL is set up with type void
6754 (*)(SSL *, int, int), as described somewhere. However, when calling the
6755 function (actually a macro) that sets it up, the type void(*)() is
6756 expected. I've put in a cast to prevent warnings from picky compilers.
6758 34. If a DNS black list lookup found a CNAME record, but there were no A
6759 records associated with the domain it pointed at, Exim crashed.
6761 35. If a DNS black list lookup returned more than one A record, Exim ignored
6762 all but the first. It now scans all returned addresses if a particular IP
6763 value is being sought. In this situation, the contents of the
6764 $dnslist_value variable are a list of all the addresses, separated by a
6767 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6768 to accept a host name and all its aliases if the forward lookup for any of
6769 them yielded the IP address of the incoming connection. Now it accepts only
6770 those names whose forward lookup yields the correct IP address. Any other
6771 names are discarded. This closes a loophole whereby a rogue DNS
6772 administrator could create reverse DNS records to break through a
6773 wildcarded host restriction in an ACL.
6775 37. If a user filter or a system filter that ran in a subprocess used any of
6776 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6777 the wrong values were passed to the pipe command ($thisaddress had the
6778 value of $0, $0 had the value of $1, etc). This bug was introduced by
6779 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6781 38. Improved the line breaking for long SMTP error messages from ACLs.
6782 Previously, if there was no break point between 40 and 75 characters, Exim
6783 left the rest of the message alone. Two changes have been made: (a) I've
6784 reduced the minimum length to 35 characters; (b) if it can't find a break
6785 point between 35 and 75 characters, it looks ahead and uses the first one
6786 that it finds. This may give the occasional overlong line, but at least the
6787 remaining text gets split now.
6789 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6790 file descriptors might be low, and that setting 1000 would always raise it.
6791 It turns out that in some environments, the limit is already over 1000 and
6792 that lowering it causes trouble. So now Exim takes care not to decrease it.
6794 40. When delivering a message, the value of $return_path is set to $sender_
6795 address at the start of routing (routers may change the value). By an
6796 oversight, this default was not being set up when an address was tested by
6797 -bt or -bv, which affected the outcome if any router or filter referred to
6800 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6801 log only when "message" or "log_message" are set. However, if one of the
6802 conditions was an address verification, or a call to a nested ACL, the
6803 messages generated by the underlying test were being passed through. This
6804 no longer happens. The underlying message is available in $acl_verify_
6805 message for both "message" and "log_message" expansions, so it can be
6806 passed through if needed.
6808 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6809 new expansion $bh_ to give the encoded byte string without charset
6810 translation. Translation happens only if iconv() is available; HAVE_ICONV
6811 indicates this at build time. HEADERS_CHARSET gives the charset to
6812 translate to; headers_charset can change it in the configuration, and
6813 "headers charset" can change it in an individual filter file.
6815 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6816 that creates RFC 2047 encoded "words" labels them as that charset instead
6817 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6818 expansion operator; (ii) when Exim creates a From: line for a local
6819 message; (iii) when a header line is rewritten to include a "phrase" part.
6821 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6822 buggy, causing it to skip the first lines of messages whose message ID
6823 ended in 'D'. This would not have bitten before Exim release 4.14, because
6824 message IDs were unlikely to end in 'D' before then. The effect was to have
6825 incorrect size information for certain domains.
6827 45. #include "config.h" was missing at the start of the crypt16.c module. This
6828 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6831 46. If there was a timeout during a "random" callout check, Exim treated it as
6832 a failure of the random address, and carried on sending RSET and the real
6833 address. If the delay was just some slowness somewhere, the response to the
6834 original RCPT would be taken as a response to RSET and so on, causing
6835 mayhem of various kinds.
6837 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6838 when I implemented it. It didn't allow for the fact that some option values
6839 may legitimately be negative (e.g. size_addition), and it didn't even do
6840 the right test for positive values.
6842 48. Domain names in DNS records are case-independent. Exim always looks them up
6843 in lower case. Some resolvers return domain names in exactly the case they
6844 appear in the zone file, that is, they may contain uppercase letters. Not
6845 all resolvers do this - some return always lower case. Exim was treating a
6846 change of case by a resolver as a change of domain, similar to a widening
6847 of a domain abbreviation. This triggered its re-routing code and so it was
6848 trying to route what was effectively the same domain again. This normally
6849 caused routing to fail (because the router wouldn't handle the domain
6850 twice). Now Exim checks for this case specially, and just changes the
6851 casing of the domain that it ultimately uses when it transmits the message
6854 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6857 50. If a filter generated a file delivery with a non-absolute name (possible if
6858 no home directory exists for the router), the forbid_file option was not
6861 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6862 the existing equality matching.
6864 52. Exim was using ints instead of ino_t variables in some places where it was
6865 dealing with inode numbers.
6867 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6868 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6869 called TMPDIR, and if it finds it is different, it changes its value.
6871 54. The smtp_printf() function is now made available to local_scan() so
6872 additional output lines can be written before returning. There is also an
6873 smtp_fflush() function to enable the detection of a dropped connection.
6874 The variables smtp_input and smtp_batched_input are exported to
6877 55. Changed the default runtime configuration: the message "Unknown user"
6878 has been removed from the ACL, and instead placed on the localuser router,
6879 using the cannot_route_message feature. This means that any verification
6880 failures that generate their own messages won't get overridden. Similarly,
6881 the "Unrouteable address" message that was in the ACL for unverifiable
6882 relay addresses has also been removed.
6884 56. Added hosts_avoid_esmtp to the smtp transport.
6886 57. The exicyclog script was not checking for the esoteric option
6887 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6888 will work only if exicyclog is run under the appropriate euid.
6890 58. Following a discussion on the list, the rules by which Exim recognises line
6891 endings on incoming messages have been changed. The -dropcr and drop_cr
6892 options are now no-ops, retained only for backwards compatibility. The
6893 following line terminators are recognized: LF CRLF CR. However, special
6894 processing applies to CR:
6896 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6897 nor a local message in the state where . is a terminator.
6899 (ii) If a bare CR is encountered in a header line, an extra space is added
6900 after the line terminator so as not to end the header. The reasoning
6901 behind this is that bare CRs in header lines are most likely either
6902 to be mistakes, or people trying to play silly games.
6904 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6905 was being incorrectly given as 18 bytes larger than it should have been.
6906 This is a VOB (very old bug).
6908 60. This may never have affected anything current, but just in case it has:
6909 When the local host is found other than at the start of a list of hosts,
6910 the local host, those with the same MX, and any that follow, are discarded.
6911 When the list in question was part of a longer list of hosts, the following
6912 hosts (not currently being processed) were also being discarded. This no
6913 longer happens. I'm not sure if this situation could ever has previously
6916 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6919 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6920 header. This is something that is recommended in a new Internet Draft, and
6921 is something that is documented as being done by Sendmail. There are two
6922 possible values. For messages generated by the autoreply transport, Exim
6925 Auto-Submitted: auto-replied
6927 whereas for all other generated messages (e.g. bounces) it adds
6929 Auto-Submitted: auto-generated
6931 63. The "personal" condition in filters now includes a test for the
6932 Auto-Submitted: header. If it contains the string "auto-" the message it
6933 not considered personal.
6935 64. Added rcpt_include_affixes as a generic transport option.
6937 65. Added queue_only_override (default true).
6939 66. Added the syslog_duplication option.
6941 67. If what should have been the first header line of a message consisted of
6942 a space followed by a colon, Exim was mis-interpreting it as a header line.
6943 It isn't of course - it is syntactically invalid and should therefore be
6944 treated as the start of the message body. The misbehaviour could have
6945 caused a number of strange effects, including loss of data in subsequent
6946 header lines, and spool format errors.
6948 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6949 client host had authenticated. This control can now be exercised by an ACL
6950 for more flexibility.
6952 69. By default, callouts do not happen when testing with -bh. There is now a
6953 variant, -bhc, which does actually run the callout code, including
6954 consulting and updating the callout cache.
6956 70. Added support for saslauthd authentication, courtesy of Alexander
6959 71. If statvfs() failed on the spool or log directories while checking their
6960 size for availability, Exim confusingly gave the error "space shortage".
6961 Furthermore, in debugging mode it crashed with a floating point exception.
6962 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6963 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6964 serious problem, Exim now writes to the main and panic logs when this
6965 happens, with details of the failure. It then refuses to accept the
6966 incoming message, giving the message "spool directory problem" or "log
6967 directory problem" with a 421 code for SMTP messages.
6969 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6970 0, 1, and 2 exist, because some OS complain for execs without them (see
6971 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6972 descriptors. However, the code omitted to check that the open succeeded,
6973 causing mysterious errors if for some reason the permissions on /dev/null
6974 got screwed. Now Exim writes a message to the main and panic logs, and
6975 bombs out if it can't open /dev/null.
6977 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6978 interact so that it is all more flexible. It is supposed to remain
6979 backwards compatible. Also added extra_local_interfaces.
6981 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6982 to bomb out with an assertion failure - to the client this appears as a
6983 connection drop. This problem occurs in the part of the code that was taken
6984 from the Samba project. Fortunately, the assertion is in a very simple
6985 function, so I have fixed this by reproducing the function inline in the
6986 one place where it is called, and arranging for authentication to fail
6987 instead of killing the process with assert().
6989 75. The SPA client code was not working when the server requested OEM rather
6990 than Unicode encoding.
6992 76. Added code to make require_files with a specific uid setting more usable in
6993 the case where statting the file as root fails - usually a non-root-mounted
6994 NFS file system. When this happens and the failure is EACCES, Exim now
6995 forks a subprocess and does the per-uid checking as the relevant uid.
6997 77. Added process_log_path.
6999 78. If log_file_path was not explicitly set, a setting of check_log_space or
7000 check_log_inodes was ignored.
7002 79. If a space check for the spool or log partitions fails, the incident is now
7003 logged. Of course, in the latter case the data may get lost...
7005 80. Added the %p formatting code to string_format() so that it can be used to
7006 print addresses in debug_print(). Adjusted all the address printing in the
7007 debugging in store.c to use %p rather than %d.
7009 81. There was a concern that a line of code in smtp_in.c could overflow a
7010 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7011 initially expressed, the concern was not well-founded, because trailing
7012 spaces are removed early. However, if the trailing spaces were followed by
7013 a NULL, they did not get removed, so the overflow was possible. Two fixes
7016 (a) I re-wrote the offending code in a cleaner fashion.
7017 (b) If an incoming SMTP command contains a NULL character, it is rejected
7020 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7021 runs initgroups(), so that if the Exim user is in any additional groups,
7022 they will be used during message reception.
7028 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob