1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
40 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
41 Disable the session-cache too, which might reduce our load. Since we
42 currrectly use a new context for every connection, both as server and
43 client, there is no benefit for these.
44 GnuTLS appears to not support tickets server-side by default (we don't
45 call gnutls_session_ticket_enable_server()) but client side is enabled
46 by default on recent versions (3.1.3 +) unless the PFS priority string
49 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
50 <https://reproducible-builds.org/specs/source-date-epoch/>.
52 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
53 the check for any unsuccessful recipients did not notice the limit, and
54 erroneously found still-pending ones.
56 JH/08 Pipeline CHUNKING command and data together, on kernels that support
57 MSG_MORE. Only in-clear (not on TLS connections).
59 JH/09 Avoid using a temporary file during transport using dkim. Unless a
60 transport-filter is involved we can buffer the headers in memory for
61 creating the signature, and read the spool data file once for the
62 signature and again for transmission.
64 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
65 4.77 as the kernel support then wasn't solid, having issues in 64bit
66 mode. Now, it's been long enough. Add support for FreeBSD also.
68 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
69 case where the routing stage had gathered several addresses to send to
70 a host before calling the transport for the first, we previously failed
71 to close down TLS in the old transport process before passing the TCP
72 connection to the new process. The new one sent a STARTTLS command
73 which naturally failed, giving a failed delivery and bloating the retry
74 database. Investigation and fix prototype from Wolfgang Breyha.
76 JH/12 Fix check on SMTP command input synchronisation. Previously there were
77 false-negatives in the check that the sender had not preempted a response
78 or prompt from Exim (running as a server), due to that code's lack of
79 awareness of the SMTP input buffering.
81 PP/04 Add commandline_checks_require_admin option.
82 Exim drops privileges sanely, various checks such as -be aren't a
83 security problem, as long as you trust local users with access to their
84 own account. When invoked by services which pass untrusted data to
85 Exim, this might be an issue. Set this option in main configuration
86 AND make fixes to the calling application, such as using `--` to stop
89 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
90 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
91 responses to those, into a single TLS record each way (this usually means
92 a single packet). As a side issue, smtp_enforce_sync now works on TLS
95 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
96 affects you only if you're dancing at the edge of the param size limits.
97 If you are, and this message makes sense to you, then: raise the
98 configured limit or use OpenSSL 1.1. Nothing we can do for older
101 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
102 cmdline element to get the documented default one. Previously it was
105 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
106 are unsuable for non-message-specific reasons - eg. connection timeout,
107 banner-time rejection.
109 JH/16 Fix logging of delivery remote port, when specified by router, under
112 PP/06 Repair manualroute's ability to take options in any order, even if one
113 is the name of a transport.
116 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
118 JH/17 Change the list-building routines interface to use the expanding-string
119 triplet model, for better allocation and copying behaviour.
121 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
122 Previously it was constructed the first time a possibly-matching string
123 was met in the configuration file input during startup; now it is done
126 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
127 compatible one, to avoid the (poorly documented) possibility of a config
128 file in the working directory redirecting the DB files, possibly correpting
129 some existing file. CVE-2017-10140 assigned for BDB.
131 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
132 cache-hot. Previously, although the result was properly cached, the
133 initial verify call returned a defer.
135 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
136 the main verify for receipient in uncached-mode.
138 JH/22 Retire historical build files to an "unsupported" subdir. These are
139 defined as "ones for which we have no current evidence of testing".
141 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
142 if present. Previously it was ignored.
144 JH/24 Start using specified-initialisers in C structure init coding. This is
145 a C99 feature (it's 2017, so now considered safe).
147 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
148 if was a fixed-sized field and bitmask ops via macros; it is now more
151 PP/07 GitHub PR 56: Apply MariaDB build fix.
152 Patch provided by Jaroslav Škarvada.
154 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
155 during Coverity cleanups [4.87 JH/47]
156 Diagnosis and fix provided by Michael Fischer v. Mollard.
158 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
159 the right size to place the terminating semicolon on its own folded
160 line, the header hash was calculated to an incorrect value thanks to
161 the (relaxed) space the fold became.
163 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
164 and confused the parent.
170 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
171 than -2003 did; needs libidn2 in addition to libidn.
173 JH/02 The path option on a pipe transport is now expanded before use.
175 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
176 Patch provided by "Björn", documentation fix added too.
178 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
179 missing a wire-to-host endian conversion.
181 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
182 close after a BDAT command line could be taken as a following command,
183 giving a synch failure. Fix by only checking for synch immediately
184 before acknowledging the chunk.
186 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
187 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
188 macro. Patches provided by Josh Soref.
190 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
191 Previously we did not; the RFC seems ambiguous and VRFY is not listed
192 by IANA as a service extension. However, John Klensin suggests that we
195 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
196 the dkim code may be unix-mode line endings rather than smtp wire-format
197 CRLF, so prepend a CR to any bare LF.
199 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
200 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
202 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
203 the first were themselves being wrongly included in the feed into dkim
204 processing; with most chunk sizes in use this resulted in an incorrect
205 body hash calculated value.
207 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
208 DKIM signature block, for verification. Although advised against by
209 standards it is specifically not ruled illegal.
211 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
213 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
214 missing a body hash (the bh= tag).
216 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
217 It seems that HAProxy sends the Proxy Protocol information in clear and
218 only then does a TLS startup, so do the same.
220 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
221 TCP connections (such as for Spamd) unless the daemon successfully set
222 Fast Open mode on its listening sockets. This fixes breakage seen on
223 too-old kernels or those not configured for Fast Open, at the cost of
224 requiring both directions being enabled for TFO, and TFO never being used
225 by non-daemon-related Exim processes.
227 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
228 endings, at least on the first header line. Try to canonify any that get
229 past that check, despite the cost.
231 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
232 now limited to an arbitrary five deep, while parsing addresses with the
233 strip_excess_angle_brackets option enabled.
235 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
236 instead leave the unprompted TLS handshake in socket buffer for the
237 TLS library to consume.
239 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
241 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
243 JH/16 Drop variables when they go out of scope. Memory management drops a whole
244 region in one operation, for speed, and this leaves assigned pointers
245 dangling. Add checks run only under the testsuite which checks all
246 variables at a store-reset and panics on a dangling pointer; add code
247 explicitly nulling out all the variables discovered. Fixes one known
248 bug: a transport crash, where a dangling pointer for $sending_ip_address
249 originally assigned in a verify callout, is re-used.
251 PP/06 Drop '.' from @INC in various Perl scripts.
253 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
255 PP/08 Reduce a number of compilation warnings under clang; building with
256 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
257 should be warning-free.
259 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
261 HS/01 Fix portability problems introduced by PP/08 for platforms where
262 realloc(NULL) is not equivalent to malloc() [SunOS et al].
264 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
265 chunk. This allows us to accept broken chunked messages. We need a more
266 general solution here.
268 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
269 already-broken messages in the queue.
271 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
273 JH/19 Fix reference counting bug in routing-generated-address tracking.
279 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
280 supports it and a size is available (ie. the sending peer gave us one).
282 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
283 years of being deprecated). The replacements are the ACLs
284 acl_smtp_mime and acl_not_smtp_mime.
286 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
287 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
288 or even an in-clear connection were permitted. Now, if the host lookup
289 was dnssec and dane was requested then the host is only used if the
290 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
291 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
292 if one fails this test.
293 This means that a poorly-configured remote DNS will make it incommunicado;
294 but it protects against a DNS-interception attack on it.
296 JH/04 Bug 1810: make continued-use of an open smtp transport connection
297 non-noisy when a race steals the message being considered.
299 JH/05 If main configuration option tls_certificate is unset, generate a
300 self-signed certificate for inbound TLS connections.
302 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
303 in rewrites and routers.
305 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
306 and logged a warning sing 4.83; now they are a configuration file error.
308 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
309 (lacking @domain). Apply the same qualification processing as RCPT.
311 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
313 JH/10 Support ${sha256:} applied to a string (as well as the previous
316 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
317 a cutthrough deliver is pending, as we always want to make a connection.
318 This also avoids re-routing the message when later placing the cutthrough
319 connection after a verify cache hit.
320 Do not update it with the verify result either.
322 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
323 when routing results in more than one destination address.
325 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
326 signing (which inhibits the cutthrough capability). Previously only
327 the presence of an option was tested; now an expansion evaluating as
328 empty is permissible (obviously it should depend only on data available
329 when the cutthrough connection is made).
331 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
332 the relevant preceding SMTP command did not note the pipelining mode.
334 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
335 Previously they were not counted.
337 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
338 as one having no matching records. Previously we deferred the message
339 that needed the lookup.
341 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
342 distinguished as "(=".
344 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
345 for missing MX records. Previously it only worked for missing A records.
347 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
349 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
350 after the data-go-ahead and data-ack. Patch from Jason Betts.
352 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
353 even for a "none" policy. Patch from Tony Meyer.
355 JH/22 Fix continued use of a connection for further deliveries. If a port was
356 specified by a router, it must also match for the delivery to be
359 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
360 When one of the recipients of a message was unsuitable for the connection
361 (has no matching addresses), we lost track of needing to mark it
362 deferred. As a result mail would be lost.
364 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
366 JH/25 Decoding ACL controls is now done using a binary search; the source code
367 takes up less space and should be simpler to maintain. Merge the ACL
368 condition decode tables also, with similar effect.
370 JH/26 Fix problem with one_time used on a redirect router which returned the
371 parent address unchanged. A retry would see the parent address marked as
372 delivered, so not attempt the (identical) child. As a result mail would
375 JH/27 Fix a possible security hole, wherein a process operating with the Exim
376 UID can gain a root shell. Credit to http://www.halfdog.net/ for
377 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
380 JH/28 Enable {spool,log} filesystem space and inode checks as default.
381 Main config options check_{log,spool}_{inodes,space} are now
382 100 inodes, 10MB unless set otherwise in the configuration.
384 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
385 Previously it only applied to the main-section connection policy
388 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
390 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
391 by me. Added RFC7919 DH primes as an alternative.
393 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
394 are not in the system include path.
396 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
397 GnuTLS, when a session startup failed (eg because the client disconnected)
398 Exim did stdio operations after fclose. This was exposed by a recent
399 change which nulled out the file handle after the fclose.
401 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
402 signed directly by the cert-signing cert, rather than an intermediate
403 OCSP-signing cert. This is the model used by LetsEncrypt.
405 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
407 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
408 an incoming connection.
410 HS/02 Bug 1802: Do not half-close the connection after sending a request
413 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
414 fallback to "prime256v1".
416 JH/34 SECURITY: Use proper copy of DATA command in error message.
417 Could leak key material. Remotely exploitable. CVE-2016-9963.
423 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
424 and 3.4.4 - once the server is enabled to respond to an OCSP request
425 it does even when not requested, resulting in a stapling non-aware
426 client dropping the TLS connection.
428 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
429 support variable-length bit vectors. No functional change.
431 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
432 The I= interface field on outgoing lines is now after the H= remote
433 host field, same as incoming lines. There is a separate
434 outgoing_interface log selector which allows you to disable the
437 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
438 If not running log_selector +smtp_connection the mainlog would be held
439 open indefinitely after a "too many connections" event, including to a
440 deleted file after a log rotate. Leave the per net connection logging
441 leaving it open for efficiency as that will be quickly detected by the
442 check on the next write.
444 HS/01 Bug 1671: Fix post transport crash.
445 Processing the wait-<transport> messages could crash the delivery
446 process if the message IDs didn't exist for some reason. When
447 using 'split_spool_directory=yes' the construction of the spool
448 file name failed already, exposing the same netto behaviour.
450 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
451 mime_regex ACL conditions.
453 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
454 to DSN fail messages (bounces): remote IP, remote greeting, remote response
455 to HELO, local diagnostic string.
457 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
458 log line to debug. Even when configured with a tls authenticator many
459 client connections are expected to not authenticate in this way, so
460 an authenticate fail is not an error.
462 HS/02 Add the Exim version string to the process info. This way exiwhat
463 gives some more detail about the running daemon.
465 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
466 matter for fast-change records such as DNSBLs.
468 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
469 retry record, even if constant. There may be multiple transports with
470 different interface settings and the retry behaviour needs to be kept
473 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
475 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
477 JH/10 Bug 840: fix log_defer_output option of pipe transport
479 JH/11 Bug 830: use same host for all RCPTS of a message, even under
480 hosts_randomize. This matters a lot when combined with mua_wrapper.
482 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
483 ${quote_pgsql:<string>} operator.
485 JH/13 Bug 1708: avoid misaligned access in cached lookup.
487 JH/14 Change header file name for freeradius-client. Relevant if compiling
488 with Radius support; from the Gentoo tree and checked under Fedora.
490 JH/15 Bug 1712: Introduce $prdr_requested flag variable
492 JH/16 Bug 1714: Permit an empty string as expansion result for transport
493 option transport_filter, meaning no filtering.
495 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
497 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
498 defaults to "*" (all hosts). The variable is now available when not built
499 with TLS, default unset, mainly to enable keeping the testsuite sane.
500 If a server certificate is not supplied (via tls_certificate) an error is
501 logged, and clients will find TLS connections fail on startup. Presumably
502 they will retry in-clear.
503 Packagers of Exim are strongly encouraged to create a server certificate
504 at installation time.
506 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
507 with the $config_file variable.
509 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
510 in transport context, after the attempt, and per-recipient. The latter type
511 is per host attempted. The event data is the error message, and the errno
512 information encodes the lookup type (A vs. MX) used for the (first) host,
513 and the trailing two digits of the smtp 4xx response.
515 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
516 to write to mainlog (or rejectlog, paniclog) in the window between file
517 creation and permissions/ownership being changed. Particularly affects
518 installations where exicyclog is run as root, rather than exim user;
519 result is that the running daemon panics and dies.
521 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
523 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
524 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
525 "pri" and "weight". Note that the previous implicit priority given by the
526 list order is no longer honoured.
528 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
531 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
532 by defining SUPPORT_SOCKS.
534 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
535 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
536 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
537 variables are renamed to proxy_{local,external}_{address,port}.
539 JH/27 Move Internationalisation support from Experimental to mainline, enabled
540 for a build by defining SUPPORT_I18N
542 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
543 of the query string, and make ${quote_redis:} do that quoting.
545 JH/29 Move Events support from Experimental to mainline, enabled by default
546 and removable for a build by defining DISABLE_EVENT.
548 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
550 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
551 cached by the daemon.
553 JH/32 Move Redis support from Experimental to mainline, enabled for a build
554 by defining LOOKUP_REDIS. The libhiredis library is required.
556 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
557 keys are given for lookup.
559 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
560 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
561 only supported when built with TLS support. The PolarSSL SHA routines
562 are still used when the TLS library is too old for convenient support.
564 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
565 openssl_options), for security. OpenSSL forces this from version 1.1.0
566 server-side so match that on older versions.
568 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
569 allocation for $value could be released as the expansion processing
570 concluded, but leaving the global pointer active for it.
572 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
573 and to use the domains and local_parts ACL conditions.
575 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
576 incorrectly not doubled on cutthrough transmission, hence seen as a
577 body-termination at the receiving system - resulting in truncated mails.
578 Commonly the sender saw a TCP-level error, and retransmitted the message
579 via the normal store-and-forward channel. This could result in duplicates
580 received - but deduplicating mailstores were liable to retain only the
581 initial truncated version.
583 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
585 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
587 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
588 we're in there, support oversigning also; bug 1309.
590 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
592 HS/04 Add support for keep_environment and add_environment options.
594 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
595 either intentional arithmetic overflow during PRNG, or testing config-
598 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
599 delivery resulted in actual delivery. Cancel cutthrough before DATA
602 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
603 rejects a recipient: pass the reject to the originator.
605 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
606 Many were false-positives and ignorable, but it's worth fixing the
609 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
610 for the new environment-manipulation done at startup. Move the routines
611 from being local to tls.c to being global via the os.c file.
613 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
614 an extract embedded as result-arg for a map, the first arg for extract
615 is unavailable so we cannot tell if this is a numbered or keyed
616 extraction. Accept either.
622 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
625 JH/02 The smtp transport option "multi_domain" is now expanded.
627 JH/03 The smtp transport now requests PRDR by default, if the server offers
630 JH/04 Certificate name checking on server certificates, when exim is a client,
631 is now done by default. The transport option tls_verify_cert_hostnames
632 can be used to disable this per-host. The build option
633 EXPERIMENTAL_CERTNAMES is withdrawn.
635 JH/05 The value of the tls_verify_certificates smtp transport and main options
636 default to the word "system" to access the system default CA bundle.
637 For GnuTLS, only version 3.0.20 or later.
639 JH/06 Verification of the server certificate for a TLS connection is now tried
640 (but not required) by default. The verification status is now logged by
641 default, for both outbound TLS and client-certificate supplying inbound
644 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
647 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
648 Status Notification (bounce) messages are now MIME format per RFC 3464.
649 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
650 under the control of the dsn_advertise_hosts option, and routers may
651 have a dsn_lasthop option.
653 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
654 default, modifiable by a malware= option. The list separator for
655 the options can now be changed in the usual way. Bug 68.
657 JH/10 The smtp_receive_timeout main option is now expanded before use.
659 JH/11 The incoming_interface log option now also enables logging of the
660 local interface on delivery outgoing connections.
662 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
663 if the interface and destination host and port all match.
665 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
668 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
669 Patch from Andrew Lewis.
671 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
672 now supports optional time-restrictions, weighting, and priority
673 modifiers per server. Patch originally by <rommer@active.by>.
675 JH/16 The spamd_address main option now supports a mixed list of local
676 and remote servers. Remote servers can be IPv6 addresses, and
677 specify a port-range.
679 JH/17 Bug 68: The spamd_address main option now supports an optional
680 timeout value per server.
682 JH/18 Bug 1581: Router and transport options headers_add/remove can
683 now have the list separator specified.
685 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
688 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
691 JH/21 Support for the A6 type of dns record is withdrawn.
693 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
694 rather than the verbs used.
696 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
697 from 255 to 1024 chars.
699 JH/24 Verification callouts now attempt to use TLS by default.
701 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
702 are generic router options now. The defaults didn't change.
704 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
705 Original patch from Alexander Shikoff, worked over by JH.
707 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
708 files if ZCAT_COMMAND is not executable.
710 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
712 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
714 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
715 Normally benign, it bites when the pair was led to by a CNAME;
716 modern usage is to not canonicalize the domain to a CNAME target
717 (and we were inconsistent anyway for A-only vs AAAA+A).
719 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
721 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
722 when evaluating $sender_host_dnssec.
724 JH/31 Check the HELO verification lookup for DNSSEC, adding new
725 $sender_helo_dnssec variable.
727 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
729 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
731 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
733 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
734 documented as working, but never had. Support all but $spam_report.
736 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
737 added for tls authenticator.
739 HS/03 Add perl_taintmode main config option
745 TL/01 When running the test suite, the README says that variables such as
746 no_msglog_check are global and can be placed anywhere in a specific
747 test's script, however it was observed that placement needed to be near
748 the beginning for it to behave that way. Changed the runtest perl
749 script to read through the entire script once to detect and set these
750 variables, reset to the beginning of the script, and then run through
751 the script parsing/test process like normal.
753 TL/02 The BSD's have an arc4random API. One of the functions to induce
754 adding randomness was arc4random_stir(), but it has been removed in
755 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
756 function when detected.
758 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
759 cause callback expansion.
761 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
762 syntax errors in an expansion can be treated as a string instead of
763 logging or causing an error, due to the internal use of bool_lax
764 instead of bool when processing it.
766 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
767 server certificates when making smtp deliveries.
769 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
771 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
773 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
775 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
776 Merged patch from Sebastian Wiedenroth.
778 JH/05 Fix results-pipe from transport process. Several recipients, combined
779 with certificate use, exposed issues where response data items split
780 over buffer boundaries were not parsed properly. This eventually
781 resulted in duplicates being sent. This issue only became common enough
782 to notice due to the introduction of connection certificate information,
783 the item size being so much larger. Found and fixed by Wolfgang Breyha.
785 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
786 size buffer was used, resulting in syntax errors when an expansion
789 JH/07 Add support for directories of certificates when compiled with a GnuTLS
790 version 3.3.6 or later.
792 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
793 is EXPERIMENTAL_EVENT, the main-configuration and transport options
794 both become "event_action", the variables become $event_name, $event_data
795 and $event_defer_errno. There is a new variable $verify_mode, usable in
796 routers, transports and related events. The tls:cert event is now also
797 raised for inbound connections, if the main configuration event_action
800 TL/06 In test suite, disable OCSP for old versions of openssl which contained
801 early OCSP support, but no stapling (appears to be less than 1.0.0).
803 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
804 server certificate names available under the smtp transport option
805 "tls_verify_cert_hostname" now do not permit multi-component wildcard
808 JH/10 Time-related extraction expansions from certificates now use the main
809 option "timezone" setting for output formatting, and are consistent
810 between OpenSSL and GnuTLS compilations. Bug 1541.
812 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
813 encoded parameter in the incoming message. Bug 1558.
815 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
816 include certificate info, eximon was claiming there were spoolfile
819 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
821 JH/14 Log delivery-related information more consistently, using the sequence
822 "H=<name> [<ip>]" wherever possible.
824 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
825 are problematic for Debian distribution, omit them from the release
828 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
830 JH/16 Fix string representation of time values on 64bit time_t architectures.
833 JH/17 Fix a null-indirection in certextract expansions when a nondefault
834 output list separator was used.
839 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
840 checkers that were complaining about end of non-void function with no
843 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
844 This was a regression introduced in 4.83 by another bugfix.
846 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
848 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
849 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
855 TF/01 Correctly close the server side of TLS when forking for delivery.
857 When a message was received over SMTP with TLS, Exim failed to clear up
858 the incoming connection properly after forking off the child process to
859 deliver the message. In some situations the subsequent outgoing
860 delivery connection happened to have the same fd number as the incoming
861 connection previously had. Exim would try to use TLS and fail, logging
862 a "Bad file descriptor" error.
864 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
865 utilities have not been installed.
867 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
868 temporary space as the ACL may create new global variables.
870 TL/01 LDAP support uses per connection or global context settings, depending
871 upon the detected version of the libraries at build time.
873 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
874 to extract and use the src ip:port in logging and expansions as if it
875 were a direct connection from the outside internet. PPv2 support was
876 updated based on HAProxy spec change in May 2014.
878 JH/02 Add ${listextract {number}{list}{success}{fail}}.
880 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
881 Properly escape header and check for NULL return.
883 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
886 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
888 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
889 characters in header names, implemented as a verify condition.
890 Contributed by Michael Fischer v. Mollard.
892 TL/05 Rename SPF condition results err_perm and err_temp to standardized
893 results permerror and temperror. Previous values are deprecated but
894 still accepted. In a future release, err_perm and err_temp will be
895 completely removed, which will be a backward incompatibility if the
896 ACL tests for either of these two old results. Patch contributed by
897 user bes-internal on the mailing list.
899 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
901 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
902 selectors, in both main and reject logs.
904 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
907 JH/07 Add malware type "sock" for talking to simple daemon.
909 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
911 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
912 routers/transports under cutthrough routing.
914 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
915 numbers. Touch up "bool" conditional to keep the same definition.
917 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
919 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
921 JH/12 Expand items in router/transport headers_add or headers_remove lists
922 individually rather than the list as a whole. Bug 1452.
924 Required for reasonable handling of multiple headers_ options when
925 they may be empty; requires that headers_remove items with embedded
926 colons must have them doubled (or the list-separator changed).
928 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
929 view the policy declared in the DMARC record. Currently, $dmarc_status
930 is a combined value of both the record presence and the result of the
933 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
935 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
936 dnslookup router and the smtp transport (applying to the forward
939 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
940 of ldap servers used for a specific lookup. Patch provided by Heiko
943 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
944 New variable $lookup_dnssec_authenticated for observability.
946 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
947 Patch submitted by Lars Timman.
949 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
951 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
952 Requires trusted mode and valid format message id, aborts otherwise.
953 Patch contributed by Heiko Schlichting.
955 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
956 certextract with support for various fields. Bug 1358.
958 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
959 is requested by default, modifiable by smtp transport option
962 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
963 operate on certificate variables to give certificate fingerprints
964 Also new ${sha256:cert_variable}.
966 JH/23 The PRDR feature is moved from being Experimental into the mainline.
968 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
969 Christian Aistleitner.
971 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
973 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
974 file. Patch from Wolfgang Breyha.
976 JH/25 Expand the coverage of the delivery $host and $host_address to
977 client authenticators run in verify callout. Bug 1476.
979 JH/26 Port service names are now accepted for tls_on_connect_ports, to
980 align with daemon_smtp_ports. Bug 72.
982 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
983 support and error reporting did not work properly.
985 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
986 and is readable. Patch from Andrew Colin Kissa.
988 TL/14 Enhance documentation of ${run expansion and how it parses the
989 commandline after expansion, particularly in the case when an
990 unquoted variable expansion results in an empty value.
992 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
994 PP/02 Fix internal collision of T_APL on systems which support RFC3123
995 by renaming away from it. Addresses GH issue 15, reported by
998 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1000 TL/15 SECURITY: prevent double expansion in math comparison functions
1001 (can expand unsanitized data). Not remotely exploitable.
1008 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1010 PP/02 Make -n do something, by making it not do something.
1011 When combined with -bP, the name of an option is not output.
1013 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1016 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1017 $sender_host_name and config options to manage this, and basic check
1020 PP/05 DSCP support for outbound connections and control modifier for inbound.
1022 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1023 (Only plugin which currently uses this is kerberos4, which nobody should
1024 be using, but we should make it available and other future plugins might
1025 conceivably use it, even though it would break NAT; stuff *should* be
1026 using channel bindings instead).
1028 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1029 name; added for Sendmail compatibility; requires admin caller.
1030 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1031 just ignore it); requires trusted caller.
1032 Also parse but ignore: -Ac -Am -X<logfile>
1035 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1037 TL/02 Add +smtp_confirmation as a default logging option.
1039 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1040 Patch by Magnus Holmgren from 2007-02-20.
1042 TL/04 Bugzilla 1281 - Spec typo.
1043 Bugzilla 1283 - Spec typo.
1044 Bugzilla 1290 - Spec grammar fixes.
1046 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1048 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1050 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1051 mailing list by by Dmitry Isaikin.
1053 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1055 JH/02 Support "G" suffix to numbers in ${if comparisons.
1057 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1059 NM/01 Bugzilla 1197 - Spec typo
1060 Bugzilla 1196 - Spec examples corrections
1062 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1064 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1065 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1068 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1069 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1071 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1072 "acl {{name}{arg}...}", and optional args on acl condition
1075 JH/05 Permit multiple router/transport headers_add/remove lines.
1077 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1079 JH/07 Avoid using a waiting database for a single-message-only transport.
1080 Performance patch from Paul Fisher. Bugzilla 1262.
1082 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1085 JH/09 Add $headers_added variable, with content from use of ACL modifier
1086 add_header (but not yet added to the message). Bugzilla 199.
1088 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1089 Pulled from Bugzilla 817 by Wolfgang Breyha.
1091 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1093 (nb: this is the same fix as in Exim 4.80.1)
1095 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1098 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1101 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1103 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1104 advertise SMTP AUTH mechanism to us, instead of a generic
1105 protocol violation error. Also, make Exim more robust to bad
1106 data from the Dovecot auth socket.
1108 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1110 When a queue runner is handling a message, Exim first routes the
1111 recipient addresses, during which it prunes them based on the retry
1112 hints database. After that it attempts to deliver the message to
1113 any remaining recipients. It then updates the hints database using
1116 So if a recipient address works intermittently, it can get repeatedly
1117 deferred at routing time. The retry hints record remains fresh so the
1118 address never reaches the final cutoff time.
1120 This is a fairly common occurrence when a user is bumping up against
1121 their storage quota. Exim had some logic in its local delivery code
1122 to deal with this. However it did not apply to per-recipient defers
1123 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1125 This change adds a proper retry rule check during routing so that the
1126 final cutoff time is checked against the message's age. We only do
1127 this check if there is an address retry record and there is not a
1128 domain retry record; this implies that previous attempts to handle
1129 the address had the retry_use_local_parts option turned on. We use
1130 this as an approximation for the destination being like a local
1131 delivery, as in LMTP.
1133 I suspect this new check makes the old local delivery cutoff check
1134 redundant, but I have not verified this so I left the code in place.
1136 TF/02 Correct gecos expansion when From: is a prefix of the username.
1138 Test 0254 submits a message to Exim with the header
1142 When I ran the test suite under the user fanf2, Exim expanded
1143 the header to contain my full name, whereas it should have added
1144 a Resent-Sender: header. It erroneously treats any prefix of the
1145 username as equal to the username.
1147 This change corrects that bug.
1149 GF/01 DCC debug and logging tidyup
1150 Error conditions log to paniclog rather than rejectlog.
1151 Debug lines prefixed by "DCC: " to remove any ambiguity.
1153 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1155 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1156 Bug spotted by Jeremy Harris; was flawed since initial commit.
1157 Would have resulted in OCSP responses post-SNI triggering an Exim
1158 NULL dereference and crash.
1160 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1162 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1163 Bug detection, analysis and fix by Samuel Thibault.
1164 Bugzilla 1331, Debian bug #698092.
1166 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1168 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1169 Server implementation by Todd Lyons, client by JH.
1170 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1171 config variable "prdr_enable" controls whether the server
1172 advertises the facility. If the client requests PRDR a new
1173 acl_data_smtp_prdr ACL is called once for each recipient, after
1174 the body content is received and before the acl_smtp_data ACL.
1175 The client is controlled by both of: a hosts_try_prdr option
1176 on the smtp transport, and the server advertisement.
1177 Default client logging of deliveries and rejections involving
1178 PRDR are flagged with the string "PRDR".
1180 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1181 fclose(). Diagnosis by Todd Lyons.
1183 PP/17 Update configure.default to handle IPv6 localhost better.
1184 Patch by Alain Williams (plus minor tweaks).
1187 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1188 This is now consistent with GnuTLS, and is now documented: the
1189 previous undocumented portable approach to treating the option as
1190 unset was to force an expansion failure. That still works, and
1191 an empty string is now equivalent.
1193 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1194 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1195 not performing validation itself.
1197 PP/20 Added force_command boolean option to pipe transport.
1198 Patch from Nick Koston, of cPanel Inc.
1200 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1203 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1205 PP/21 Fix eximon continuous updating with timestamped log-files.
1206 Broken in a format-string cleanup in 4.80, missed when I repaired the
1207 other false fix of the same issue.
1208 Report and fix from Heiko Schlichting.
1211 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1212 Report from Prashanth Katuri.
1214 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1215 It's SecureTransport, so affects any MacOS clients which use the
1216 system-integrated TLS libraries, including email clients.
1218 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1219 using a MIME ACL for non-SMTP local injection.
1220 Report and assistance in diagnosis by Warren Baker.
1222 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1224 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1226 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1227 last id that failed so it may be referenced in subsequent ACL's.
1229 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1232 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1233 ldap library initialization, allowing self-signed CA's to be
1234 used. Also properly sets require_cert option later in code by
1235 using NULL (global ldap config) instead of ldap handle (per
1236 session). Bug diagnosis and testing by alxgomz.
1238 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1239 the src/util/ subdirectory.
1241 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1242 renamed to Transport Post Delivery Action by Jeremy Harris, as
1245 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1246 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1247 redis_servers = needs to be configured which will be used by the redis
1248 lookup. Patch from Warren Baker, of The Packet Hub.
1250 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1252 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1253 hostname or reverse DNS when processing a host list. Used suggestions
1254 from multiple comments on this bug.
1256 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1258 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1259 Missed a few lines, added it to make the runtest require no keyboard
1262 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1263 contains upper case chars. Make router use caseful_local_part.
1265 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1266 support when GnuTLS has been built with p11-kit.
1272 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1274 This, or similar/improved, will also be change PP/11 of 4.82.
1280 PP/01 Handle short writes when writing local log-files.
1281 In practice, only affects FreeBSD (8 onwards).
1282 Bugzilla 1053, with thanks to Dmitry Isaikin.
1284 NM/01 Bugzilla 949 - Documentation tweak
1286 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1289 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1291 PP/02 Implemented gsasl authenticator.
1293 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1295 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1296 `pkg-config foo` for cflags/libs.
1298 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1299 with rest of GSASL and with heimdal_gssapi.
1301 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1302 `pkg-config foo` for cflags/libs for the TLS implementation.
1304 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1305 properties get this fed in as external SSF. A number of robustness
1306 and debugging improvements to the cyrus_sasl authenticator.
1308 PP/08 cyrus_sasl server now expands the server_realm option.
1310 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1311 Patch by Jeremy Harris.
1313 PP/10 Added dbmjz lookup type.
1315 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1317 PP/12 MAIL args handles TAB as well as SP, for better interop with
1318 non-compliant senders.
1319 Analysis and variant patch by Todd Lyons.
1321 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1322 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1323 Patch from Dirk Mueller <dmueller@suse.com>
1325 PP/13 tls_peerdn now print-escaped for spool files.
1326 Observed some $tls_peerdn in wild which contained \n, which resulted
1327 in spool file corruption.
1329 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1330 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1331 or write after TLS renegotiation, which otherwise led to messages
1334 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1335 as a tracking header (ie: a signed header comes before the signature).
1336 Patch from Wolfgang Breyha.
1338 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1339 comma-sep list; embedded commas doubled.
1341 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1343 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1345 Report and patch from Dmitry Banschikov.
1347 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1348 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1349 failures. We appear to now support TLS1.1+ with Exim.
1351 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1352 lets Exim select keys and certificates based upon TLS SNI from client.
1353 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1354 before an outbound SMTP session. New log_selector, +tls_sni.
1356 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1357 NULL dereference. Report and patch from Alun Jones.
1359 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1360 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1361 Not seeing resolver debug output on NetBSD, but suspect this is a
1362 resolver implementation change.
1364 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1365 Left warnings. Added "eximon gdb" invocation mode.
1367 PP/21 Defaulting "accept_8bitmime" to true, not false.
1369 PP/22 Added -bw for inetd wait mode support.
1371 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1372 locate the relevant includes and libraries. Made this the default.
1374 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1375 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1377 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1378 This may cause build issues on older platforms.
1380 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1381 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1382 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1383 Added SNI support via GnuTLS too.
1384 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1386 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1388 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1389 Applied second patch from Janne, implementing suggestion to default
1390 multiple-strings-in-record handling to match SPF spec.
1392 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1394 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1395 read-only, out of scope).
1396 Patch from Wolfgang Breyha, report from Stuart Northfield.
1398 PP/29 Fix three issues highlighted by clang analyser static analysis.
1399 Only crash-plausible issue would require the Cambridge-specific
1400 iplookup router and a misconfiguration.
1401 Report from Marcin Mirosław.
1403 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1405 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1406 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1407 As part of this, removing so much warning spew let me fix some minor
1408 real issues in debug logging.
1410 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1411 assignment on my part. Fixed.
1413 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1414 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1415 Janne Snabb (who went above and beyond: thank you).
1417 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1418 string otherwise requires a connection and a bunch more work and it's
1419 relatively easy to get wrong. Should also expose TLS library linkage
1422 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1423 64-bit ${eval} (JH/03).
1425 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1426 GNU libc to support some of the 64-bit stuff, should not lead to
1427 conflicts. Defined before os.h is pulled in, so if a given platform
1428 needs to override this, it can.
1430 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1431 protection layer was required, which is not implemented.
1432 Bugzilla 1254, patch from Wolfgang Breyha.
1434 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1435 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1436 tls_dhparam take prime identifiers. Also unbreak combination of
1437 OpenSSL+DH_params+TLSSNI.
1439 PP/39 Disable SSLv2 by default in OpenSSL support.
1445 PP/01 Solaris build fix for Oracle's LDAP libraries.
1446 Bugzilla 1109, patch from Stephen Usher.
1448 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1450 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1453 TF/02 Fix a couple more cases where we did not log the error message
1454 when unlink() failed. See also change 4.74-TF/03.
1456 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1457 lock up or crash if it happened to be inside a call to libc when it
1458 got a SIGUSR1 from exiwhat.
1460 The SIGUSR1 handler appends the current process status to the process
1461 log which is later printed by exiwhat. It used to use the general
1462 purpose logging code to do this, but several functions it calls are
1463 not safe for signals.
1465 The new output code in the SIGUSR1 handler is specific to the process
1466 log, and simple enough that it's easy to inspect for signal safety.
1467 Removing some special cases also simplifies the general logging code.
1468 Removing the spurious timestamps from the process log simplifies
1471 TF/04 Improved ratelimit ACL condition.
1473 The /noupdate option has been deprecated in favour of /readonly which
1474 has clearer semantics. The /leaky, /strict, and /readonly update modes
1475 are mutually exclusive. The update mode is no longer included in the
1476 database key; it just determines when the database is updated. (This
1477 means that when you upgrade Exim will forget old rate measurements.)
1479 Exim now checks that the per_* options are used with an update mode that
1480 makes sense for the current ACL. For example, when Exim is processing a
1481 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1482 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1483 must specify per_mail/readonly. If you omit the update mode it defaults to
1484 /leaky where that makes sense (as before) or /readonly where required.
1486 The /noupdate option is now undocumented but still supported for
1487 backwards compatibility. It is equivalent to /readonly except that in
1488 ACLs where /readonly is required you may specify /leaky/noupdate or
1489 /strict/noupdate which are treated the same as /readonly.
1491 A useful new feature is the /count= option. This is a generalization
1492 of the per_byte option, so that you can measure the throughput of other
1493 aggregate values. For example, the per_byte option is now equivalent
1494 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1496 The per_rcpt option has been generalized using the /count= mechanism
1497 (though it's more complicated than the per_byte equivalence). When it is
1498 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1499 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1500 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1501 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1502 Note that using per_rcpt with a non-readonly update mode in more than
1503 one ACL will cause the recipients to be double-counted. (The per_mail
1504 and per_byte options don't have this problem.)
1506 The handling of very low rates has changed slightly. If the computed rate
1507 is less than the event's count (usually one) then this event is the first
1508 after a long gap. In this case the rate is set to the same as this event's
1509 count, so that the first message of a spam run is counted properly.
1511 The major new feature is a mechanism for counting the rate of unique
1512 events. The new per_addr option counts the number of different
1513 recipients that someone has sent messages to in the last time period. It
1514 behaves like per_rcpt if all the recipient addresses are different, but
1515 duplicate recipient addresses do not increase the measured rate. Like
1516 the /count= option this is a general mechanism, so the per_addr option
1517 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1518 example, measure the rate that a client uses different sender addresses
1519 with the options per_mail/unique=$sender_address. There are further
1520 details in the main documentation.
1522 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1524 TF/06 Removed a few PCRE remnants.
1526 TF/07 Automatically extract Exim's version number from tags in the git
1527 repository when doing development or release builds.
1529 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1530 Bugzilla 879. Patch from Paul Fisher.
1532 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1533 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1536 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1538 PP/05 Variable $av_failed, true if the AV scanner deferred.
1539 Bugzilla 1078. Patch from John Horne.
1541 PP/06 Stop make process more reliably on build failure.
1542 Bugzilla 1087. Patch from Heiko Schlittermann.
1544 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1545 Bugzilla 1089. Patch from Heiko Schlittermann.
1547 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1548 Bugzilla 1131. Patch from Holger Weiß.
1550 PP/09 Handle IPv6 addresses with SPF.
1551 Bugzilla 860. Patch from Wolfgang Breyha.
1553 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1555 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1558 PP/11 match_* no longer expand right-hand-side by default.
1559 New compile-time build option, EXPAND_LISTMATCH_RHS.
1560 New expansion conditions, "inlist", "inlisti".
1562 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1564 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1566 PP/14 fix log_write() format string regression from TF/03.
1567 Bugzilla 1152. Patch from Dmitry Isaikin.
1573 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1575 PP/02 Harmonised TLS library version reporting; only show if debugging.
1576 Layout now matches that introduced for other libraries in 4.74 PP/03.
1578 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1580 PP/04 New "dns_use_edns0" global option.
1582 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1585 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1586 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1588 TK/01 Updated PolarSSL code to 0.14.2.
1589 Bugzilla 1097. Patch from Andreas Metzler.
1591 PP/07 Catch divide-by-zero in ${eval:...}.
1592 Fixes bugzilla 1102.
1594 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1597 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1598 format-string attack -- SECURITY: remote arbitrary code execution.
1600 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1601 time unintentionally subject to list matching rules, letting the header
1602 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1603 arbitrary string expansion). This allowed for information disclosure.
1605 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1606 INT_MIN/-1 -- value coerced to INT_MAX.
1612 NM/01 Workaround for PCRE version dependency in version reporting
1615 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1616 This fixes portability to compilers other than gcc, notably
1617 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1619 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1620 makefiles for portability to HP-UX and POSIX correctness.
1622 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1623 Also via indented variable definition in the Makefile.
1624 (Debugging by Oliver Heesakkers).
1626 PP/02 Restore caching of spamd results with expanded spamd_address.
1627 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1629 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1630 Improves build reliability. Fix from: Frank Elsner
1632 NM/02 Fix wide character breakage in the rfc2047 coding
1633 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1635 NM/03 Allow underscore in dnslist lookups
1636 Fixes bug 1026. Patch from Graeme Fowler
1638 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1639 Code patches from Adam Ciarcinski of NetBSD.
1641 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1644 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1645 is logged, to avoid truncation. Patch from John Horne.
1647 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1648 Patch from Jakob Hirsch.
1650 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1651 SQL string expansion failure details.
1652 Patch from Andrey Oktyabrski.
1654 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1655 Patch from Simon Arlott.
1657 PP/09 New lookups functionality failed to compile on old gcc which rejects
1658 extern declarations in function scope.
1659 Patch from Oliver Fleischmann
1661 PP/10 Use sig_atomic_t for flags set from signal handlers.
1662 Check getgroups() return and improve debugging.
1663 Fixed developed for diagnosis in bug 927 (which turned out to be
1666 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1667 Patch from Mark Zealey.
1669 PP/12 Bugzilla 1056: Improved spamd server selection.
1670 Patch from Mark Zealey.
1672 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1673 Based on patch from Heiko Schlittermann.
1675 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1676 Patch from Uwe Doering, sign-off by Michael Haardt.
1678 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1679 variable declaration deep within a block. Bug and patch from
1682 PP/15 lookups-Makefile IRIX compatibility coercion.
1684 PP/16 Make DISABLE_DKIM build knob functional.
1686 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1687 Patch by Simon Arlott
1689 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1690 variable argument macros. Our copy now differs from upstream.
1696 TF/01 Failure to get a lock on a hints database can have serious
1697 consequences so log it to the panic log.
1699 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1700 controlled using the smtp_confirmation log selector.
1702 TF/03 Include the error message when we fail to unlink a spool file.
1704 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1705 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1706 for maintaining out-of-tree patches for some time.
1708 PP/01 Bugzilla 139: Documentation and portability issues.
1709 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1710 Handle per-OS dynamic-module compilation flags.
1712 PP/02 Let /dev/null have normal permissions.
1713 The 4.73 fixes were a little too stringent and complained about the
1714 permissions on /dev/null. Exempt it from some checks.
1715 Reported by Andreas M. Kirchwitz.
1717 PP/03 Report version information for many libraries, including
1718 Exim version information for dynamically loaded libraries. Created
1719 version.h, now support a version extension string for distributors
1720 who patch heavily. Dynamic module ABI change.
1722 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1723 privilege escalation vulnerability whereby the Exim run-time user
1724 can cause root to append content of the attacker's choosing to
1727 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1730 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1731 If dropping privileges for untrusted macros, we disabled normal logging
1732 on the basis that it would fail; for the Exim run-time user, this is not
1733 the case, and it resulted in successful deliveries going unlogged.
1734 Fixed. Reported by Andreas Metzler.
1740 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1741 only prepend for the Resent-* case. Fixes regression introduced in
1742 Exim 4.70 by NM/22 for Bugzilla 607.
1744 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1745 increasing numbers of administrators be bitten by this.
1747 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1749 PP/03 Bugzilla 994: added openssl_options main configuration option.
1751 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1753 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1755 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1757 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1758 without a peer certificate, leading to a segfault because of an
1759 assumption that peers always have certificates. Be a little more
1760 paranoid. Problem reported by Martin Tscholak.
1762 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1763 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1764 NB: ClamAV planning to remove STREAM in "middle of 2010".
1765 CL also introduces -bmalware, various -d+acl logging additions and
1766 more caution in buffer sizes.
1768 PP/09 Implemented reverse_ip expansion operator.
1770 PP/10 Bugzilla 937: provide a "debug" ACL control.
1772 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1774 PP/12 Bugzilla 973: Implement --version.
1776 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1778 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1780 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1782 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1783 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1784 ignore trailing whitespace.
1786 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1788 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1791 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1792 Notification from Dr Andrew Aitchison.
1794 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1795 ExtendedDetectionInfo response format.
1796 Notification from John Horne.
1798 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1801 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1802 XSL and documented dependency on system catalogs, with examples of how
1805 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1808 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1809 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1810 configuration file which is writeable by the Exim user or group.
1812 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1813 of configuration files to cover files specified with the -C option if
1814 they are going to be used with root privileges, not just the default
1817 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1818 option (effectively making it always true).
1820 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1821 files to be used while preserving root privileges.
1823 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1824 that rogue child processes cannot use them.
1826 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1827 run-time user, instead of root.
1829 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1830 Exim run-time user without dropping privileges.
1832 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1833 result string, instead of calling string_vformat() twice with the same
1836 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1837 for other users. Others should always drop root privileges if they use
1838 -C on the command line, even for a whitelisted configure file.
1840 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1842 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1848 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1849 $data_path, and $header_path variables; fixed documentation bugs and
1852 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1853 exipick to access non-standard spools, including the "frozen" queue
1856 NM/01 Bugzilla 965: Support mysql stored procedures.
1857 Patch from Alain Williams
1859 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1861 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1862 Patch from Andreas Metzler
1864 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1865 Patch from Kirill Miazine
1867 NM/05 Bugzilla 671: Added umask to procmail example.
1869 JJ/03 installed exipick 20100323.0, fixing doc bug
1871 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1872 directory. Notification and patch from Dan Rosenberg.
1874 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1876 TK/02 Improve log output when DKIM signing operation fails.
1878 MH/01 Treat the transport option dkim_domain as a colon separated
1879 list, not as a single string, and sign the message with each element,
1880 omitting multiple occurences of the same signer.
1882 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1883 Bugzilla 985, 986. Patch by Simon Arlott
1885 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1886 Patch by Simon Arlott
1888 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1889 MBX locking. Notification from Dan Rosenberg.
1895 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1897 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1899 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1901 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1903 NM/04 Bugzilla 915: Flag broken perl installation during build.
1909 TK/01 Added patch by Johannes Berg that expands the main option
1910 "spamd_address" if it starts with a dollar sign.
1912 TK/02 Write list of recipients to X-Envelope-Sender header when building
1913 the mbox-format spool file for content scanning (suggested by Jakob
1916 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1917 (http://www.dcc-servers.net/) support via dccifd. Activated by
1918 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1920 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1921 by Mark Daniel Reidel <mr@df.eu>.
1923 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1924 When building exim an external PCRE library is now needed -
1925 PCRE is a system library on the majority of modern systems.
1926 See entry on PCRE_LIBS in EDITME file.
1928 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1929 conversation. Added nologin parameter to request.
1930 Patch contributed by Kirill Miazine.
1932 TF/01 Do not log submission mode rewrites if they do not change the address.
1934 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1936 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1937 log files in place. Contributed by Roberto Lima.
1939 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1941 TF/03 Bugzilla 615: When checking the local_parts router precondition
1942 after a local_part_suffix or local_part_prefix option, Exim now
1943 does not use the address's named list lookup cache, since this
1944 contains cached lookups for the whole local part.
1946 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1947 Robert Millan. Documentation is in experimental-spec.txt.
1949 TF/04 Bugzilla 668: Fix parallel build (make -j).
1951 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1953 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1954 Patch provided by Jan Srzednicki.
1956 TF/05 Leading white space used to be stripped from $spam_report which
1957 wrecked the formatting. Now it is preserved.
1959 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1960 that they are available at delivery time.
1962 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1964 TF/08 TLS error reporting now respects the incoming_interface and
1965 incoming_port log selectors.
1967 TF/09 Produce a more useful error message if an SMTP transport's hosts
1968 setting expands to an empty string.
1970 NM/06 Bugzilla 744: EXPN did not work under TLS.
1971 Patch provided by Phil Pennock.
1973 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1974 Patch provided by Richard Godbee.
1976 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1977 acl_smtp_notquit, added index entry.
1979 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1980 Patch provided by Eugene Bujak.
1982 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1983 accept(). Patch provided by Maxim Dounin.
1985 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1986 Patch provided by Phil Pennock.
1988 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1990 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1991 Patch provided by Brad "anomie" Jorsch.
1993 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1994 Patch provided by Dean Brooks.
1996 TK/05 Add native DKIM support (does not depend on external libraries).
1998 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1999 Patch provided by Graeme Fowler.
2001 NM/16 Bugzilla 851: Documentation example syntax fix.
2003 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2005 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2008 NM/19 Bugzilla 745: TLS version reporting.
2009 Patch provided by Phil Pennock.
2011 NM/20 Bugzilla 167: bool: condition support.
2012 Patch provided by Phil Pennock.
2014 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2015 clients. Patch provided by Phil Pennock.
2017 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2018 Patch provided by Brad "anomie" Jorsch.
2020 NM/23 Bugzilla 687: Fix misparses in eximstats.
2021 Patch provided by Heiko Schlittermann.
2023 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2024 Patch provided by Heiko Schlittermann.
2026 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2027 plus update to original patch.
2029 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2031 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2032 Patch provided by David Brownlee.
2034 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2036 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2038 NM/30 Bugzilla 888: TLS documentation bugfixes.
2040 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2042 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2043 Unlike the original bugzilla I have changed all shell scripts in src tree.
2045 NM/33 Bugzilla 898: Transport filter timeout fix.
2046 Patch by Todd Rinaldo.
2048 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2049 Patch by Serge Demonchaux.
2051 NM/35 Bugzilla 39: Base64 decode bug fixes.
2052 Patch by Jakob Hirsch.
2054 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2056 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2058 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2060 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2066 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2067 ALT-N's libdkim that I have put here:
2068 http://duncanthrax.net/exim-experimental/
2070 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2071 were called 'true' and it seems that C99 defines that as a reserved
2072 keyword to be used with 'bool' variable types. That means you could
2073 not include C99-style headers which use bools without triggering
2074 build errors in sieve.c.
2076 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2077 as mailq or other aliases. Changed the --help handling significantly
2078 to do whats expected. exim_usage() emits usage/help information.
2080 SC/01 Added the -bylocaldomain option to eximstats.
2082 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2084 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2086 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2092 PH/01 Another patch from the Sieve maintainer.
2094 PH/02 When an IPv6 address is converted to a string for single-key lookup
2095 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2096 dots are used instead of colons so that keys in lsearch files need not
2097 contain colons. This was done some time before quoting was made available
2098 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2099 (notated using the quote facility) so as to distinguish them from IPv4
2100 keys. This meant that lookups for IP addresses in host lists did not work
2101 for iplsearch lookups.
2103 This has been fixed by arranging for IPv6 addresses to be expressed with
2104 colons if the lookup type is iplsearch. This is not incompatible, because
2105 previously such lookups could never work.
2107 The situation is now rather anomalous, since one *can* have colons in
2108 ordinary lsearch keys. However, making the change in all cases is
2109 incompatible and would probably break a number of configurations.
2111 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2114 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2115 conversion specification without a maximum field width, thereby enabling
2116 a rogue spamd server to cause a buffer overflow. While nobody in their
2117 right mind would setup Exim to query an untrusted spamd server, an
2118 attacker that gains access to a server running spamd could potentially
2119 exploit this vulnerability to run arbitrary code as the Exim user.
2121 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2122 $primary_hostname instead of what libspf2 thinks the hosts name is.
2124 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2125 a directory entry by the name of the lookup key. Previously, if a
2126 symlink pointed to a non-existing file or a file in a directory that
2127 Exim lacked permissions to read, a lookup for a key matching that
2128 symlink would fail. Now it is enough that a matching directory entry
2129 exists, symlink or not. (Bugzilla 503.)
2131 PH/03 The body_linecount and body_zerocount variables are now exported in the
2134 PH/04 Added the $dnslist_matched variable.
2136 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2137 This means they are set thereafter only if the connection becomes
2140 PH/06 Added the client_condition to authenticators so that some can be skipped
2141 by clients under certain conditions.
2143 PH/07 The error message for a badly-placed control=no_multiline_responses left
2144 "_responses" off the end of the name.
2146 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2148 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2149 (without spaces) instead of just copying the configuration text.
2151 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2153 PH/11 Added $max_received_linelength.
2155 PH/12 Added +ignore_defer and +include_defer to host lists.
2157 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2158 way in which PCRE > 7.0 is built.
2160 PH/14 Implemented queue_only_load_latch.
2162 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2163 MAIL command. The effect was to mangle the value on 64-bit systems.
2165 PH/16 Another patch from the Sieve maintainer.
2167 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2169 PH/18 If a system quota error occurred while trying to create the file for
2170 a maildir delivery, the message "Mailbox is full" was not appended to the
2171 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2172 only to a quota excession during the actual writing of the file.
2174 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2175 characters?) which causes problems in log lines. The DN values are now
2176 passed through string_printing() before being added to log lines.
2178 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2179 and InterBase are left for another time.)
2181 PH/21 Added message_body_newlines option.
2183 PH/22 Guard against possible overflow in moan_check_errorcopy().
2185 PH/23 POSIX allows open() to be a macro; guard against that.
2187 PH/24 If the recipient of an error message contained an @ in the local part
2188 (suitably quoted, of course), incorrect values were put in $domain and
2189 $local_part during the evaluation of errors_copy.
2195 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2196 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2199 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2200 issue a MAIL command.
2202 PH/02 In an ACL statement such as
2204 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2206 if a client was not listed at all, or was listed with a value other than
2207 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2208 the condition was not true (as it should be), so access was not denied.
2209 The bug was that the ! inversion was incorrectly passed on to the second
2210 item. This has been fixed.
2212 PH/03 Added additional dnslists conditions == and =& which are different from
2213 = and & when the dns lookup returns more than one IP address.
2215 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2216 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2218 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2219 FSYNC, which compiles an option called disable_fsync that allows for
2220 bypassing fsync(). The documentation is heavily laced with warnings.
2222 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2224 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2225 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2226 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2227 including adding "make clean"; (3) Added -fPIC when compiling the test
2228 dynamically loaded module, to get rid of a warning.
2230 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2231 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2232 = 0s. The bug is otherwise harmless.
2234 PH/07 There was a bug in the dovecot authenticator such that the value of
2235 $auth1 could be overwritten, and so not correctly preserved, after a
2236 successful authentication. This usually meant that the value preserved by
2237 the server_setid option was incorrect.
2239 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2241 PH/09 Installed PCRE release 7.0.
2243 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2244 run for batched SMTP input. It is now run at the start of every message
2245 in the batch. While fixing this I discovered that the process information
2246 (output by running exiwhat) was not always getting set for -bs and -bS
2247 input. This is fixed, and it now also says "batched" for BSMTP.
2249 PH/11 Added control=no_pipelining.
2251 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2252 patch, slightly modified), and move the expansion of helo_data till after
2253 the connection is made in the smtp transport (so it can use these
2256 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2258 PH/14 Added log_selector = +pid.
2260 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2262 PH/16 Add ${if forany and ${if forall.
2264 PH/17 Added dsn_from option to vary the From: line in DSNs.
2266 PH/18 Flush SMTP output before performing a callout, unless control =
2267 no_callout_flush is set.
2269 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2270 was true (the default) a successful delivery failed to delete the retry
2271 item, thus causing premature timeout of the address. The bug is now
2274 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2276 PH/21 Long custom messages for fakedefer and fakereject are now split up
2277 into multiline responses in the same way that messages for "deny" and
2278 other ACL rejections are.
2280 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2281 with slight modification.
2283 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2284 draft, changing the syntax and factoring some duplicate code".
2286 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2287 for deliveries of the second and subsequent messages over the same SMTP
2290 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2291 ${reduce, with only minor "tidies".
2293 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2295 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2296 expansion side effects.
2298 PH/27 When a message times out after an over-quota error from an Exim-imposed
2299 quota, the bounce message says "mailbox is full". This message was not
2300 being given when it was a system quota that was exceeded. It now should
2303 MH/03 Made $recipients available in local_scan(). local_scan() already has
2304 better access to the recipient list through recipients_list[], but
2305 $recipients can be useful in postmaster-provided expansion strings.
2307 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2308 in the case of a MAIL command with additional options following the
2309 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2310 were accidentally chopped off.
2312 PH/29 SMTP synchronization checks are implemented when a command is read -
2313 there is a check that no more input is waiting when there shouldn't be
2314 any. However, for some commands, a delay in an ACL can mean that it is
2315 some time before the response is written. In this time, more input might
2316 arrive, invalidly. So now there are extra checks after an ACL has run for
2317 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2318 pipelining has not been advertised.
2320 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2322 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2323 returned in the SMTP response when smtp_return_error_details was false.
2324 This has been fixed.
2326 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2327 instead of the C I/O that was originally supplied, because problems were
2328 reported on Solaris.
2330 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2331 Exim which did not show up earlier: it was assuming that a call to
2332 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2333 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2334 was a macro that became an assignment, so it seemed to work. This has
2335 changed to a proper function call with a void return, hence the compile
2336 error. Exim's code has been fixed.
2338 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2341 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2343 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2345 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2346 because it turns out to be too big or there is a timeout) there is no
2347 "Completed" line in the log. When some messages of this type were
2348 selected by exigrep, they were listed as "not completed". Others were
2349 picked up by some special patterns. I have improved the selection
2350 criteria to be more general.
2352 PH/38 The host_find_failed option in the manualroute router can now be set
2353 to "ignore", to completely ignore a host whose IP address cannot be
2354 found. If all hosts are ignored, the behaviour is controlled by the new
2355 host_all_ignored option.
2357 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2358 homing or because of multiple MX records with /mx) generated more than
2359 one IP address, and the following item turned out to be the local host,
2360 all the secondary addresses of the first item were incorrectly removed
2361 from the list, along with the local host and any following hosts (which
2362 is what is supposed to happen).
2364 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2365 whoever called Exim into the -H file. In the case of the daemon it was
2366 behaving confusingly. When first started, it used values for whoever
2367 started the daemon, but after a SIGHUP it used the Exim user (because it
2368 calls itself on a restart). I have changed the code so that it now always
2371 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2372 message are rejected with the same error (e.g. no authentication or bad
2373 sender address), and a DATA command is nevertheless sent (as can happen
2374 with PIPELINING or a stupid MUA), the error message that was given to the
2375 RCPT commands is included in the rejection of the DATA command. This is
2376 intended to be helpful for MUAs that show only the final error to their
2379 PH/42 Another patch from the Sieve maintainer.
2381 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2382 Eximstats - Fixed some broken HTML links and added missing column headers
2384 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2385 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2387 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2389 PH/43 Yet another patch from the Sieve maintainer.
2391 PH/44 I found a way to check for a TCP/IP connection going away before sending
2392 the response to the final '.' that terminates a message, but only in the
2393 case where the client has not sent further data following the '.'
2394 (unfortunately, this is allowed). However, in many cases there won't be
2395 any further data because there won't be any more messages to send. A call
2396 to select() can be used: if it shows that the input is "ready", there is
2397 either input waiting, or the socket has been closed. An attempt to read
2398 the next input character can distinguish the two cases. Previously, Exim
2399 would have sent an OK response which the client would never have see.
2400 This could lead to message repetition. This fix should cure that, at
2401 least in a lot of common cases.
2403 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2404 advertised in response to EHLO.
2410 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2411 fixed by 4.65/MH/01 (is this a record?) are fixed:
2413 (i) An empty string was always treated as zero by the numeric comparison
2414 operators. This behaviour has been restored.
2416 (ii) It is documented that the numeric comparison operators always treat
2417 their arguments as decimal numbers. This was broken in that numbers
2418 starting with 0 were being interpreted as octal.
2420 While fixing these problems I realized that there was another issue that
2421 hadn't been noticed. Values of message_size_limit (both the global option
2422 and the transport option) were treated as octal if they started with 0.
2423 The documentation was vague. These values are now always treated as
2424 decimal, and I will make that clear in the documentation.
2430 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2431 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2434 MH/01 Don't check that the operands of numeric comparison operators are
2435 integers when their expansion is in "skipping" mode (fixes bug
2436 introduced by 4.64-PH/07).
2438 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2439 child addresses, Exim now panics and dies. Previously, because the count
2440 is held in a short int, deliveries were likely to be lost. As such a
2441 large number of recipients for a single message is ridiculous
2442 (performance will be very, very poor), I have chosen to impose a limit
2443 rather than extend the field.
2449 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2450 leftover -K file (the existence of which was triggered by #402).
2451 While we were at it, introduced process PID as part of the -K
2452 filename. This should rule out race conditions when creating
2455 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2456 processing considerably. Previous code took too long for large mails,
2457 triggering a timeout which in turn triggers #401.
2459 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2460 in the DK code in transports.c. sendfile() is not really portable,
2461 hence the _LINUX specificness.
2463 TF/01 In the add_headers option to the mail command in an Exim filter,
2464 there was a bug that Exim would claim a syntax error in any
2465 header after the first one which had an odd number of characters
2468 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2469 callout verification, Exim cached a "reject" for the entire domain. This
2470 is correct for most verifications, but it is not correct for a recipient
2471 verification with use_sender or use_postmaster set, because in that case
2472 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2473 case of MAIL FROM:<> rejection from other early rejections (e.g.
2474 rejection of HELO). When verifying a recipient using a non-null MAIL
2475 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2476 Whatever the result of the callout, the value of the domain cache is
2477 left unchanged (for any other kind of callout, getting as far as trying
2478 RCPT means that the domain itself is ok).
2480 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2483 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2484 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2485 session, the daemon ignores ECONNECT errors and logs others; it now
2486 ignores EPIPE as well.
2488 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2489 (quoted-printable decoding).
2491 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2492 later the small subsequent patch to fix an introduced bug.
2494 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2496 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2498 PH/08 An error is now given if message_size_limit is specified negative.
2500 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2501 to be given (somewhat) arbitrary names.
2503 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2506 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2507 miscellaneous code fixes
2509 PH/10 Added the log_reject_target ACL modifier to specify where to log
2512 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2513 hostname. This is wrong, because it relates to the incoming message (and
2514 probably the interface on which it is arriving) and not to the outgoing
2515 callout (which could be using a different interface). This has been
2516 changed to use the value of the helo_data option from the smtp transport
2517 instead - this is what is used when a message is actually being sent. If
2518 there is no remote transport (possible with a router that sets up host
2519 addresses), $smtp_active_hostname is used.
2521 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2522 tweaks were necessary in order to get it to work (see also 21 below):
2523 (a) The code assumed that strncpy() returns a negative number on buffer
2524 overflow, which isn't the case. Replaced with Exim's string_format()
2526 (b) There were several signed/unsigned issues. I just did the minimum
2527 hacking in of casts. There is scope for a larger refactoring.
2528 (c) The code used strcasecmp() which is not a standard C function.
2529 Replaced with Exim's strcmpic() function.
2530 (d) The code set only $1; it now sets $auth1 as well.
2531 (e) A simple test gave the error "authentication client didn't specify
2532 service in request". It would seem that Dovecot has changed its
2533 interface. Fortunately there's a specification; I followed it and
2534 changed what the client sends and it appears to be working now.
2536 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2539 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2540 address A is aliased to B and C, where B exists and C does not. Without
2541 -v the output is "A verified" because verification stops after a
2542 successful redirection if more than one address is generated. However,
2543 with -v the child addresses are also verified. Exim was outputting "A
2544 failed to verify" and then showing the successful verification for C,
2545 with its parentage. It now outputs "B failed to verify", showing B's
2546 parentage before showing the successful verification of C.
2548 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2549 look up a TXT record in a specific list after matching in a combined
2552 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2553 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2554 they consult the DNS. I had assumed they would set it the way they
2555 wanted; and indeed my experiments on Linux seem to show that in some
2556 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2557 To be on the safe side, however, I have now made the interface to
2558 host_find_byname() similar to host_find_bydns(), with an argument
2559 containing the DNS resolver options. The host_find_byname() function now
2560 sets these options at its start, just as host_find_bydns() does. The smtp
2561 transport options dns_qualify_single and dns_search_parents are passed to
2562 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2563 of host_find_byname() use the default settings of RES_DEFNAMES
2564 (qualify_single) but not RES_DNSRCH (search_parents).
2566 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2567 spool_read_header() do less string testing, by means of a preliminary
2568 switch on the second character of optional "-foo" lines. (This is
2569 overdue, caused by the large number of possibilities that now exist.
2570 Originally there were few.) While I was there, I also converted the
2571 str(n)cmp tests so they don't re-test the leading "-" and the first
2572 character, in the hope this might squeeze out yet more improvement.
2574 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2575 flag allowing group syntax was set by the header_syntax check but not
2576 turned off, possible causing trouble later; (2) The flag was not being
2577 set at all for the header_verify test, causing "group"-style headers to
2578 be rejected. I have now set it in this case, and also caused header_
2579 verify to ignore an empty address taken from a group. While doing this, I
2580 came across some other cases where the code for allowing group syntax
2581 while scanning a header line wasn't quite right (mostly, not resetting
2582 the flag correctly in the right place). These bugs could have caused
2583 trouble for malformed header lines. I hope it is now all correct.
2585 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2586 with the "reply" argument non-NULL. The code, however (which originally
2587 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2588 but it didn't always do it. This confused somebody who was copying the
2589 code for some other use. I have removed all the tests.
2591 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2592 feature that was used to support insecure browsers during the U.S. crypto
2593 embargo. It requires special client support, and Exim is probably the
2594 only MTA that supported it -- and would never use it because real RSA is
2595 always available. This code has been removed, because it had the bad
2596 effect of slowing Exim down by computing (never used) parameters for the
2597 RSA_EXPORT functionality.
2599 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2600 authenticator to fail if there's a tab character in the incoming data
2601 (there should never be unless someone is messing about, as it's supposed
2602 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2603 if the connection is using TLS or if the remote IP is the same as the
2604 local IP, and the "valid-client-cert option" if a client certificate has
2607 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2608 authenticators. This can be used for authorization after authentication
2609 succeeds. (In the case of plaintext, it servers for both authentication
2612 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2613 if any retry times were supplied.
2615 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2616 connection, where there is no client IP address to check. In this
2617 situation, the verify now always succeeds.
2619 PH/25 Applied John Jetmore's -Mset patch.
2621 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2623 PH/27 In a string expansion for a processed (not raw) header when multiple
2624 headers of the same name were present, leading whitespace was being
2625 removed from all of them, but trailing whitespace was being removed only
2626 from the last one. Now trailing whitespace is removed from each header
2627 before concatenation. Completely empty headers in a concatenation (as
2628 before) are ignored.
2630 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2631 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2633 PH/29 [Removed. This was a change that I later backed out, and forgot to
2634 correct the ChangeLog entry (that I had efficiently created) before
2635 committing the later change.]
2637 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2638 address errors (4xx response to RCPT) over the same connection as other
2639 messages routed to the same hosts. Such deliveries are always "forced",
2640 so retry times are not inspected. This resulted in far too many retries
2641 for the affected addresses. The effect occurred only when there were more
2642 hosts than the hosts_max_try setting in the smtp transport when it had
2643 the 4xx errors. Those hosts that it had tried were not added to the list
2644 of hosts for which the message was waiting, so if all were tried, there
2645 was no problem. Two fixes have been applied:
2647 (i) If there are any address or message errors in an SMTP delivery, none
2648 of the hosts (tried or untried) are now added to the list of hosts
2649 for which the message is waiting, so the message should not be a
2650 candidate for sending over the same connection that was used for a
2651 successful delivery of some other message. This seems entirely
2652 reasonable: after all the message is NOT "waiting for some host".
2653 This is so "obvious" that I'm not sure why it wasn't done
2654 previously. Hope I haven't missed anything, but it can't do any
2655 harm, as the worst effect is to miss an optimization.
2657 (ii) If, despite (i), such a delivery is accidentally attempted, the
2658 routing retry time is respected, so at least it doesn't keep
2659 hammering the server.
2661 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2662 in ${readsocket because some servers need this prod.
2664 PH/32 Added some extra debug output when updating a wait-xxx database.
2666 PH/33 The hint "could be header name not terminated by colon", which has been
2667 given for certain expansion errors for a long time, was not being given
2668 for the ${if def:h_colon_omitted{... case.
2670 PH/34 The spec says: "With one important exception, whenever a domain list is
2671 being scanned, $domain contains the subject domain." There was at least
2672 one case where this was not true.
2674 PH/35 The error "getsockname() failed: connection reset by peer" was being
2675 written to the panic log as well as the main log, but it isn't really
2676 panic-worthy as it just means the connection died rather early on. I have
2677 removed the panic log writing for the ECONNRESET error when getsockname()
2680 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2681 runs only) independently of the message's sender address. This meant
2682 that, if the 4xx error was in fact related to the sender, a different
2683 message to the same recipient with a different sender could confuse
2684 things. In particular, this can happen when sending to a greylisting
2685 server, but other circumstances could also provoke similar problems.
2686 I have changed the default so that the retry time for these errors is now
2687 based a combination of the sender and recipient addresses. This change
2688 can be overridden by setting address_retry_include_sender=false in the
2691 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2692 remote server are returned as part of bounce messages. This was not
2693 happening for LMTP over a pipe (the lmtp transport), but now it is the
2694 same for both kinds of LMTP.
2696 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2697 in header lines that were in fact CNAMEs. This is no longer the case.
2699 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2700 and queue runs started by the daemon processed all messages. This has
2701 been fixed so that -R and -S can now usefully be given with -q<time>.
2703 PH/40 Import PCRE release 6.7 (fixes some bugs).
2705 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2707 PH/42 Give an error if -q is specified more than once.
2709 PH/43 Renamed the variables $interface_address and $interface_port as
2710 $received_ip_address and $received_port, to make it clear that these
2711 values apply to message reception, and not to the outgoing interface when
2712 a message is delivered. (The old names remain recognized, of course.)
2714 PH/44 There was no timeout on the connect() call when using a Unix domain
2715 socket in the ${readsocket expansion. There now is.
2717 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2718 be meaningful with "accept".
2720 SC/01 Eximstats V1.43
2721 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2723 SC/02 Eximstats V1.44
2724 Use a glob alias rather than an array ref in the generated
2725 parser. This improves both readability and performance.
2727 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2728 Collect SpamAssassin and rejection statistics.
2729 Don't display local sender or destination tables unless
2730 there is data to show.
2731 Added average volumes into the top table text output.
2733 SC/04 Eximstats V1.46
2734 Collect data on the number of addresses (recipients)
2735 as well as the number of messages.
2737 SC/05 Eximstats V1.47
2738 Added 'Message too big' to the list of mail rejection
2739 reasons (thanks to Marco Gaiarin).
2741 SC/06 Eximstats V1.48
2742 Mainlog lines which have GMT offsets and are too short to
2743 have a flag are now skipped.
2745 SC/07 Eximstats V1.49 (Alain Williams)
2746 Added the -emptyok flag.
2748 SC/08 Eximstats V1.50
2749 Fixes for obtaining the IP address from reject messages.
2751 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2752 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2753 whitespace changes from 4.64-PH/27
2755 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2758 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2759 are found, allow negative numbers in numeric criteria)
2761 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2763 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2764 to match changes made in 4.64-PH/43
2766 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2768 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2769 "domain" is actually an IP address, and force a failure. This locks out
2770 those revolvers/nameservers that support "A-for-A" lookups, in
2771 contravention of the specifications.
2773 PH/48 When a host name was looked up from an IP address, and the subsequent
2774 forward lookup of the name timed out, the host name was left in
2775 $sender_host_name, contrary to the specification.
2777 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2778 restricted to single-key lookups, Exim was not diagnosing an error if
2779 * or *@ was used with a query-style lookup.
2781 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2783 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2784 long ago, but noone interested enough thought of it. Let's just say that
2785 the "1.1" means that there are some new functions that weren't there at
2786 some point in the past.
2788 PH/51 Error processing for expansion failure of helo_data from an smtp
2789 transport during callout processing was broken.
2791 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2792 tested/used via the -bh/-bhc/-bs options.
2794 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2795 bug, fixed in subsequent PCRE releases).
2797 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2798 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2800 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2806 SC/01 Use a glob alias rather than an array ref in eximstats generated
2807 parser. This improves both readability and performance.
2809 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2810 Don't display local sender or destination tables in eximstats unless
2811 there is data to show.
2812 Added average volumes into the eximstats top table text output.
2814 SC/03 Collect data on the number of addresses (recipients) as well
2815 as the number of messages in eximstats.
2817 TF/01 Correct an error in the documentation for the redirect router. Exim
2818 does (usually) call initgroups() when daemonizing.
2820 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2821 with consistent privilege compared to when running as a daemon.
2823 TF/03 Note in the spec that $authenticated_id is not set for local
2824 submissions from trusted users.
2826 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2827 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2829 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2830 by adding some example configuration directives to the default
2831 configuration file. A little bit of work is required to uncomment the
2832 directives and define how usernames and passwords are checked, but
2833 there is now a framework to start from.
2835 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2836 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2837 without this. I don't know how relevant this is to other LDAP libraries.
2839 PH/02 Add the verb name to the "unknown ACL verb" error.
2841 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2843 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2845 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2846 directory not expanded when it should be if an expanded home directory
2847 was set for the address (which is overridden by the transport).
2849 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2852 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2853 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2854 because it is too late at that time, and has no effect.
2856 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2857 security issue with \' (bugzilla #107). I could not use the
2858 PQescapeStringConn() function, because it needs a PGconn value as one of
2861 PH/08 When testing addresses using -bt, indicate those final addresses that
2862 are duplicates that would not cause an additional delivery. At least one
2863 person was confused, thinking that -bt output corresponded to deliveries.
2864 (Suppressing duplicates isn't a good idea as you lose the information
2865 about possibly different redirections that led to the duplicates.)
2867 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2868 systems where poll() doesn't work, in particular OS X.
2870 PH/10 Added more information to debugging output for retry time not reached.
2872 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2873 operations in malware.c.
2875 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2878 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2879 syslog_duplication set false, log lines that would normally be written
2880 both the the main log and to the reject log were not written to syslog at
2883 PH/14 In the default configuration, change the use of "message" in ACL warn
2884 statements to "add_header".
2886 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2887 not followed by a command (e.g. "seen endif").
2889 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2890 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2893 PH/17 Added extra conditions to the default value of delay_warning_condition
2897 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2898 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2899 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2902 The Auto-Submitted: and various List- headers are standardised, whereas I
2903 don't think Precedence: ever was.
2905 PH/18 Refactored debugging code in route_finduser() to show more information,
2906 in particular, the error code if getpwnam() issues one.
2908 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2909 This is apparently needed in addition to the PH/07 change above to avoid
2910 any possible encoding problems.
2912 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2913 but not after initializing Perl.
2915 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2916 output them only if debugging. By default they are written stderr,
2917 apparently, which is not desirable.
2919 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2922 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2925 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2927 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2928 authenticated or an ident call has been made. Suppress the default
2929 values for $authenticated_id and $authenticated_sender (but permit -oMai
2930 and -oMas) when testing with -bh.
2932 PH/24 Re-jigged the order of the tests in the default configuration so that the
2933 tests for valid domains and recipients precede the DNS black list and CSA
2934 tests, on the grounds that those ones are more expensive.
2936 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2937 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2938 This bug exists in every version of Exim that I still have, right back to
2941 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2942 However, an attempt to turn on case-sensitivity in a regex key by
2943 including (?-i) didn't work because the subject string was already
2944 lowercased, and the effects were non-intuitive. It turns out that a
2945 one-line patch can be used to allow (?-i) to work as expected.
2951 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2952 other effects) broke the use of negated acl sub-conditions.
2954 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2957 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2958 "Deny" causes Exim to reject the incoming connection with a 554 error.
2959 Unfortunately, if there is a major crisis, such as a disk failure,
2960 tcp-wrappers gives "deny", whereas what one would like would be some
2961 kind of temporary error. A kludge has been added to help with this.
2962 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2963 554 error is used if errno is still zero or contains ENOENT (which occurs
2964 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2967 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2969 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2970 errors. Otherwise a message that provokes a temporary error (when other
2971 messages do not) can cause a whole host to time out.
2973 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2974 addresses were routed directly to files or pipes from a redirect router.
2975 File deliveries just didn't batch; pipe deliveries might have suffered
2978 PH/06 A failure to get a lock for a hints database would erroneously always say
2979 "Failed to get write lock", even when it was really a read lock.
2981 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2982 of 0600. This has been changed to use the value of the lockfile_mode
2983 option (which defaults to 0600).
2985 PH/08 Applied small patch from the Sieve maintainer.
2987 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2988 folder from quota calculations, a direct delivery into this folder messed
2989 up the contents of the maildirsize file. This was because the regex was
2990 used only to exclude .Trash (or whatever) when the size of the mailbox
2991 was calculated. There was no check that a delivery was happening into an
2992 excluded directory. This bug has been fixed by ignoring all quota
2993 processing for deliveries into excluded directories.
2995 PH/10 Added the maildirfolder_create_regex option to appendfile.
3001 PH/01 The code for finding all the local interface addresses on a FreeBSD
3002 system running IPv6 was broken. This may well have applied to all BSD
3003 systems, as well as to others that have similar system calls. The broken
3004 code found IPv4 interfaces correctly, but gave incorrect values for the
3005 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3006 that it would not match correctly against @[] and not recognize the IPv6
3009 PH/02 The ipliteral router was not recognizing addresses of the form user@
3010 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3012 PH/03 Added disable_ipv6.
3014 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3015 decoded form, because it is most often used to construct To: headers
3016 lines in autoreplies, and the decoded form may well be syntactically
3017 invalid. However, $reply_address has leading white space removed, and all
3018 newlines turned into spaces so that the autoreply transport does not
3021 PH/05 If group was specified without a user on a router, and no group or user
3022 was specified on a transport, the group from the router was ignored.
3024 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3025 for visible compile-time settings that can be used to change these
3026 numbers, for those that want even more. Backwards compatibility with old
3027 spool files has been maintained. However, going back to a previous Exim
3028 release will lost any variables that are in spool files.
3030 PH/07 Two small changes when running in the test harness: increase delay when
3031 passing a TCP/IP connection to a new process, in case the original
3032 process has to generate a bounce, and remove special handling of
3033 127.0.0.2 (sic), which is no longer necessary.
3035 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3036 be the same on different OS.
3038 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3041 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3042 whether --show-vars was specified or not
3044 JJ/02 exipick: Added support for new ACL variable spool format introduced
3047 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3048 syntactically invalid From: or Reply-to: line, and a filter used this to
3049 generate an autoreply, and therefore failed to obtain an address for the
3050 autoreply, Exim could try to deliver to a non-existent relative file
3051 name, causing unrelated and misleading errors. What now happens is that
3052 it logs this as a hard delivery error, but does not attempt to create a
3055 PH/11 The exinext utility has a -C option for testing purposes, but although
3056 the given file was scanned by exinext itself; it wasn't being passed on
3057 when Exim was called.
3059 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3060 an end-of-file indication when reading a command response.
3062 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3063 compiled. In many other places in Exim, IPv6 addresses are always
3064 recognized, so I have changed this. It also means that IPv4 domain
3065 literals of the form [IPV4:n.n.n.n] are now always recognized.
3067 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3068 used if the router is not running as root, for example, when verifying at
3069 ACL time, or when using -bh. The debugging output from this situation was
3070 non-existent - all you got was a failure to exec. I have made two
3073 (a) Failures to set uid/gid, the current directory, or a process leader
3074 in a subprocess such as that created by queryprogram now generate
3075 suitable debugging output when -d is set.
3077 (b) The queryprogram router detects when it is not running as root,
3078 outputs suitable debugging information if -d is set, and then runs
3079 the subprocess without attempting to change uid/gid.
3081 PH/15 Minor change to Makefile for building test_host (undocumented testing
3084 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3085 additional section of a DNS packet that returns MX or SRV records.
3086 Instead, it always explicitly searches for A/AAAA records. This avoids
3087 major problems that occur when a DNS server includes only records of one
3088 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3089 fixed another bug: if SRV records were looked up and the corresponding
3090 address records were *not* found in the additional section, the port
3091 values from the SRV records were lost.
3093 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3094 using the correct key (the original address) when searching the retry
3095 rules in order to find which one to use for generating the retry hint.
3097 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3098 adding the default one. Similarly, if it contains a Reply-To: header, the
3099 errors_reply_to option, if set, is not used.
3101 PH/19 When calculating a retry time, Exim used to measure the "time since
3102 failure" by looking at the "first failed" field in the retry record. Now
3103 it does not use this if it is later than than the arrival time of the
3104 message. Instead it uses the arrival time. This makes for better
3105 behaviour in cases where some deliveries succeed, thus re-setting the
3106 "first failed" field. An example is a quota failure for a huge message
3107 when small messages continue to be delivered. Without this change, the
3108 "time since failure" will always be short, possible causing more frequent
3109 delivery attempts for the huge message than are intended.
3110 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3112 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3113 $1, $2, $3) because the numerical variables can be reset during some
3114 expansion items (e.g. "match"), thereby losing the authentication data.
3116 PH/21 Make -bV show the size of off_t variables so that the test suite can
3117 decide whether to run tests for quotas > 2G.
3119 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3120 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3121 filecount value is greater than 2G or if a quota value is greater than 2G
3122 on a system where the size of off_t is not greater than 4, a panic error
3125 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3126 never match. The debug and -bh output now contains an explicit error
3127 message indicating a malformed IPv4 address or mask.
3129 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3130 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3131 PH/23 above applies.
3133 PH/25 Do not write to syslog when running in the test harness. The only
3134 occasion when this arises is a failure to open the main or panic logs
3135 (for which there is an explicit test).
3137 PH/26 Added the /no_tell option to "control=freeze".
3139 PH/27 If a host name lookup failed very early in a connection, for example, if
3140 the IP address matched host_lookup and the reverse lookup yielded a name
3141 that did not have a forward lookup, an error message of the form "no IP
3142 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3143 could be logged. Now it outputs the IP address instead of "NULL".
3145 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3146 allows the sender and the authenticated sender to be set when
3147 submitting a message from within Exim. Since child_open_exim() is
3148 documented for local_scan(), the new function should be too.
3150 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3151 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3152 results in an empty string is now treated as unset.
3154 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3156 PH/31 Added sender_verify_fail logging option.
3158 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3159 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3160 tidied the source and removed it altogether.
3162 PH/33 When a queue run was abandoned because the load average was too high, a
3163 log line was always written; now it is written only if the queue_run log
3164 selector is set. In addition, the log line for abandonment now contains
3165 information about the queue run such as the pid. This is always present
3166 in "start" and "stop" lines but was omitted from the "abandon" line.
3168 PH/34 Omit spaces between a header name and the colon in the error message that
3169 is given when verify = headers_syntax fails (if there are lots of them,
3170 the message gets confusing).
3172 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3173 names, as there are now some PTR records that contain slashes. This check
3174 is only to protect against broken name servers that fall over on strange
3175 characters, so the fact that it applies to all lookups doesn't matter.
3177 PH/36 Now that the new test suite is complete, we can remove some of the
3178 special code in Exim that was needed for the old test suite. For example,
3179 sorting DNS records because real resolvers return them in an arbitrary
3180 order. The new test suite's fake resolver always returns records in the
3183 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3184 bounces) except when queue_only is set, to avoid logging races between
3185 the different processes.
3187 PH/38 Panic-die if .include specifies a non-absolute path.
3189 PH/39 A tweak to the "H" retry rule from its user.
3191 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3192 a label. They prevented compilation on older perls.
3194 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3195 a warning to be raised on newish perls.
3197 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3198 on queue. Changes to match documented behaviour of showing count of
3199 messages matching specified criteria.
3201 PH/40 Changed the default ident timeout from 30s to 5s.
3203 PH/41 Added support for the use of login_cap features, on those BSD systems
3204 that have them, for controlling the resources used by pipe deliveries.
3206 PH/42 The content-scanning code uses fopen() to create files in which to put
3207 message data. Previously it was not paying any attention to the mode of
3208 the files. Exim runs with umask(0) because the rest of the code creates
3209 files with open(), and sets the required mode explicitly. Thus, these
3210 files were ending up world-writeable. This was not a big issue, because,
3211 being within the spool directory, they were not world-accessible. I have
3212 created a function called modefopen, which takes an additional mode
3213 argument. It sets umask(777), creates the file, chmods it to the required
3214 mode, then resets the umask. All the relevant calls to fopen() in the
3215 content scanning code have been changed to use this function.
3217 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3218 to 24 hours. This avoids potential overflow problems when processing G
3219 and H retry rules. I suspect nobody ever tinkers with this value.
3221 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3223 PH/45 When the plaintext authenticator is running as a client, the server's
3224 challenges are checked to ensure they are valid base64 strings. By
3225 default, the authentication attempt is cancelled if an invalid string is
3226 received. Setting client_ignore_invalid_base64 true ignores these errors.
3227 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3228 they are received. Thus, the responses can be made to depend on the
3229 challenges. If an invalid string is ignored, an empty string is placed in
3232 PH/46 Messages that are created by the autoreply transport now contains a
3233 References: header, in accordance with RFCs 2822 and 3834.
3235 PH/47 Added authenticated_sender_force to the smtp transport.
3237 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3239 PH/49 Installed latest patch from the Sieve maintainer.
3241 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3242 was also set, the appendfile transport was unnecessarily scanning a
3243 directory of message files (e.g. for maildir delivery) to find the count
3244 of files (along with the size), even though it did not need this
3245 information. It now does the scan only if it needs to find either the
3246 size of the count of files.
3248 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3250 PH/52 Two bugs concerned with error handling when the smtp transport is
3253 (i) Exim was not creating retry information for temporary errors given
3254 for individual recipients after the DATA command when the smtp transport
3255 was used in LMTP mode. This meant that they could be retried too
3256 frequently, and not timed out correctly.
3258 (ii) Exim was setting the flag that allows error details to be returned
3259 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3260 recipients that were returned after the DATA command.
3262 PH/53 This is related to PH/52, but is more general: for any failing address,
3263 when detailed error information was permitted to be returned to the
3264 sender, but the error was temporary, then after the final timeout, only
3265 "retry timeout exceeded" was returned. Now it returns the full error as
3266 well as "retry timeout exceeded".
3268 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3269 do this, and (what is worse) MTAs that accept it.
3271 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3272 will now be deprecated.
3274 PH/56 New os.c-cygwin from the Cygwin maintainer.
3276 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3277 formats (previously only available in exim formats via -bpr, -bpru,
3278 and -bpra. Now also available in native and exiqgrep formats)
3280 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3281 with very large, slow to parse queues
3283 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3285 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3287 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3288 responses to authentication challenges, though it was showing the
3289 challenges; (ii) I've removed the CR characters from the debug output for
3292 PH/58 Allow for the insertion of a newline as well as a space when a string
3293 is turned into more than one encoded-word during RFC 2047 encoding. The
3294 Sieve code now uses this.
3296 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3297 data_4xx, lost_connection, tls_required.
3299 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3300 message was being sent as an SMTP response.
3302 PH/61 Add -l and -k options to exicyclog.
3304 PH/62 When verifying, if an address was redirected to one new address, so that
3305 verification continued, and the new address failed or deferred after
3306 having set something in $address_data, the value of $address_data was not
3307 passed back to the ACL. This was different to the case when no
3308 redirection occurred. The value is now passed back in both cases.
3310 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3311 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3312 use login_cap.h, so on its own it isn't the distinguishing feature. The
3313 new name refers directly to the setclassresources() function.
3315 PH/65 Added configuration files for NetBSD3.
3317 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3319 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3320 is preferred over IPv4.
3322 PH/68 The bounce_return_message and bounce_return_body options were not being
3323 honoured for bounces generated during the reception of non-SMTP messages.
3324 In particular, this applied to messages rejected by the ACL. This bug has
3325 been fixed. However, if bounce_return_message is true and bounce_return_
3326 body is false, the headers that are returned for a non-SMTP message
3327 include only those that have been read before the error was detected.
3328 (In the case of an ACL rejection, they have all been read.)
3330 PH/69 The HTML version of the specification is now built in a directory called
3331 spec_html instead of spec.html, because the latter looks like a path with
3332 a MIME-type, and this confuses some software.
3334 PH/70 Catch two compiler warnings in sieve.c.
3336 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3337 function verify_get_ident() calls ip_connect() to connect a socket, but
3338 if the "connect()" function timed out, ip_connect() used to close the
3339 socket. However, verify_get_ident() also closes the socket later, and in
3340 between Exim writes to the log, which may get opened at this point. When
3341 the socket was closed in ip_connect(), the log could get the same file
3342 descriptor number as the socket. This naturally causes chaos. The fix is
3343 not to close the socket in ip_connect(); the socket should be closed by
3344 the function that creates it. There was only one place in the code where
3345 this was missing, in the iplookup router, which I don't think anybody now
3346 uses, but I've fixed it anyway.
3348 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3349 well as to direct DNS lookups. Otherwise the handling of names in host
3350 lists is inconsistent and therefore confusing.
3356 PH/01 Two changes to the default runtime configuration:
3358 (1) Move the checks for relay_from_hosts and authenticated clients from
3359 after to before the (commented out) DNS black list checks.
3361 (2) Add control=submission to the relay_from_hosts and authenticated
3362 clients checks, on the grounds that messages accepted by these
3363 statements are most likely to be submissions.
3365 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3367 (1) Generate an error if the third argument for the ${prvs expansion is
3370 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3373 (3) Reset the variables that are obtained from the first argument of
3374 ${prvscheck and used in the second argument before leaving the code,
3375 because their memory is reclaimed, so using them afterwards may do
3378 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3379 one (it's much easier than Tom thought :-).
3381 (5) Because of (4), we can now allow for the use of $prvscheck_result
3382 inside the third argument.
3384 PH/03 For some reason, the default setting of PATH when running a command from
3385 a pipe transport was just "/usr/bin". I have changed it to
3388 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3389 anything to be listed in the output from -bV.
3391 PH/05 When a filter generated an autoreply, the entire To: header line was
3392 quoted in the delivery log line, like this:
3394 => >A.N.Other <ano@some.domain> <original@ddress> ...
3396 This has been changed so that it extracts the operative address. There
3397 may be more than one such address. If so, they are comma-separated, like
3400 => >ano@some.domain,ona@other.domain <original@ddress> ...
3402 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3403 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3404 not being looked up in the rDNS to get a host name, Exim was showing the
3405 IP address twice in Received: lines, even though the IP addresses were
3406 identical. For example:
3408 Received: from [1.2.3.4] (helo=[1.2.3.4])
3410 However, if the real host name was known, it was omitting the HELO data
3411 if it matched the actual IP address. This has been tidied up so that it
3412 doesn't show the same IP address twice.
3414 PH/07 When both +timestamp and +memory debugging was on, the value given by
3415 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3416 called by the debug printing, thereby overwriting the timestamp buffer.
3417 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3419 PH/08 When the original message was included in an autoreply transport, it
3420 always said "this is a copy of the message, including all the headers",
3421 even if body_only or headers_only was set. It now gives an appropriate
3424 PH/09 Applied a patch from the Sieve maintainer which:
3426 o fixes some comments
3427 o adds the (disabled) notify extension core
3428 o adds some debug output for the result of if/elsif tests
3429 o points to the current vacation draft in the documentation
3430 and documents the missing references header update
3434 o fixes a bug in processing the envelope test (when testing
3435 multiple envelope elements, the last element determined the
3438 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3439 Electronic Mail") by including:
3441 Auto-submitted: auto-generated
3443 in the messages that it generates (bounce messages and others, such as
3444 warnings). In the case of bounce messages for non-SMTP messages, there was
3445 also a typo: it was using "Auto_submitted" (underscore instead of
3446 hyphen). Since every message generated by Exim is necessarily in response
3447 to another message, thes have all been changed to:
3449 Auto-Submitted: auto-replied
3451 in accordance with these statements in the RFC:
3453 The auto-replied keyword:
3455 - SHOULD be used on messages sent in direct response to another
3456 message by an automatic process,
3458 - MUST NOT be used on manually-generated messages,
3460 - MAY be used on Delivery Status Notifications (DSNs) and Message
3461 Disposition Notifications (MDNs),
3463 - MUST NOT be used on messages generated by automatic or periodic
3464 processes, except for messages which are automatic responses to
3467 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3468 to the default Received: header definition.
3470 PH/12 Added log selector acl_warn_skipped (default on).
3472 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3473 variables because (a) they are in the wrong storage pool and (b) even if
3474 they were copied, it wouldn't work properly because of the caching.
3476 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3477 checking when decoding. Apparently there are clients that generate
3478 overlong encoded strings. Why am I not surprised?
3480 PH/15 If the first argument of "${if match_address" was not empty, but did not
3481 contain an "@" character, Exim crashed. Now it writes a panic log message
3482 and treats the condition as false.
3484 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3486 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3487 extension "envelope-auth". The code is finished and in agreement with
3488 other implementations, but there is no documentation so far and in fact,
3489 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3490 not changing the active code.
3492 Print executed "if" and "elsif" statements when debugging is used. This
3493 helps a great deal to understand what a filter does.
3495 Document more things not specified clearly in RFC3028. I had all this
3496 sorted out, when out of a sudden new issues came to my mind. Oops."
3498 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3501 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3502 canonical form (as documented). However, after a host name lookup from
3503 the IP address, check_host() was doing a simple string comparison with
3504 addresses acquired from the DNS when checking that the found name did
3505 have the original IP as one of its addresses. Since any found IPv6
3506 addresses are likely to be in abbreviated form, the comparison could
3507 fail. Luckily, there already exists a function for doing the comparison
3508 by converting both addresses to binary, so now that is used instead of
3509 the text comparison.
3511 PH/20 There was another similar case to PH/19, when a complete host name was
3512 given in a host list; looking up its IP address could give an abbreviated
3513 form, whereas the current host's name might or might not be abbreviated.
3514 The same fix has been applied.
3520 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3521 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3524 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3525 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3527 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3529 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3530 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3531 Also, for "Auto-Submitted", treat anything other than "no" as
3532 non-personal, in accordance with RFC 3834. (Previously it treated
3533 anything starting "auto-" as non-personal.)
3535 TF/01 The control=submission/name=... option had a problem with syntax
3536 errors if the name included a slash character. The /name= option
3537 now slurps the rest of the string, so it can include any characters
3538 but it must come last in the list of options (after /sender_retain
3541 PH/05 Some modifications to the interface to the fake nameserver for the new
3549 TK/01 Added the "success_on_redirect" address verification option. See
3550 NewStuff for rationale and an example.
3552 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3554 PH/02 Patch to exigrep to allow it to work on syslog lines.
3556 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3557 fread() to skip over the body file's header line, because in Cygwin the
3558 header line is locked and is inaccessible.
3560 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3561 co-exist for some time) to make it clear that it is the Exim ID that is
3562 referenced, not the Message-ID: header line.
3564 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3565 string_format() function, because snprintf() does not exist on all
3568 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3569 ${stat: expansion item.
3571 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3572 protocol synchronization error", to keep the pedants happy.
3574 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3575 well as for IRIX systems, when gcc is being used. See the host.c source
3578 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3580 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3583 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3584 transport and to the smtp transport in LMTP mode.
3586 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3588 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3589 run a filter in a subprocess. This could lead to confusion in subsequent
3590 lookups in the parent process. There should also be a search_tidyup() at
3591 the end of the subprocess.
3593 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3594 only if the host matched helo_try_verify_hosts, which caused the
3595 verification to occur when the EHLO/HELO command was issued. The ACL just
3596 tested the remembered result. Now, if a previous verification attempt has
3597 not happened, "verify = helo" does it there and then.
3599 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3601 TK/03 Fix log output including CR from clamd.
3603 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3604 exist provoked a memory error which could cause a segfault.
3606 PH/15 Installed PCRE 6.2
3608 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3610 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3611 of the problem. Specifically, suggested +O2 rather than +O1 for the
3614 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3616 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3617 record, the port was not correctly logged when the outgoing_port log
3618 selector was set (it logged the transort's default port).
3620 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3621 fallback_hosts, and "hosts" in the smtp transport.
3623 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3624 host errors such as "Connection refused".
3626 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3627 authentication with radiusclient 0.4.9:
3629 - Error returned from rc_read_config was caught wrongly
3630 - Username/password not passed on to radius server due to wrong length.
3632 The presumption is that some radiusclient API changes for 4.51/PH/17
3633 were not taken care of correctly. The code is still untested by me (my
3634 Linux distribution still has 0.3.2 of radiusclient), but it was
3635 contributed by a Radius user.
3637 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3638 expanding the "port" option of the smtp transport.
3640 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3641 while reading a MIME header. Thanks to Tom Hughes for a patch.
3643 PH/24 Include config.h inside local_scan.h so that configuration settings are
3646 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3647 that in an ACL for RCPT (for example), you can examine exactly what was
3650 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3651 commands, but it was not correctly comparing the address with the actual
3652 client host address. Thus, it would show the EHLO address in Received:
3653 header lines when this was not necessary.
3655 PH/27 Added the % operator to ${eval:}.
3657 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3658 it should be ignoring failures (because with -C, for example, it has lost
3659 privilege). It wasn't ignoring creation failures other than "already
3662 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3665 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3666 because an input line was too long, either on its own, or by virtue of
3667 too many continuations, the temporary file was not being removed, and the
3668 return code was incorrect.
3670 PH/31 Missing "BOOL" in function definition in filtertest.c.
3672 PH/32 Applied Sieve patches from the maintainer.
3674 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3676 PH/33 Added "verify = not_blind".
3678 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3679 Local/Makefile (with some defaults set). These are used in built scripts
3680 such as exicyclog, but they have never been used in the exim_install
3681 script (though there are many overriding facilities there). I have
3682 arranged that the exim_install script now takes note of these two
3685 PH/35 Installed configuration files for Dragonfly.
3687 PH/36 When a locally submitted message by a trusted user did not contain a
3688 From: header, and the sender address was obtained from -f or from an SMTP
3689 MAIL command, and the trusted user did not use -F to supply a sender
3690 name, $originator_name was incorrectly used when constructing a From:
3691 header. Furthermore, $originator_name was used for submission mode
3692 messages from external hosts without From: headers in a similar way,
3693 which is clearly wrong.
3695 PH/37 Added control=suppress_local_fixups.
3697 PH/38 When log_selector = +received_sender was set, and the addition of the
3698 sender made the log line's construction buffer exactly full, or one byte
3699 less than full, an overflow happened when the terminating "\n" was
3702 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3703 when the result of a list match is failure because a DNS lookup failed.
3705 PH/40 RM_COMMAND is now used in the building process.
3707 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3708 the "build-* directories that it finds.
3710 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3711 address in a domain literal was a prefix of an interface address.
3713 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3714 when verifying a sender address, unless rewrite_headers is false.
3716 PH/44 Wrote a long comment about why errors_to addresses are verified as
3717 recipients, not senders.
3719 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3720 the ratelimit ACL was added.
3722 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3724 PH/46 Added extra information about PostgreSQL errors to the error string.
3726 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3727 avoiding the need to install special zones in a real server. This is
3728 backwards compatible; if it can't find the fake resolver, it drops back.
3729 Thus, both old and new test suites can be run.
3731 TF/02 Added util/ratelimit.pl
3733 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3734 clock is set back in time.
3736 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3737 Candler <B.Candler@pobox.com>.
3739 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3740 OK for addresses that are the result of redirections.
3742 PH/48 A number of further additions for the benefit of the new test suite,
3743 including a fake gethostbyname() that interfaces to the fake DNS resolver
3746 TF/06 The fix for widen_domains has also been applied to qualify_single and
3747 search_parents which are the other dnslookup options that can cause
3750 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3753 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3755 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3756 service" if it encountered an MX record with an empty target hostname.
3757 The message is now "an MX or SRV record indicated no SMTP service".
3759 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3760 if the DNS of the sending site is misconfigured. This is quite a
3761 common situation. This change restores the behaviour of treating a
3762 helo verification defer as a failure.
3764 PH/51 If self=fail was set on a router, the bounce message did not include the
3765 actual error message.
3771 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3773 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3774 command itself ended in error, the underlying message about the transport
3775 filter timeout was being overwritten with the pipe command error. Now the
3776 underlying error message should be appended to the second error message.
3778 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3780 PH/02 Reduce the amount of output that "make" produces by default. Full output
3781 can still be requested.
3783 PH/03 The warning log line about a condition test deferring for a "warn" verb
3784 was being output only once per connection, rather than after each
3785 occurrence (because it was using the same function as for successful
3786 "warn" verbs). This seems wrong, so I have changed it.
3788 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3789 it should not have, which might have caused a crash in the right
3790 circumstances, but probably never did.
3792 PH/04 Installed a modified version of Tony Finch's patch to make submission
3793 mode fix the return path as well as the Sender: header line, and to
3794 add a /name= option so that you can make the user's friendly name appear
3797 TF/03 Added the control = fakedefer ACL modifier.
3799 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3800 Mark Lowes for thorough testing.
3802 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3804 TK/03 Merged latest SRS patch from Miles Wilton.
3806 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3807 with the definition in sysexits.h (which is #included earlier).
3808 Fortunately, Exim does not actually use EX_OK. The code used to try to
3809 preserve the sysexits.h value, by assuming that macro definitions were
3810 scanned for macro replacements. I have been disabused of this notion,
3811 so now the code just undefines EX_OK before #including unistd.h.
3813 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3814 in the smtp transport. When a block could not be written in a single
3815 write() function, the timeout was being re-applied to each part-write.
3816 This seems wrong - if the receiver was accepting one byte at a time it
3817 would take for ever. The timeout is now adjusted when this happens. It
3818 doesn't have to be particularly precise.
3820 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3821 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3823 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3824 without a domain if the check to <postmaster@domain> fails.
3826 SC/01 Eximstats: added -xls and the ability to specify output files
3827 (patch written by Frank Heydlauf).
3829 SC/02 Eximstats: use FileHandles for outputting results.
3831 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3833 SC/04 Eximstats: fixed display of large numbers with -nvr option
3835 SC/05 Eximstats: fixed merging of reports with empty tables.
3837 SC/06 Eximstats: added the -include_original_destination flag
3839 SC/07 Eximstats: removed tabs and trailing whitespace.
3841 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3843 TK/06 MBOX spool code: Add real "From " MBOX separator line
3844 so the .eml file is really in mbox format (even though
3845 most programs do not really care). Patch from Alex Miller.
3847 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3848 The latter is generated from $received_to and is only set if the
3849 message has one envelope recipient. SA can use these headers,
3850 obviously out-of-the-box. Patch from Alex Miller.
3852 PH/08 The ${def test on a variable was returning false if the variable's
3853 value was "0", contrary to what the specification has always said!
3854 The result should be true unless the variable is empty.
3856 PH/09 The syntax error of a character other than { following "${if
3857 def:variable_name" (after optional whitespace) was not being diagnosed.
3858 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3859 accidental colon was present, for example, could give incorrect results.
3861 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3862 result is used (not including appendfile, where other changes are about
3865 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3866 This involved changing a lot of size variables from int to off_t. It
3867 should work with maildirs and everything.
3869 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3870 spamd dying while we are connected to it.
3872 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3875 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3876 function for BDB 4.3.
3878 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3880 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3881 the sources for PCRE have been re-arranged and more files are now
3884 PH/15 The code I had for printing potentially long long variables in PH/11
3885 above was not the best (it lost precision). The length of off_t variables
3886 is now inspected at build time, and an appropriate printing format (%ld
3887 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3888 to be "long long int" or "long int". This is needed for the internal
3889 formatting function string_vformat().
3891 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3892 the configuration file to be ":syslog", then the script "guesses" where
3893 the logs files are, rather than using the compiled in default. In our
3894 case the guess is not the same as the compiled default, so the script
3895 suddenly stopped working when I started to use syslog. The patch checks
3896 to see if log_file_path is "". If so, it attempts to read it from exim
3897 with no configuration file to get the compiled in version, before it
3898 falls back to the previous guessing code."
3900 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3901 implementing BATV in an Exim configuration. See NewStuff for the gory
3904 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3905 Makefile that are specific to HP-UX.
3907 PH/18 If the "use_postmaster" option was set for a recipient callout together
3908 with the "random" option, the postmaster address was used as the MAIL
3909 FROM address for the random test, but not for the subsequent recipient
3910 test. It is now used for both.
3912 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3913 patch removes a few documentation additions to RFC 3028, because the
3914 latest draft now contains them. It adds the new en;ascii-case comparator
3915 and a new error check for 8bit text in MIME parts. Comparator and
3916 require names are now matched exactly. I enabled the subaddress
3917 extension, but it is not well tested yet (read: it works for me)."
3919 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3920 rework some of the code of TK/09 above to avoid the hardwired use of
3921 "%lld" and "long long". Replaced the call to snprintf() with a call to
3924 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3925 records point to non-existent hosts", "retry timeout exceeded", and
3926 "retry time not reached for any host after a long failure period".
3928 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3929 experimental DomainKeys support:
3931 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3932 (2) On an error such as an illegally used "control", the wrong name for
3933 the control was given.
3935 These problems did NOT occur unless DomainKeys support was compiled.
3937 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3939 PH/24 Added ${if match_ip condition.
3941 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3942 databases so that it will be absolutely obvious if a crash occurs in the
3943 DB library. This is a regular occurrence (often caused by mis-matched
3946 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3947 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3948 picked up on a user's system that detects such things. There doesn't seem
3949 to be a gcc warning option for this - only an attribute that has to be
3950 put on the function's prototype. It seems that in Fedora Core 4 they have
3951 set this on a number of new functions. No doubt there will be more in due
3954 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3955 specify a transport. However, if an address that was verified by such a
3956 router was the subject of a callout, Exim crashed because it tried to
3957 read the rcpt_include_affixes from the non-existent transport. Now it
3958 just assumes that the setting of that option is false. This bug was
3959 introduced by 4.51/PH/31.
3961 PH/28 Changed -d+all to exclude +memory, because that information is very
3962 rarely of interest, but it makes the output a lot bigger. People tend to
3963 do -d+all out of habit.
3965 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3966 code in os-type was giving problems when libc.so lives in lib64, like on
3969 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3970 aren't the modern standard, and it seems that some systems' include files
3971 don't always have them. Exim was already checking for some of the newer
3972 ones like T_AAAA, and defining it itself. I've added checks for all the
3973 record types that Exim uses.
3975 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3976 not automatically generating a new one, as it is supposed to. This
3977 prevented TLS from working. If the file did exist, but contained invalid
3978 data, a new version was generated, as expected. It was only the case of a
3979 non-existent file that was broken.
3981 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3982 with a change in libdomainkeys > 0.64.
3984 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3985 from DNS. If the selector record carries the flag, it now has
3986 precedence over the domain-wide flag.
3988 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3990 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3991 the use of a port name that isn't defined in /etc/services) occurred, the
3992 message was deferred as in a normal delivery, and thus remained on the
3993 spool, instead of being failed because of the mua_wrapper setting. This
3994 is now fixed, and I tidied up some of the mua_wrapper messages at the
3997 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3998 the messages in a hash of arrays rather than using individual hashes.
3999 This is a bit cleaner and results in dramatic memory savings, albeit
4000 at a slight CPU cost.
4002 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4003 as requested by Marc Sherman.
4005 SC/10 Eximstats: added histograms for user specified patterns as requested
4008 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4010 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4011 fopen() in the content-scanning modules that did not already have it.
4017 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4018 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4020 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4022 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4024 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4025 to test Sieve filters that use "vacation".
4027 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4028 that changes the way the GnuTLS parameters are stored in the cache file.
4029 The new format can be generated externally. For backward compatibility,
4030 if the data in the cache doesn't make sense, Exim assumes it has read an
4031 old-format file, and it generates new data and writes a new file. This
4032 means that you can't go back to an older release without removing the
4035 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4036 work if there are any delivery delays because "one_time" forces the
4037 parent to be marked "delivered", so its unseen clone is never tried
4038 again. For this reason, Exim now forbids the simultaneous setting of
4041 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4042 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4043 entry if you want to know the details. The fix, however, neglected to
4044 consider the case when local delivery batching is involved. The test for
4045 "previously delivered" was not happening when checking to see if an
4046 address could be batched with a previous (undelivered) one; under
4047 certain circumstances this could lead to multiple deliveries to the same
4050 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4051 in its include files, and this causes problems building Exim.
4053 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4054 header_syntax) but Exim was just ignoring anything given after a slash.
4055 In particular, this caused confusion with an attempt to use "verify =
4056 reverse_host_lookup/defer_ok". An error is now given when options are
4057 supplied for verify items that do not have them. (Maybe reverse_host_
4058 lookup should have a defer_ok option, but that's a different point.)
4060 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4061 defined by RFC 821) to 2048, because there were problems with some AUTH
4062 commands, and RFC 1869 says the size should be increased for extended
4063 SMTP commands that take arguments.
4065 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4068 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4069 "unknown" error; now it says that the functionality isn't in the binary.
4071 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4072 an address' error message when a string expansion fails (syntax or
4073 whatever). Otherwise the password may appear in the log. Following change
4074 PH/42 below, there is no longer a chance of it appearing in a bounce
4077 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4079 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4080 of its addresses was ever tried. (Bugzilla bug #2.)
4082 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4083 the result incorrectly in the debug output. (It correctly added a newline
4084 to what was transported.)
4086 TF/01 Added $received_time.
4088 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4089 commented out examples of how to interface to a virus scanner and to
4090 SpamAssassin. Also added commented examples of av_scanner and
4091 spamd_address settings.
4093 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4094 and controls are allowed in which ACLs. There were a couple of minor
4095 errors. Some of the entries in the conditions table (which is a table of
4096 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4097 negation of where the condition IS allowed.
4099 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4101 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4102 header file does not have a version number, so I've had to invent a new
4103 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4104 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4105 radiusclient), but it was contributed by a Radius user.
4107 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4108 files or directories, for OpenSSL.
4110 PH/19 When an Exim process that is running as root has to create an Exim log
4111 file, it does so in a subprocess that runs as exim:exim so as to get the
4112 ownership right at creation (otherwise, other Exim processes might see
4113 the file with the wrong ownership). There was no test for failure of this
4114 fork() call, which would lead to the process getting stuck as it waited
4115 for a non-existent subprocess. Forks do occasionally fail when resources
4116 run out. I reviewed all the other calls to fork(); they all seem to check
4119 PH/20 When checking for unexpected SMTP input at connect time (before writing
4120 the banner), Exim was not dealing correctly with a non-positive return
4121 from the read() function. If the client had disconnected by this time,
4122 the result was a log entry for a synchronization error with an empty
4123 string after "input=" when read() returned zero. If read() returned -1
4124 (an event I could not check), uninitialized data bytes were printed.
4125 There were reports of junk text (parts of files, etc) appearing after
4128 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4130 PH/22 Added support for macro redefinition, and (re)definition in between
4131 driver and ACL definitions.
4133 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4134 forgetting to use the resulting value; it was using the unexpanded value.
4136 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4137 hadn't been configured. The fix is from Juergen Kreileder, who
4138 understands it better than I do:
4140 "Here's what I see happening with three configured cyrus_sasl
4141 authenticators configured (plain, login, cram-md5):
4143 On startup auth_cyrus_sasl_init() gets called for each of these.
4144 This means three calls to sasl_listmech() without a specified mech_list.
4145 => SASL tests which mechs of all available mechs actually work
4146 => three warnings about OTP not working
4147 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4149 With the patch, sasl_listmech() also gets called three times. But now
4150 SASL's mech_list option is set to the server_mech specified in the the
4151 authenticator. Or in other words, the answer from sasl_listmech()
4152 gets limited to just the mech you're testing for (which is different
4154 => the return list contains just 'plain' or 'login', 'cram-md5' or
4155 nothing depending on the value of ob->server_mech.
4157 I've just tested the patch: Authentication still works fine,
4158 unavailable mechs specified in the exim configuration are still
4159 caught, and the auth.log warnings about OTP are gone."
4161 PH/25 When debugging is enabled, the contents of the command line are added
4162 to the debugging output, even when log_selector=+arguments is not
4165 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4166 answer is "GNU", and only if the return is "GNU/something" is the answer
4169 PH/27 $acl_verify_message is now set immediately after the failure of a
4170 verification in an ACL, and so is available in subsequent modifiers. In
4171 particular, the message can be preserved by coding like this:
4173 warn !verify = sender
4174 set acl_m0 = $acl_verify_message
4176 Previously, $acl_verify_message was set only while expanding "message"
4177 and "log_message" when a very denied access.
4179 PH/28 Modified OS/os.c-Linux with
4181 -#ifndef OS_LOAD_AVERAGE
4182 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4184 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4185 nomenclature these days.)
4187 PH/29 Installed patch from the Sieve maintainer that adds the options
4188 sieve_useraddress and sieve_subaddress to the redirect router.
4190 PH/30 In these circumstances:
4191 . Two addresses routed to the same list of hosts;
4192 . First host does not offer TLS;
4193 . First host accepts first address;
4194 . First host gives temporary error to second address;
4195 . Second host offers TLS and a TLS session is established;
4196 . Second host accepts second address.
4197 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4198 and peerdn, if requested) that were in fact used only for the second
4201 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4202 attention to any local part prefix or suffix that was matched by the
4203 router that accepted the address. It now behaves in the same way as it
4204 does for delivery: the affixes are removed from the local part unless
4205 rcpt_include_affixes is set on the transport.
4207 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4208 timeout during the DATA phase of an incoming message.
4210 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4211 applied a patch sanctioned by the Sieve maintainer.
4213 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4214 the uid or gid is negative. A case of a negative gid caused this to be
4215 noticed. The fix allows for either to be negative.
4217 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4218 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4221 PH/36 The change PH/12 above was broken. Fixed it.
4223 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4224 the grounds that routing the same address twice would always produce the
4225 same answer. This might have been true once, but it is certainly no
4226 longer true now. Routing a child address may depend on the previous
4227 routing that produced that child. Some complicated redirection strategies
4228 went wrong when messages had multiple recipients, and made Exim's
4229 behaviour dependent on the order in which the addresses were given.
4231 I have moved the duplicate checking until after the routing is complete.
4232 Exim scans the addresses that are assigned to local and remote
4233 transports, and removes any duplicates. This means that more work will be
4234 done, as duplicates will always all be routed, but duplicates are
4235 presumably rare, so I don't expect this is of any significance.
4237 For deliveries to pipes, files, and autoreplies, the duplicate checking
4238 still happens during the routing process, since they are not going to be
4241 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4242 It corrects a timeout issue with spamd. This is Ian's comment: "The
4243 background is that sometimes spamd either never reads data from a
4244 connection it has accepted, or it never writes response data. The exiscan
4245 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4246 blindly assumes that writes won't block so it may never time out."
4248 PH/39 Allow G after quota size as well as K and M.
4250 PH/40 The value set for $authenticated_id in an authenticator may not contain
4251 binary zeroes or newlines because the value is written to log lines and
4252 to spool files. There was no check on this. Now the value is run through
4253 the string_printing() function so that such characters are converted to
4254 printable escape sequences.
4256 PH/41 $message_linecount is a new variable that contains the total number of
4257 lines in the message. Compare $body_linecount, which is the count for the
4260 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4261 bounce and delay warning messages, except in certain special cases, which
4264 (a) An SMTP error message from a remote host;
4265 (b) A message specified in a :fail: redirection;
4266 (c) A message specified in a "fail" command in a system filter;
4267 (d) A message specified in a FAIL return from the queryprogram router;
4268 (e) A message specified by the cannot_route_message router option.
4270 In these cases only, Exim does include the error details in bounce and
4271 warning messages. There are also a few cases where bland messages such
4272 as "unrouteable address" or "local delivery error" are given.
4274 PH/43 $value is now also set for the "else" part of a ${run expansion.
4276 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4277 being worked on, but at least Exim now implements the latest version to
4280 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4281 process to complete was treated as a delivery failure, a timeout while
4282 writing the message to the pipe was logged, but erroneously treated as a
4283 successful delivery. Such timeouts include transport filter timeouts. For
4284 consistency with the overall process timeout, these timeouts are now
4285 treated as errors, giving rise to delivery failures by default. However,
4286 there is now a new Boolean option for the pipe transport called
4287 timeout_defer, which, if set TRUE, converts the failures into defers for
4288 both kinds of timeout. A transport filter timeout is now identified in
4291 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4292 systems where "make" and "gmake" are different, calling "gmake" at top
4293 level broke things. I've arranged for the value of $(MAKE) to be passed
4294 from the Makefile to this script so that it can call the same version of
4298 A note about Exim versions 4.44 and 4.50
4299 ----------------------------------------
4301 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4302 changes of various kinds. As a consequence, a big documentation update was
4303 needed. This delayed the release for rather longer than seemed good, especially
4304 in the light of a couple of (minor) security issues. Therefore, the changes
4305 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4306 release. So 4.44 and 4.50 are in effect two different branches that both start
4309 I have left the 4.50 change log unchanged; it contains all the changes since
4310 4.43. The change log for 4.44 is below; many of its items are identical to
4311 those for 4.50. This seems to be the most sensible way to preserve the
4312 historical information.
4318 1. Minor wording change to the doc/README.SIEVE file.
4320 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4321 computation of the current number of files was incorrect.
4323 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4324 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4327 4. Give more explanation in the error message when the command for a transport
4328 filter fails to execute.
4330 5. There are several places where Exim runs a non-Exim command in a
4331 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4332 was being done only for the command run by the queryprogram router. It is
4333 now done for all such subprocesses. The other cases are: ${run, transport
4334 filters, and the commands run by the lmtp and pipe transports.
4336 6. Added CONFIGURE_GROUP build-time option.
4338 7. Some older OS have a limit of 256 on the maximum number of file
4339 descriptors. Exim was using setrlimit() to set 1000 as a large value
4340 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4341 systems. I've change it so that if it can't get 1000, it tries for 256.
4343 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4344 was an oversight, and furthermore, ever since the addition of extra
4345 controls (e.g. 4.43/32), the checks on when to allow different forms of
4346 "control" were broken. There should now be diagnostics for all cases when a
4347 control that does not make sense is encountered.
4349 9. Added the /retain_sender option to "control=submission".
4351 10. $recipients is now available in the predata ACL (oversight).
4353 11. Tidy the search cache before the fork to do a delivery from a message
4354 received from the command line. Otherwise the child will trigger a lookup
4355 failure and thereby defer the delivery if it tries to use (for example) a
4356 cached ldap connection that the parent has called unbind on.
4358 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4359 of $address_data from the recipient verification was clobbered by the
4360 sender verification.
4362 13. The value of address_data from a sender verification is now available in
4363 $sender_address_data in subsequent conditions in the ACL statement.
4365 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4367 15. Added a new option "connect=<time>" to callout options, to set a different
4370 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4371 was its contents. (It was OK if the option was not defined at all.)
4373 17. A "Completed" log line is now written for messages that are removed from
4374 the spool by the -Mrm option.
4376 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4377 information about exactly what failed.
4379 19. Added -dd to debug only the daemon process.
4381 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4382 handles timeouts, both on the server side and network timeouts. Renamed the
4383 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4385 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4386 It is now set to "smtps".
4388 22. $host_address is now set to the target address during the checking of
4389 ignore_target_hosts.
4391 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4392 being passed; this would have caused $sender_host_name to have been used if
4393 matching the list had actually called for a host name (not very likely,
4394 since this list is usually IP addresses). A host name is now passed as
4397 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4398 code that specifies a non-restarting handler (typically sigaction() in
4399 modern systems) in an attempt to fix a rare and obscure crash bug.
4401 25. Narrowed the window for a race in the daemon that could cause it to ignore
4402 SIGCHLD signals. This is not a major problem, because they are used only to
4403 wake it up if nothing else does.
4405 26. A malformed maildirsize file could cause Exim to calculate negative values
4406 for the mailbox size or file count. Odd effects could occur as a result.
4407 The maildirsize information is now recalculated if the size or filecount
4410 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4411 support for a long time. Removed HAVE_SYS_VFS_H.
4413 28. Installed the latest version of exipick from John Jetmore.
4415 29. In an address list, if the pattern was not a regular expression, an empty
4416 subject address (from a bounce message) matched only if the pattern was an
4417 empty string. Non-empty patterns were not even tested. This was the wrong
4418 because it is perfectly reasonable to use an empty address as part of a
4419 database query. An empty address is now tested by patterns that are
4420 lookups. However, all the other forms of pattern expect the subject to
4421 contain a local part and a domain, and therefore, for them, an empty
4422 address still always fails if the pattern is not itself empty.
4424 30. Exim went into a mad DNS loop when attempting to do a callout where the
4425 host was specified on an smtp transport, and looking it up yielded more
4426 than one IP address.
4428 31. Re-factored the code for checking spool and log partition space into a
4429 function that finds that data and another that does the check. The former
4430 is then used to implement four new variables: $spool_space, $log_space,
4431 $spool_inodes, and $log_inodes.
4433 32. The RFC2047 encoding function was originally intended for short strings
4434 such as real names; it was not keeping to the 75-character limit for
4435 encoded words that the RFC imposes. It now respects the limit, and
4436 generates multiple encoded words if necessary. To be on the safe side, I
4437 have increased the buffer size for the ${rfc2047: expansion operator from
4440 33. It is now permitted to omit both strings after an "if" condition; if the
4441 condition is true, the result is "true". As before, when the second string
4442 is omitted, a false condition yields an empty string. This makes it less
4443 cumbersome to write custom ACL and router conditions.
4445 34. Failure to deliver a bounce message always caused it to be frozen, even if
4446 there was an errors_to setting on the router. The errors_to setting is now
4449 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4450 canonical form (fully expanded) before being placed in
4451 $sender_host_address.
4453 36. The table in the code that translates DNS record types into text (T_A to
4454 "A" for instance) was missing entries for NS and CNAME. It is just possible
4455 that this could have caused confusion if both these types were looked up
4456 for the same domain, because the text type is used as part of Exim's
4457 per-process caching. But the chance of anyone hitting this buglet seems
4460 37. The dnsdb lookup has been extended in a number of ways.
4462 (1) There is a new type, "zns", which walks up the domain tree until it
4463 finds some nameserver records. It should be used with care.
4465 (2) There is a new type, "mxh", which is like "mx" except that it returns
4466 just the host names, not the priorities.
4468 (3) It is now possible to give a list of domains (or IP addresses) to be
4469 looked up. The behaviour when one of the lookups defers can be
4470 controlled by a keyword.
4472 (4) It is now possible to specify the separator character for use when
4473 multiple records are returned.
4475 38. The dnslists ACL condition has been extended: it is now possible to supply
4476 a list of IP addresses and/or domains to be looked up in a particular DNS
4479 39. Added log_selector=+queue_time_overall.
4481 40. When running the queue in the test harness, wait just a tad after forking a
4482 delivery process, to get repeatability of debugging output.
4484 41. Include certificate and key file names in error message when GnuTLS fails
4485 to set them up, because the GnuTLS error message doesn't include the name
4486 of the failing file when there is a problem reading it.
4488 42. Allow both -bf and -bF in the same test run.
4490 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4492 44. The "Exiscan patch" is now merged into the mainline Exim source.
4494 45. Sometimes the final signoff response after QUIT could fail to get
4495 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4496 < 0 before doing a fflush(). This bug looks as though it goes back to the
4497 introduction of TLS in release 3.20, but "sometimes" must have been rare
4498 because the tests only now provoked it.
4500 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4501 (this can affect the format of dates).
4503 47. exim_tidydb, when checking for the continued existence of a message for
4504 which it has found a message-specific retry record, was not finding
4505 messages that were in split spool directories. Consequently, it was
4506 deleting retry records that should have stayed in existence.
4508 48. Steve fixed some bugs in eximstats.
4510 49. The SPA authentication driver was not abandoning authentication and moving
4511 on to the next authenticator when an expansion was forced to fail,
4512 contradicting the general specification for all authenticators. Instead it
4513 was generating a temporary error. It now behaves as specified.
4515 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4516 (the order specifies the preference for clients). The order is now AES256,
4517 AES128, 3DES, ARCFOUR128.
4519 51. Small patch to Sieve code - explicitly set From: when generating an
4522 52. Exim crashed if a remote delivery caused a very long error message to be
4523 recorded - for instance if somebody sent an entire SpamAssassin report back
4524 as a large number of 550 error lines. This bug was coincidentally fixed by
4525 increasing the size of one of Exim's internal buffers (big_buffer) that
4526 happened as part of the Exiscan merge. However, to be on the safe side, I
4527 have made the code more robust (and fixed the comments that describe what
4530 53. Now that there can be additional text after "Completed" in log lines (if
4531 the queue_time_overall log selector is set), a one-byte patch to exigrep
4532 was needed to allow it to recognize "Completed" as not the last thing in
4535 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4536 patch that reportedly fixes this has been added. I am not expert enough to
4537 create a test for it. This is what the patch creator wrote:
4539 "I found a little strange behaviour of ldap code when working with
4540 Windows 2003 AD Domain, where users was placed in more than one
4541 Organization Units. When I tried to give exim partial DN, the exit code
4542 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4543 But simultaneously result of request was absolutely normal ldap result,
4544 so I produce this patch..."
4546 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4547 REFERENCE, so I have modified the code to exclude the patch when that macro
4550 55. Some experimental protocols are using DNS PTR records for new purposes. The
4551 keys for these records are domain names, not reversed IP addresses. The
4552 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4553 leaves it alone. Component reversal etc. now happens only for IP addresses.
4556 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4558 57. Double the size of the debug message buffer (to 2048) so that more of very
4559 long debug lines gets shown.
4561 58. The exicyclog utility now does better if the number of log files to keep
4562 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4564 59. Two changes related to the smtp_active_hostname option:
4566 (1) $smtp_active_hostname is now available as a variable.
4567 (2) The default for smtp_banner uses $smtp_active_hostname instead
4568 of $primary_hostname.
4570 60. The host_aton() function is supposed to be passed a string that is known
4571 to be a valid IP address. However, in the case of IPv6 addresses, it was
4572 not checking this. This is a hostage to fortune. Exim now panics and dies
4573 if the condition is not met. A case was found where this could be provoked
4574 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4575 components; fortuitously, this particular loophole had already been fixed
4576 by change 4.50/55 above.
4578 If there are any other similar loopholes, the new check in host_aton()
4579 itself should stop them being exploited. The report I received stated that
4580 data on the command line could provoke the exploit when Exim was running as
4581 exim, but did not say which command line option was involved. All I could
4582 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4583 running as the user.
4586 61. There was a buffer overflow vulnerability in the SPA authentication code
4587 (which came originally from the Samba project). I have added a test to the
4588 spa_base64_to_bits() function which I hope fixes it.
4591 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4592 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4594 63. The daemon start-up calls getloadavg() while still root for those OS that
4595 need the first call to be done as root, but it missed one case: when
4596 deliver_queue_load_max is set with deliver_drop_privilege. This is
4597 necessary for the benefit of the queue runner, because there is no re-exec
4598 when deliver_drop_privilege is set.
4600 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4601 This has been fixed.
4603 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4604 was in use, was not putting the data itself into the right store pool;
4605 consequently, it could be overwritten for a subsequent message in the same
4606 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4609 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4611 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4612 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4613 the function were treating the return as a boolean value, which happened to
4614 work because 0=false and not-0=true, but is not correct code.
4616 68. The host_aton() function was not handling scoped IPv6 addresses (those
4617 with, for example, "%eth0" on the end) correctly.
4619 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4620 negated items (that is, ~something) in unsigned ints. Some compilers
4621 apparently mutter when there is no cast.
4623 70. If an address verification called from an ACL failed, and did not produce a
4624 user-specific message (i.e. there was only a "system" message), nothing was
4625 put in $acl_verify_message. In this situation, it now puts the system
4628 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4629 session; change 4.31/43 added the unwanted input to the log line - except
4630 that it did not do this in the start of session case. It now does.
4632 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4633 This is wrong and can cause the other end to generate a synchronization
4634 error if it is another Exim or anything else that does the synchronization
4635 check. A QUIT command is no longer sent after a timeout.
4637 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4638 during host lookups.
4640 74. The defer_ok option of callout verification was not working if it was used
4641 when verifying addresses in header lines, that is, for this case:
4643 verify = header_sender/callout=defer_ok
4645 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4646 those file descriptors could be used for SMTP connections. If anything
4647 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4648 could be sent to the SMTP client, causing chaos. The daemon now opens
4649 stdin, stdout, and stderr to /dev/null when it puts itself into the
4652 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4653 log by default. The user can override this with suitable Perl magic.
4655 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4656 the log message when discard triggers, was not working for the DATA ACL or
4657 for the non-SMTP ACL.
4659 78. Error message wording change in sieve.c.
4661 79. If smtp_accept_max_per_host was set, the number of connections could be
4662 restricted to fewer than expected, because the daemon was trying to set up
4663 a new connection before checking whether the processes handling previous
4664 connections had finished. The check for completed processes is now done
4665 earlier. On busy systems, this bug wouldn't be noticed because something
4666 else would have woken the daemon, and it would have reaped the completed
4669 80. If a message was submitted locally by a user whose login name contained one
4670 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4671 It caused a spool format error. I have fixed the spool reading code. A
4672 related problem was that the "from" clause in the Received: line became
4673 illegal because of the space(s). It is now covered by ${quote_local_part.
4675 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4678 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4680 83. Patch from Sieve maintainer to fix unterminated string problem in
4681 "vacation" handling.
4683 84. Some minor changes to the Linux configuration files to help with other
4684 OS variants using glibc.
4686 85. One more patch for Sieve to update vacation handling to latest spec.
4689 ----------------------------------------------------
4690 See the note above about the 4.44 and 4.50 releases.
4691 ----------------------------------------------------
4697 1. Change 4.43/35 introduced a bug that caused file counts to be
4698 incorrectly computed when quota_filecount was set in an appendfile
4701 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4702 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4705 3. Give more explanation in the error message when the command for a transport
4706 filter fails to execute.
4708 4. There are several places where Exim runs a non-Exim command in a
4709 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4710 was being done only for the command run by the queryprogram router. It is
4711 now done for all such subprocesses. The other cases are: ${run, transport
4712 filters, and the commands run by the lmtp and pipe transports.
4714 5. Some older OS have a limit of 256 on the maximum number of file
4715 descriptors. Exim was using setrlimit() to set 1000 as a large value
4716 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4717 systems. I've change it so that if it can't get 1000, it tries for 256.
4719 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4720 was an oversight, and furthermore, ever since the addition of extra
4721 controls (e.g. 4.43/32), the checks on when to allow different forms of
4722 "control" were broken. There should now be diagnostics for all cases when a
4723 control that does not make sense is encountered.
4725 7. $recipients is now available in the predata ACL (oversight).
4727 8. Tidy the search cache before the fork to do a delivery from a message
4728 received from the command line. Otherwise the child will trigger a lookup
4729 failure and thereby defer the delivery if it tries to use (for example) a
4730 cached ldap connection that the parent has called unbind on.
4732 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4733 of $address_data from the recipient verification was clobbered by the
4734 sender verification.
4736 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4737 was its contents. (It was OK if the option was not defined at all.)
4739 11. A "Completed" log line is now written for messages that are removed from
4740 the spool by the -Mrm option.
4742 12. $host_address is now set to the target address during the checking of
4743 ignore_target_hosts.
4745 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4746 being passed; this would have caused $sender_host_name to have been used if
4747 matching the list had actually called for a host name (not very likely,
4748 since this list is usually IP addresses). A host name is now passed as
4751 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4752 code that specifies a non-restarting handler (typically sigaction() in
4753 modern systems) in an attempt to fix a rare and obscure crash bug.
4755 15. Narrowed the window for a race in the daemon that could cause it to ignore
4756 SIGCHLD signals. This is not a major problem, because they are used only to
4757 wake it up if nothing else does.
4759 16. A malformed maildirsize file could cause Exim to calculate negative values
4760 for the mailbox size or file count. Odd effects could occur as a result.
4761 The maildirsize information is now recalculated if the size or filecount
4764 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4765 support for a long time. Removed HAVE_SYS_VFS_H.
4767 18. Updated exipick to current release from John Jetmore.
4769 19. Allow an empty sender to be matched against a lookup in an address list.
4770 Previously the only cases considered were a regular expression, or an
4773 20. Exim went into a mad DNS lookup loop when doing a callout where the
4774 host was specified on the transport, if the DNS lookup yielded more than
4777 21. The RFC2047 encoding function was originally intended for short strings
4778 such as real names; it was not keeping to the 75-character limit for
4779 encoded words that the RFC imposes. It now respects the limit, and
4780 generates multiple encoded words if necessary. To be on the safe side, I
4781 have increased the buffer size for the ${rfc2047: expansion operator from
4784 22. Failure to deliver a bounce message always caused it to be frozen, even if
4785 there was an errors_to setting on the router. The errors_to setting is now
4788 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4789 canonical form (fully expanded) before being placed in
4790 $sender_host_address.
4792 24. Updated eximstats to version 1.33
4794 25. Include certificate and key file names in error message when GnuTLS fails
4795 to set them up, because the GnuTLS error message doesn't include the name
4796 of the failing file when there is a problem reading it.
4798 26. Expand error message when OpenSSL has problems setting up cert/key files.
4801 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4802 (this can affect the format of dates).
4804 28. exim_tidydb, when checking for the continued existence of a message for
4805 which it has found a message-specific retry record, was not finding
4806 messages that were in split spool directories. Consequently, it was
4807 deleting retry records that should have stayed in existence.
4809 29. eximstats updated to version 1.35
4810 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4811 1.35 - bugfix such that pie charts by volume are generated correctly
4813 30. The SPA authentication driver was not abandoning authentication and moving
4814 on to the next authenticator when an expansion was forced to fail,
4815 contradicting the general specification for all authenticators. Instead it
4816 was generating a temporary error. It now behaves as specified.
4818 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4819 (the order specifies the preference for clients). The order is now AES256,
4820 AES128, 3DES, ARCFOUR128.
4822 31. Small patch to Sieve code - explicitly set From: when generating an
4825 32. Exim crashed if a remote delivery caused a very long error message to be
4826 recorded - for instance if somebody sent an entire SpamAssassin report back
4827 as a large number of 550 error lines. This bug was coincidentally fixed by
4828 increasing the size of one of Exim's internal buffers (big_buffer) that
4829 happened as part of the Exiscan merge. However, to be on the safe side, I
4830 have made the code more robust (and fixed the comments that describe what
4833 33. Some experimental protocols are using DNS PTR records for new purposes. The
4834 keys for these records are domain names, not reversed IP addresses. The
4835 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4836 leaves it alone. Component reversal etc. now happens only for IP addresses.
4839 34. The host_aton() function is supposed to be passed a string that is known
4840 to be a valid IP address. However, in the case of IPv6 addresses, it was
4841 not checking this. This is a hostage to fortune. Exim now panics and dies
4842 if the condition is not met. A case was found where this could be provoked
4843 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4844 components; fortuitously, this particular loophole had already been fixed
4845 by change 4.50/55 or 4.44/33 above.
4847 If there are any other similar loopholes, the new check in host_aton()
4848 itself should stop them being exploited. The report I received stated that
4849 data on the command line could provoke the exploit when Exim was running as
4850 exim, but did not say which command line option was involved. All I could
4851 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4852 running as the user.
4855 35. There was a buffer overflow vulnerability in the SPA authentication code
4856 (which came originally from the Samba project). I have added a test to the
4857 spa_base64_to_bits() function which I hope fixes it.
4860 36. The daemon start-up calls getloadavg() while still root for those OS that
4861 need the first call to be done as root, but it missed one case: when
4862 deliver_queue_load_max is set with deliver_drop_privilege. This is
4863 necessary for the benefit of the queue runner, because there is no re-exec
4864 when deliver_drop_privilege is set.
4866 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4867 was in use, was not putting the data itself into the right store pool;
4868 consequently, it could be overwritten for a subsequent message in the same
4869 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4872 38. Sometimes the final signoff response after QUIT could fail to get
4873 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4874 < 0 before doing a fflush(). This bug looks as though it goes back to the
4875 introduction of TLS in release 3.20, but "sometimes" must have been rare
4876 because the tests only now provoked it.
4882 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4883 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4884 Now it returns an integer. A number of calls were still expecting a T/F
4885 return. Fortuitously, in all cases, the tests worked in OK situations,
4886 which is the norm. However, things would have gone wrong on any write
4887 failures on the smtp file descriptor. This function is used when sending
4888 messages over SMTP and also when doing verify callouts.
4890 2. When Exim is called to do synchronous delivery of a locally submitted
4891 message (the -odf or -odi options), it no longer closes stderr before doing
4894 3. Implemented the mua_wrapper option.
4896 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4898 5. Implemented the functions header_remove(), header_testname(),
4899 header_add_at_position(), and receive_remove_recipient(), and exported them
4902 6. If an ACL "warn" statement specified the addition of headers, Exim already
4903 inserted X-ACL-Warn: at the start if there was no header name. However, it
4904 was not making this test for the second and subsequent header lines if
4905 there were newlines in the string. This meant that an invalid header could
4906 be inserted if Exim was badly configured.
4908 7. Allow an ACL "warn" statement to add header lines at the start or after all
4909 the Received: headers, as well as at the end.
4911 8. Added the rcpt_4xx retry error code.
4913 9. Added postmaster_mailfrom=xxx to callout verification option.
4915 10. Added mailfrom=xxxx to the callout verification option, for verify=
4918 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4919 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4921 12. Inserted some casts to stop certain compilers warning when using pointer
4922 differences as field lengths or precisions in printf-type calls (mostly
4923 affecting debugging statements).
4925 13. Added optional readline() support for -be (dynamically loaded).
4927 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4928 same clock tick as a message's arrival, so that its received time was the
4929 same as the "first fail" time on the retry record, and that message
4930 remained on the queue past the ultimate address timeout, every queue runner
4931 would try a delivery (because it was past the ultimate address timeout) but
4932 after another failure, the ultimate address timeout, which should have then
4933 bounced the address, did not kick in. This was a "< instead of <=" error;
4934 in most cases the first failure would have been in the next clock tick
4935 after the received time, and all would be well.
4937 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4938 being recognized when the domain list was tested by the match_domain
4939 condition in an expansion string.
4941 16. Added the ${str2b64: operator.
4943 17. Exim was always calling setrlimit() to set a large limit for the number of
4944 processes, without checking whether the existing limit was already
4945 adequate. (It did check for the limit on file descriptors.) Furthermore,
4946 errors from getrlimit() and setrlimit() were being ignored. Now they are
4947 logged to the main and panic logs, but Exim does carry on, to try to do its
4948 job under whatever limits there are.
4950 18. Imported PCRE 5.0.
4952 19. Trivial typo in log message " temporarily refused connection" (the leading
4955 20. If the log selector return_path_on_delivery was set and an address was
4956 redirected to /dev/null, the delivery process crashed because it assumed
4957 that a return path would always be set for a "successful" delivery. In this
4958 case, the whole delivery is bypassed as an optimization, and therefore no
4961 21. Internal re-arrangement: the function for sending a challenge and reading
4962 a response while authentication was assuming a zero-terminated challenge
4963 string. It's now changed to take a pointer and a length, to allow for
4964 binary data in such strings.
4966 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4968 23. Exim was not respecting finduser_retries when seeking the login of the
4969 uid under which it was called; it was always trying 10 times. (The default
4970 setting of finduser_retries is zero.) Also, it was sleeping after the final
4971 failure, which is pointless.
4973 24. Implemented tls_on_connect_ports.
4975 25. Implemented acl_smtp_predata.
4977 26. If the domain in control=submission is set empty, Exim assumes that the
4978 authenticated id is a complete email address when it generates From: or
4979 Sender: header lines.
4981 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4982 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4983 chown and chgrp in /bin and hostname in /usr/bin.
4985 28. Exim was keeping the "process log" file open after each use, just as it
4986 does for the main log. This opens the possibility of it remaining open for
4987 long periods when the USR1 signal hits a daemon. Occasional processlog
4988 errors were reported, that could have been caused by this. Anyway, it seems
4989 much more sensible not to leave this file open at all, so that is what now
4992 29. The long-running daemon process does not normally write to the log once it
4993 has entered its main loop, and it closes the log before doing so. This is
4994 so that log files can straightforwardly be renamed and moved. However,
4995 there are a couple of unusual error situations where the daemon does write
4996 log entries, and I had neglected to close the log afterwards.
4998 30. The text of an SMTP error response that was received during a remote
4999 delivery was being truncated at 512 bytes. This is too short for some of
5000 the long messages that one sometimes sees. I've increased the limit to
5003 31. It is now possible to make retry rules that apply only when a message has a
5004 specific sender, in particular, an empty sender.
5006 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5007 it possible to be selective about when SMTP synchronization is enforced.
5009 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5011 32. Implemented hosts_connection_nolog.
5013 33. Added an ACL for QUIT.
5015 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5018 35. Added mailbox_size and mailbox_filecount to appendfile.
5020 36. Added control = no_multiline_responses to ACLs.
5022 37. There was a bug in the logic of the code that waits for the clock to tick
5023 in the case where the clock went backwards by a substantial amount such
5024 that the microsecond fraction of "now" was more than the microsecond
5025 fraction of "then" (but the whole seconds number was less).
5027 38. Added support for the libradius Radius client library this is found on
5028 FreeBSD (previously only the radiusclient library was supported).
5034 1. When certain lookups returned multiple values in the form name=value, the
5035 quoting of the values was not always being done properly. Specifically:
5036 (a) If the value started with a double quote, but contained no whitespace,
5038 (b) If the value contained whitespace other than a space character (i.e.
5039 tabs or newlines or carriage returns) it was not quoted.
5040 This fix has been applied to the mysql and pgsql lookups by writing a
5041 separate quoting function and calling it from the lookup code. The fix
5042 should probably also be applied to nisplus, ibase and oracle lookups, but
5043 since I cannot test any of those, I have not disturbed their existing code.
5045 2. A hit in the callout cache for a specific address caused a log line with no
5046 reason for rejecting RCPT. Now it says "Previous (cached) callout
5047 verification failure".
5049 3. There was an off-by-one bug in the queryprogram router. An over-long
5050 return line was truncated at 256 instead of 255 characters, thereby
5051 overflowing its buffer with the terminating zero. As well as fixing this, I
5052 have increased the buffer size to 1024 (and made a note to document this).
5054 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5055 when Exim is waiting for an SMTP response from a remote server, Exim
5056 restarts its select() call on the socket, thereby resetting its timeout.
5057 This is not a problem when such interrupts are rare. Somebody set up a cron
5058 job to run exiwhat every 2 minutes, which is less than the normal select()
5059 timeout (5 or 10 minutes). This meant that the select() timeout never
5060 kicked in because it was always reset. I have fixed this by comparing the
5061 time when an interrupt arrives with the time at the start of the first call
5062 to select(). If more time than the timeout has elapsed, the interrupt is
5063 treated as a timeout.
5065 5. Some internal re-factoring in preparation for the addition of Sieve
5066 extensions (by MH). In particular, the "personal" test is moved to a
5067 separate function, and given an option for scanning Cc: and Bcc: (which is
5068 not set for Exim filters).
5070 6. When Exim created an email address using the login of the caller as the
5071 local part (e.g. when creating a From: or Sender: header line), it was not
5072 quoting the local part when it contained special characters such as @.
5074 7. Installed new OpenBSD configuration files.
5076 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5077 try to make them clearer.
5079 9. Callout options, other than the timeout value, were being ignored when
5080 verifying sender addresses in header lines. For example, when using
5082 verify = header_sender/callout=no_cache
5084 the cache was (incorrectly) being used.
5086 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5087 only the Cygwin environment.
5089 11. When return_path_on_delivery was set as a log selector, if different remote
5090 addresses in the same message used different return paths and parallel
5091 remote delivery occurred, the wrong values would sometimes be logged.
5092 (Whenever a remote delivery process finished, the return path value from
5093 the most recently started remote delivery process was logged.)
5095 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5096 lines when AUTH and/or TLS are in use. This is the "received protocol"
5097 field. Exim used to use "asmtp" for authenticated SMTP, without any
5098 indication (in the protocol name) for TLS use. Now it follows the RFC and
5099 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5100 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5101 names appear in log lines as well as in Received: header lines.
5103 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5104 extensions, and comparison tests, and to fix some bugs.
5106 14. Changes to the "personal" filter test:
5108 (1) The test was buggy in that it was just doing the equivalent of
5109 "contains" tests on header lines. For example, if a user's address was
5110 anne@some.where, the "personal" test would incorrectly be true for
5112 To: susanne@some.where
5114 This test is now done by extracting each address from the header in turn,
5115 and checking the entire address. Other tests that are part of "personal"
5116 are now done using regular expressions (for example, to check local parts
5117 of addresses in From: header lines).
5119 (2) The list of non-personal local parts in From: addresses has been
5120 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5121 taken from the Sieve specification recommendations.
5123 (3) If the message contains any header line starting with "List-" it is
5124 treated as non-personal.
5126 (4) The test for "circular" in the Subject: header line has been removed
5127 because it now seems ill-conceived.
5129 15. Minor typos in src/EDITME comments corrected.
5131 16. Installed latest exipick from John Jetmore.
5133 17. If headers_add on a router specified a text string that was too long for
5134 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5135 of string_sprintf() is now avoided.
5137 18. $message_body_size was not set (it was always zero) when running the DATA
5138 ACL and the local_scan() function.
5140 19. For the "mail" command in an Exim filter, no default was being set for
5141 the once_repeat time, causing a random time value to be used if "once" was
5142 specified. (If the value happened to be <= 0, no repeat happened.) The
5143 default is now 0s, meaning "never repeat". The "vacation" command was OK
5144 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5145 (I found it when inspecting the code).
5147 20. There is now an overall timeout for performing a callout verification. It
5148 defaults to 4 times the callout timeout, which applies to individual SMTP
5149 commands during the callout. The overall timeout applies when there is more
5150 than one host that can be tried. The timeout is checked before trying the
5151 next host. This prevents very long delays if there are a large number of
5152 hosts and all are timing out (e.g. when the network connections are timing
5153 out). The value of the overall timeout can be changed by specifying an
5154 additional sub-option for "callout", called "maxwait". For example:
5156 verify = sender/callout=5s,maxwait=20s
5158 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5159 to the end before writing, but this should make it even safer).
5161 22. Exim was forgetting that it had advertised PIPELINING for the second and
5162 subsequent messages on an SMTP connection. It was also not resetting its
5163 memory on STARTTLS and an internal HELO.
5165 23. When Exim logs an SMTP synchronization error within a session, it now
5166 records whether PIPELINING has been advertised or not.
5168 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5169 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5170 rather than long int.
5172 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5174 26. Added the never_mail option to autoreply.
5180 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5181 crash if the getsockname() call failed; this can happen if a connection is
5182 closed very soon after it is established. The problem was simply in the
5183 order in which certain operations were done, causing Exim to try to write
5184 to the SMTP stream before it had set up the file descriptor. The bug has
5185 been fixed by making things happen in the correct order.
5191 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5192 before the connection was closed, thus losing the rejection response.
5194 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5195 some early Solaris releases, but causes trouble in current releases where
5196 socklen_t is defined.
5198 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5201 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5204 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5205 that releasing the top of it at the end releases what was used for sub-
5206 expansions (unless the block got too big). However, discard this block if
5207 the first thing is a variable or header, so that we can use its block when
5208 it is dynamic (useful for very large $message_headers, for example).
5210 6. Lookups now cache *every* query, not just the most recent. A new, separate
5211 store pool is used for this. It can be recovered when all lookup caches are
5212 flushed. Lookups now release memory at the end of their result strings.
5213 This has involved some general refactoring of the lookup sources.
5215 7. Some code has been added to the store_xxx() functions to reduce the amount
5216 of flapping under certain conditions.
5218 8. log_incoming_interface used to affect only the <= reception log lines. Now
5219 it causes the local interface and port to be added to several more SMTP log
5220 lines, for example "SMTP connection from", and rejection lines.
5222 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5224 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5226 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5227 could be overwritten at the end of the current message (or the start of a
5228 new message if it was set in a HELO ACL). The value is now preserved for
5229 the duration of the SMTP connection.
5231 12. If a transport had a headers_rewrite setting, and a matching header line
5232 contained an unqualified address, that address was qualified, even if it
5233 did not match any rewriting rules. The underlying bug was that the values
5234 of the flags that permit the existence of unqualified sender and recipient
5235 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5236 non-local messages, and by -bnq for local messages) were not being
5237 preserved with the message after it was received.
5239 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5240 "next input=" as part of the text comprising the host identity instead of
5241 the correct text. The code was using the same buffer for two different
5242 strings. However, depending on which order the printing function evaluated
5243 its arguments, the bug did not always show up. Under Linux, for example, my
5244 test suite worked just fine.
5246 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5247 doesn't work with some older versions of Perl. It has been changed to "my",
5248 which in any case is probably the better facility to use.
5250 15. A really picky compiler found some instances of statements for creating
5251 error messages that either had too many or two few arguments for the format
5254 16. The size of the buffer for calls to the DNS resolver has been increased
5255 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5256 for addresses that have a lot of PTR records. This alleviates a problem; it
5257 does not fully solve it.
5259 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5260 buffer now truncates the list and logs the incident, which is the same
5261 action as happens when Exim is looking up a host name and its aliases.
5262 Previously in this situation something unpredictable would happen;
5263 sometimes it was "internal error: store_reset failed".
5265 18. If a server dropped the connection unexpectedly when an Exim client was
5266 using GnuTLS and trying to read a response, the client delivery process
5267 crashed while trying to generate an error log message.
5269 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5270 string, for example:
5272 warn message = H1: something\nH2: something
5274 the text was added as a single header line from Exim's point of view
5275 though it ended up OK in the delivered message. However, searching for the
5276 second and subsequent header lines using $h_h2: did not work. This has been
5277 fixed. Similarly, if a system filter added multiple headers in this way,
5278 the routers could not see them.
5280 20. Expanded the error message when iplsearch is called with an invalid key to
5281 suggest using net-iplsearch in a host list.
5283 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5284 ACLs are no longer actually imposed (and a message to that effect is
5287 22. If a "gecos" field in a passwd entry contained escaped characters, in
5288 particular, if it contained a \" sequence, Exim got it wrong when building
5289 a From: or a Sender: header from that name. A second bug also caused
5290 incorrect handling when an unquoted " was present following a character
5291 that needed quoting.
5293 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5294 was not being matched caselessly.
5296 24. Arranged for all hyphens in the exim.8 source to be escaped with
5299 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5300 properly. Recipient callout cache records were still being keyed to include
5301 the sender, even when use_sender was set false. This led to far more
5302 callouts that were necessary. The sender is no longer included in the key
5303 when use_sender is false.
5305 26. Added "control = submission" modifier to ACLs.
5307 27. Added the ${base62d: operator to decode base 62 numbers.
5309 28. dnsdb lookups can now access SRV records.
5311 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5312 the configuration file.
5314 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5315 (-v) mode. This makes the output for a verbose queue run more intelligible.
5317 31. Added a use_postmaster feature to recipient callouts.
5319 32. Added the $body_zerocount variable, containing the number of binary zero
5320 bytes in the message body.
5322 33. The time of last modification of the "new" subdirectory is now used as the
5323 "mailbox time last read" when there is a quota error for a maildir
5326 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5328 35. Added +ignore_unknown as a special item in host lists.
5330 36. Code for decoding IPv6 addresses in host lists is now included, even if
5331 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5332 address was recognized as an IP address, but was then not correctly decoded
5333 into binary, causing unexpected and incorrect effects when compared with
5340 1. Very minor rewording of debugging text in manualroute to say "list of
5341 hosts" instead of "hostlist".
5343 2. If verify=header_syntax was set, and a header line with an unqualified
5344 address (no domain) and a large number of spaces between the end of the
5345 name and the colon was received, the reception process suffered a buffer
5346 overflow, and (when I tested it) crashed. This was caused by some obsolete
5347 code that should have been removed. The fix is to remove it!
5349 3. When running in the test harness, delay a bit after writing a bounce
5350 message to get a bit more predictability in the log output.
5352 4. Added a call to search_tidyup() just before forking a reception process. In
5353 theory, someone could use a lookup in the expansion of smtp_accept_max_
5354 per_host which, without the tidyup, could leave open a database connection.
5356 5. Added the variables $recipient_data and $sender_data which get set from a
5357 lookup success in an ACL "recipients" or "senders" condition, or a router
5358 "senders" option, similar to $domain_data and $local_part_data.
5360 6. Moved the writing of debug_print from before to after the "senders" test
5363 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5364 problems for message scanning, either using a data ACL, or using
5365 local_scan() because the Received: header was not generated till after they
5366 were called (in order to set the time as the time of reception completion).
5367 I have revised the way this works. The header is now generated after the
5368 body is received, but before the ACL or local_scan() are called. After they
5369 are run, the timestamp in the header is updated.
5375 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5376 before starting a queue runner without re-exec. This happened only when
5377 deliver_drop_privilege was set or when the Exim user was set to root. The
5378 effect of the bug was that timeouts during subsequent deliveries caused
5379 crashes instead of being properly handled. The handler is now left at its
5380 default (and expected) setting.
5382 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5383 message, again when deliver_drop_privilege is set or Exim is run as root.
5384 The bug described in (1) was not present in this case, but the tidying up
5385 of the other signals was missing. I have made the two cases consistent.
5387 3. The ignore_target_hosts setting on a manualroute router was being ignored
5388 for hosts that were looked up using the /MX notation.
5390 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5393 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5394 operated on the sender address. After changing the $sender_address to <>
5395 for the sender address verify, Exim was re-instated it as the original
5396 (before rewriting) address, but remembering that it had rewritten it, so it
5397 wasn't rewriting it again. This bug also had the effect of breaking the
5398 sender address verification caching when the sender address was rewritten.
5400 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5401 This has been changed so that if the ip literal address matches
5402 ignore_target_hosts, the router declines.
5404 7. Added expansion conditions match_domain, match_address, and match_local_
5405 part (NOT match_host).
5407 8. The placeholder for the Received: header didn't have a length field set.
5409 9. Added code to Exim itself and to exim_lock to test for a specific race
5410 condition that could lead to file corruption when using MBX delivery. The
5411 issue is with the lockfile that is created in /tmp. If this file is removed
5412 after a process has opened it but before that process has acquired a lock,
5413 there is the potential for a second process to recreate the file and also
5414 acquire a lock. This could lead to two Exim processes writing to the file
5415 at the same time. The added code performs the same test as UW imapd; it
5416 checks after acquiring the lock that its file descriptor still refers to
5417 the same named file.
5419 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5420 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5423 11. Added the smtp_active_hostname option. If used, this will typically be made
5424 to depend on the incoming interface address. Because $interface_address is
5425 not set up until the daemon has forked a reception process, error responses
5426 that can happen earlier (such as "too many connections") no longer contain
5429 12. If an expansion in a condition on a "warn" statement fails because a lookup
5430 defers, the "warn" statement is abandoned, and the next ACL statement is
5431 processed. Previously this caused the whole ACL to be aborted.
5433 13. Added the iplsearch lookup type.
5435 14. Added ident_timeout as a log selector.
5437 15. Added tls_certificate_verified as a log selector.
5439 16. Added a global option tls_require_ciphers (compare the smtp transport
5440 option of the same name). This controls incoming TLS connections.
5442 17. I finally figured out how to make tls_require_ciphers do a similar thing
5443 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5444 before starting the TLS session.
5446 18. Tabs are now shown as \t in -bP output.
5448 19. If the log selector return_path_on_delivery was set, Exim crashed when
5449 bouncing a message because it had too many Received: header lines.
5451 20. If two routers both had headers_remove settings, and the first one included
5452 a superfluous trailing colon, the final name in the first list and the
5453 first name in the second list were incorrectly joined into one item (with a
5454 colon in the middle).
5460 1. Added -C and -D options to the exinext utility, mainly to make it easier
5461 to include in the automated testing, but these could be helpful when
5462 multiple configurations are in use.
5464 2. The exinext utility was not formatting the output nicely when there was
5465 an alternate port involved in the retry record key, nor when there was a
5466 message id as well (for retries that were specific to a specific message
5467 and a specific host). It was also confused by IPv6 addresses, because of
5468 the additional colons they contain. I have fixed the IPv4 problem, and
5469 patched it up to do a reasonable job for IPv6.
5471 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5472 delivery, the log line now contains "pipelined" if PIPELINING was used.
5474 4. An SMTP transport process used to panic and die if the bind() call to set
5475 an explicit outgoing interface failed. This has been changed; it is now
5476 treated in the same way as a connect() failure.
5478 5. A reference to $sender_host_name in the part of a conditional expansion
5479 that was being skipped was still causing a DNS lookup. This no longer
5482 6. The def: expansion condition was not recognizing references to header lines
5483 that used bh_ and bheader_.
5485 7. Added the _cache feature to named lists.
5487 8. The code for checking quota_filecount in the appendfile transport was
5488 allowing one more file than it should have been.
5490 9. For compatibility with Sendmail, the command line option
5498 and sets the incoming protocol and host name (for trusted callers). The
5499 host name and its colon can be omitted when only the protocol is to be set.
5500 Note the Exim already has two private options, -pd and -ps, that refer to
5501 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5502 "s", but I don't think that's a major issue.
5504 10. A number of refactoring changes to the code, none of which should affect
5507 (a) The number of logging options was getting close to filling up the
5508 32-bit word that was used as a bit map. I have split them into two classes:
5509 those that are passed in the argument to log_write(), and those that are
5510 only ever tested independently outside of that function. These are now in
5511 separate 32-bit words, so there is plenty of room for expansion again.
5512 There is no change in the user interface or the logging behaviour.
5514 (b) When building, for example, log lines, the code previously used a
5515 macro that called string_cat() twice, in order to add two strings. This is
5516 not really sufficiently general. Furthermore, there was one instance where
5517 it was actually wrong because one of the argument was used twice, and in
5518 one call a function was used. (As it happened, calling the function twice
5519 did not affect the overall behaviour.) The macro has been replaced by a
5520 function that can join an arbitrary number of extra strings onto a growing
5523 (c) The code for expansion conditions now uses a table and a binary chop
5524 instead of a serial search (which was left over from when there were very
5525 few conditions). Also, it now recognizes conditions like "pam" even when
5526 the relevant support is not compiled in: a suitably worded error message is
5527 given if an attempt is made to use such a condition.
5529 11. Added ${time_interval:xxxxx}.
5531 12. A bug was causing one of the ddress fields not to be passed back correctly
5532 from remote delivery subprocesses. The field in question was not being
5533 subsequently used, so this caused to problems in practice.
5535 13. Added new log selectors queue_time and deliver_time.
5537 14. Might have fixed a bug in maildirsizefile handling that threw up
5538 "unexpected character" debug warnings, and recalculated the data
5539 unnecessarily. In any case, I expanded the warning message to give more
5542 15. Added the message "Restricted characters in address" to the statements in
5543 the default ACL that block characters like @ and % in local parts.
5545 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5546 Three changes have been made:
5548 (a) There was a serious bug; a negative response to MAIL caused the whole
5549 recipient domain to be cached as invalid, thereby blocking all messages
5550 to all local parts at the same domain, from all senders. This bug has
5551 been fixed. The domain is no longer cached after a negative response to
5552 MAIL if the sender used is not empty.
5554 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5557 (c) A new callout option, "use_sender" has been added for people who want
5558 the modified behaviour.
5564 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5567 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5568 indeed breaks things for older releases.
5570 3. Added additional logging to the case where there is a problem reading data
5571 from a filter that is running in a subprocess using a pipe, in order to
5572 try to track down a specific problem.
5574 4. Testing facility fudge: when running in the test harness and attempting
5575 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5576 getting "No route to host". Convert this to a timeout.
5578 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5581 6. Some OS don't have socklen_t but use size_t instead. This affects the
5582 fifth argument of getsockopt() amongst other things. This is now
5583 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5584 can be set for individual OS. I have set it for SunOS5, OSF1, and
5585 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5586 some earlier ones do not.
5588 7. Change 4.30/15 was not doing the test caselessly.
5590 8. The standard form for an IPv6 address literal was being rejected by address
5591 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5592 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5593 this, as well as the form without the "IPv6" on the front (but only when
5594 address literals are enabled, of course).
5596 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5598 10. Exim crashed if a message with an empty sender address specified by -f
5599 encountered a router with an errors_to setting. This could be provoked only
5600 by a command such as
5604 where an empty string was supplied; "<>" did not hit this bug.
5606 11. Installed PCRE release 4.5.
5608 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5609 remained set. It is now erased.
5611 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5612 times from message ids (which are base 36 rather than the normal 62).
5614 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5615 were being counted as actual protocol errors, and logged if the log
5616 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5617 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5618 and DATA following a set of rejected RCPTs do not count as protocol errors.
5619 In other words, Exim assumes they were pipelined, though this may not
5620 actually be the case. Of course, in all cases the client gets an
5621 appropriate error code.
5623 15. If a lookup fails in an ACL condition, a message about the failure may
5624 be available; it is used if testing the ACL cannot continue, because most
5625 such messages specify what the cause of the deferral is. However, some
5626 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5627 that caused an old message to be retained and used if a later statement
5628 caused a defer, replacing the real cause of the deferral.
5630 16. If an IP address had so many PTR records that the DNS lookup buffer
5631 was not large enough to hold them, Exim could crash while trying to process
5632 the truncated data. It now detects and logs this case.
5634 17. Further to 4.21/58, another change has been made: if (and only if) the
5635 first line of a message (the first header line) ends with CRLF, a bare LF
5636 in a subsequent header line has a space inserted after it, so as not to
5637 terminate the header.
5639 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5640 unnecessarily, used atoi() instead of strtol(), and didn't check the
5641 termination when getting file sizes from file names by regex.
5643 19. Completely re-implemented the support for maildirsize files, in the light
5644 of a number of problems with the previous contributed implementation
5645 (4.30/29). In particular:
5647 . If the quota is zero, the maildirsize file is maintained, but no quota is
5650 . If the maildir directory does not exist, it is created before any attempt
5651 to write a maildirsize file.
5653 . The quota value in the file is just a cache; if the quota is changed in
5654 the transport, the new value overrides.
5656 . A regular expression is available for excluding directories from the
5659 20. The autoreply transport checks the characters in options that define the
5660 message's headers; it allows continued headers, but it was checking with
5661 isspace() after an embedded newline instead of explicitly looking for a
5664 21. If all the "regular" hosts to which an address was routed had passed their
5665 expiry times, and had not reached their retry times, the address was
5666 bounced, even if fallback hosts were defined. Now Exim should go on to try
5669 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5670 equivalent code in the SMTP transport. Some hosts send humungous responses
5671 to HELO/EHLO, more than 1024 it seems.
5673 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5674 gives compiler warnings in some environments. I've now done it "properly",
5677 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5678 (because of problems with the built-in one) was declared to return uschar *
5679 instead of char *, causing compiler failure.
5681 25. Fixed a file descriptor leak when processing alias/forward files.
5683 26. Fixed a minor format string issue in dbfn.c.
5685 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5687 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5688 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5691 29. When a sender address is verified, it is cached, to save repeating the test
5692 when there is more than one recipient in a message. However, when the
5693 verification involves a callout, it is possible for different callout
5694 options to be set for different recipients. It is too complicated to keep
5695 track of this in the cache, so now Exim always runs a verification when a
5696 callout is required, relying on the callout cache for the optimization.
5697 The overhead is duplication of the address routing, but this should not be
5700 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5701 to be verified with callout=postmaster, and the main callout worked but the
5702 postmaster check failed, the verification correctly failed. However, if a
5703 subsequent RCPT command asked for sender verification *without* the
5704 postmaster check, incorrect caching caused this verification also to fail,
5707 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5708 it was not caching the DNS options (qualify_single, search_parents) that
5709 were used when the lookup failed. A subsequent lookup with different
5710 options therefore always gave the same answer, though there were cases
5711 where it should not have. (Example: a "domains = !$mx_any" option on a
5712 dnslookup router: the "domains" option is always processed without any
5713 widening, but the router might have qualify_single set.) Now Exim uses the
5714 cached value only when the same options are set.
5716 32. Added John Jetmore's "exipick" utility to the distribution.
5718 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5719 than a timeout (e.g. a certificate is required, and is not provided), an
5720 Exim server now closes the connection immediately. Previously it waited for
5721 the client to close - but if the client is SSL, it seems that they each
5722 wait for each other, leading to a delay before one of them times out.
5724 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5725 maintained 0.8.x compatibility because I don't think many are using it, and
5726 it is clearly obsolete.
5728 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5731 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5732 client certificate was expired. A simple patch fixes this, though I don't
5733 understand the full logic of why the verify callback is called multiple
5736 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5737 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5738 which causes problems with some clients (such as the Certicom SSL Plus
5739 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5740 disables the coutermeasure allowing Eudora to connect."
5742 38. Exim was not checking that a write() to a log file succeeded. This could
5743 lead to Bad Things if a log got too big, in particular if it hit a file
5744 size limit. Exim now panics and dies if it cannot write to a log file, just
5745 as it does if it cannot open a log file.
5747 39. Modified OS/Makefile-Linux so that it now contains
5749 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5751 The two -D definitions ensure that Exim is compiled with large file
5752 support, which makes it possible to handle log files that are bigger than
5755 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5756 instance) a domain was checked against a named list that involved a lookup,
5757 causing $domain_data to be set, then another domain was checked against the
5758 same list, then the first domain was re-checked, the value of $domain_data
5759 after the final check could be wrong. In particular, if the second check
5760 failed, it could be set empty. This bug probably also applied to
5763 41. The strip_trailing_dot option was not being applied to the address given
5764 with the -f command-line option.
5766 42. The code for reading a message's header from the spool was incrementing
5767 $received_count, but never initializing it. This meant that the value was
5768 incorrect (doubled) while delivering a message in the same process in which
5769 it was received. In the most common configuration of Exim, this never
5770 happens - a fresh exec is done - but it can happen when
5771 deliver_drop_privilege is set.
5773 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5774 it now includes up to 150 characters of the unexpected data in the log
5777 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5778 and building data strings. The size of both of these buffers was 10 000
5779 bytes - far larger than anybody would *ever* want, thought I. Needless to
5780 say, somebody hit the limit. I have increased the maximum line length to
5781 20 000 and the maximum data length of concatenated lines to 100 000. I have
5782 also fixed two bugs, because there was no checking on these buffers. Tsk,
5783 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5784 buffer is too small.
5786 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5787 lsearch lookups. Now it does.
5789 46. When parsing a route_list item in a manualroute router, a fixed-length
5790 buffer was used for the list of hosts. I made this 1024 bytes long,
5791 thinking that nobody would ever have a list of hosts that long. Wrong.
5792 Somebody had a whole pile of complicated expansion conditions, and the
5793 string was silently truncated, leading to an expansion error. It turns out
5794 that it is easier to change to an unlimited length (owing to other changes
5795 that have happened since this code was originally written) than to build
5796 structure for giving a limitation error. The length of the item that
5797 expands into the list of hosts is now unlimited.
5799 47. The lsearch lookup could not handle data where the length of text line was
5800 more than 4095 characters. Such lines were truncated, leading to shortened
5801 data being returned. It should now handle lines of any length.
5803 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5804 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5807 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5808 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5809 no longer happens in comment lines. A list of replacements is now placed
5810 at the head of all of the source files, except those whose only change is
5811 to replace PERL_COMMAND in the very first #! line.
5813 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5814 messages on the queue, with a bottom-up merge sort, using code contributed
5815 by Michael Haardt. This should make operations like -bp somewhat faster on
5816 large queues. It won't affect queue runners, except when queue_run_in_order
5819 51. Installed eximstats 1.31 in the distribution.
5821 52. Added support for SRV lookups to the dnslookup router.
5823 53. If an ACL referred to $message_body or $message_body_end, the value was not
5824 reset for any messages that followed in the same SMTP session.
5826 54. The store-handling optimization for building very long strings was not
5827 differentiating between the different store pools. I don't think this
5828 actually made any difference in practice, but I've tidied it.
5830 55. While running the routers to verify a sender address, $sender_address
5831 was still set to the sender address. This is wrong, because when routing to
5832 send a bounce to the sender, it would be empty. Therefore, I have changed
5833 it so that, while verifying a sender address, $sender_address is set to <>.
5834 (There is no change to what happens when verifying a recipient address.)
5836 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5837 target A or AAAA records (if not already returned) without resetting the
5838 qualify_single or search_parents options of the DNS resolver. These are
5839 inappropriate in this case because the targets of MX and SRV records must
5840 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5841 target that, when qualified, matched something in the local domain. These
5842 two options are now turned off when doing these lookups.
5844 57. It seems that at least some releases of Reiserfs (which does not have the
5845 concept of a fixed number of inodes) returns zero and not -1 for the
5846 number of available inodes. This interacted badly with check_spool_inodes,
5847 which assumed that -1 was the "no such thing" setting. What I have done is
5848 to check that the total number of inodes is greater than zero before doing
5849 the test of how many are available.
5851 58. When a "warn" ACL statement has a log_message modifier, the message is
5852 remembered, and not repeated. This is to avoid a lot of repetition when a
5853 message has many recipients that cause the same warning to be written.
5854 However, Exim was preserving the list of already written lines for an
5855 entire SMTP session, which doesn't seem right. The memory is now reset if a
5856 new message is started.
5858 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5859 debugging output unless log_rewrite was also set.
5861 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5862 of (int)(handle) when we know that handle contains (void *)(-1).
5864 61. The Exim daemon panic-logs an error return when it closes the incoming
5865 connection. However "connection reset by peer" seems to be common, and
5866 isn't really an error worthy of noting specially, so that particular error
5869 62. When Exim is trying to find all the local interfaces, it used to panic and
5870 die if the ioctl to get the interface flags failed. However, it seems that
5871 on at least one OS (Solaris 9) it is possible to have an interface that is
5872 included in the list of interfaces, but for which you get a failure error
5873 for this call. This happens when the interface is not "plumbed" into a
5874 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5875 failure of the "get flags" call assumes that the interface is down.
5877 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5878 makes life easier for people who are doing arithmetic on fields extracted
5879 from dates, where you often get leading zeros that should not be
5880 interpreted as octal.
5882 64. Added qualify_domain to the redirect router, to override the global
5885 65. If a pathologically long header line contained very many addresses (the
5886 report of this problem mentioned 10 000) and each of them was rewritten,
5887 Exim could use up a very large amount of memory. (It kept on making new
5888 copies of the header line as it rewrote, and never released the old ones.)
5889 At the expense of a bit more processing, the header rewriting function has
5890 been changed so that it no longer eats memory in this way.
5892 66. The generation of the Received: header has been moved from the time that a
5893 message starts to be received, to the time that it finishes. The timestamp
5894 in the Received: header should now be very close to that of the <= log
5895 line. There are two side-effects of this change:
5897 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5898 logged header lines no longer include the local Received: line, because
5899 it has not yet been created. The same applies to a copy of the message
5900 that is returned to a non-SMTP sender when a message is rejected.
5902 (b) When a filter file is tested using -bf, no additional Received: header
5903 is added to the test message. After some thought, I decided that this
5906 This change does not affect the value of $received_for. It is still set
5907 after address rewriting, but before local_scan() is called.
5909 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5911 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5912 gave an unhelpful panic error message, and a defer error. I have managed to
5913 change this behaviour so that it now rejects any supplied certificate,
5914 which seems right, as the list of acceptable certificates is empty.
5916 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5917 gave an unhelpful defer error. I have not managed to make this reject any
5918 supplied certificates, but the error message it gives is "no certificate
5919 supplied", which is not helpful.
5921 70. exigrep's output now also includes lines that are not associated with any
5922 message, but which match the given pattern. Implemented by a patch from
5923 Martin Sluka, which also tidied up the Perl a bit.
5925 71. Recipient callout verification, like sender verification, was using <> in
5926 the MAIL FROM command. This isn't really the right thing, since the actual
5927 sender may affect whether the remote host accepts the recipient or not. I
5928 have changed it to use the actual sender in the callout; this means that
5929 the cache record is now keyed on a recipient/sender pair, not just the
5930 recipient address. There doesn't seem to be a real danger of callout loops,
5931 since a callout by the remote host to check the sender would use <>.
5932 [SEE ABOVE: changed after hitting problems.]
5934 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5935 temporary errors. However, in the case of such a code being given after
5936 the end of a data transmission (i.e. after ".") Exim was failing to write
5937 a retry record for the message. (Yes, there was some broken host that was
5938 actually sending 8xx at this point.)
5940 73. An unknown lookup type in a host list could cause Exim to panic-die when
5941 the list was checked. (An example that provoked this was putting <; in the
5942 middle of a list instead of at the start.) If this happened during a DATA
5943 ACL check, a -D file could be left lying around. This kind of configuration
5944 error no longer causes Exim to die; instead it causes a defer error. The
5945 incident is still logged to the main and panic logs.
5947 74. Buglet left over from Exim 3 conversion. The message "too many messages
5948 in one connection" was written to the rejectlog but not the mainlog, except
5949 when address rewriting (yes!) was being logged.
5951 75. Added write_rejectlog option.
5953 76. When a system filter was run not as root (that is, when system_filter_user
5954 was set), the values of the $n variables were not being returned to the
5955 main process; thus, they were not subsequently available in the $sn
5958 77. Added +return_path_on_delivery log selector.
5960 78. A connection timeout was being treated differently from recipients deferred
5961 when testing hosts_max_try with a message that was older than the host's
5962 retry timeout. (The host should not be counted, thus allowing all hosts to
5963 be tried at least once before bouncing.) This may have been the cause of an
5964 occasionally reported bug whereby a message would remain on the queue
5965 longer than the retry timeout, but would be bounced if a delivery was
5966 forced. I say "may" because I never totally pinned down the problem;
5967 setting up timeout/retry tests is difficult. See also the next item.
5969 79. The ultimate address timeout was not being applied to errors that involved
5970 a combination of host plus message (for example, a timeout on a MAIL
5971 command). When an address resolved to a number of possible hosts, and they
5972 were not all tried for each delivery (e.g. because of hosts_max_try), a
5973 message could remain on the queue longer than the retry timeout.
5975 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5978 81. Fixed an obscure SMTP outgoing bug which required at least the following
5979 conditions: (a) there was another message waiting for the same server;
5980 (b) the server returned 5xx to all RCPT commands in the first message so
5981 that the message was not completed; (c) the server dropped the connection
5982 or gave a negative response to the RSET that Exim sends to abort the
5983 transaction. The observed case was a dropped connection after DATA that had
5984 been sent in pipelining mode. That is, the server had advertised PIPELINING
5985 but was not implementing it correctly. The effect of the bug was incorrect
5986 behaviour, such as trying another host, and this could lead to a crash.
5992 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5993 and daemon.c were passed as pointers to ints; they should have been
5994 pointers to socklen_t variables (which are typically unsigned ints).
5996 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5999 3. Fixed a really odd bug that affected only the testing scheme; patching a
6000 certain fixed string in the binary changed the value of another string that
6001 happened to be identical to the end of the original first string.
6003 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6004 name", it returns that address as the IP address. On some operating
6005 systems (e.g. Solaris), it also passes back the IP address string as the
6006 "host name". However, on others (e.g. Linux), it passes back an empty
6007 string. Exim wasn't checking for this, and was changing the host name to an
6008 empty string, assuming it had been canonicalized.
6010 5. Although rare, it is permitted to have more than one PTR record for a given
6011 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6012 all the names associated with an address, because they do in Solaris.
6013 However, it seems that they do not in Linux for data that comes from the
6014 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6015 I found this out when I moved to a new Linux workstation and tried to run
6016 the Exim test suite.
6018 To get round this problem I have changed the code so that it now does its
6019 own call to the DNS to look up PTR records when searching for a host name.
6020 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6021 addresses that are only in /etc/hosts are still found.
6023 This behaviour is, however, controlled by an option called host_lookup_
6024 order, which defaults to "bydns:byaddr". If people want to use the other
6025 order, or indeed, just use one or the other means of lookup, they can
6026 specify it in this variable.
6028 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6029 some operating systems, this comes back from gethostbyaddr() as an empty
6030 string, and this is what Exim used to test for. However, it seems that in
6031 other systems, "." is yielded. Exim now tests for this case too.
6033 7. The values of check_spool_space and check_log_space are now held internally
6034 as a number of kilobytes instead of an absolute number of bytes. If a
6035 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6036 kilobyte. This means that much larger values can be stored.
6038 8. Exim monitor: an attempt to get the action menu when not actually pointing
6039 at a message produces an empty menu entitled "No message selected". This
6040 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6041 no entries in it ("Shell widget menu has zero width and/or height"). So I
6042 have added a single, blank menu entry in this case.
6044 9. Added ${quote_local_part.
6046 10. MIME decoding is now applied to the contents of Subject: header lines when
6049 11. Now that a reference to $sender_host_address automatically causes a reverse
6050 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6051 host lookup before query-style lookups in lists that might use this
6052 variable. This has therefore been abolished, and the "net-" prefix is no
6053 longer necessary for query-style lookups.
6055 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6056 have been a typo for LFLAGS, so it has been changed.
6058 13. The install script calls Exim with "-C /dev/null" in order to find the
6059 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6060 to be output. However, since Exim outputs its version number before the
6061 error, it didn't break the script. It just looked ugly. I fixed this by
6062 always allowing "-C /dev/null" if the caller is root.
6064 14. Ignore overlarge ACL variable number when reading spool file - insurance
6065 against a later release with more variables having written the file.
6067 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6068 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6069 this, as well as the form without the "IPv6" on the front.
6071 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6072 OS/Makefile-Darwin file.
6074 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6075 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6077 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6078 variable. However, some OS use ints for inodes. I've added cast to long int
6079 to get rid of the compiler warning.
6081 19. I had forgotten to lock out "/../" in configuration file names when
6082 ALT_CONFIG_PREFIX was set.
6084 20. Routers used for verification do not need to specify transports. However,
6085 if such a router generated a host list, and callout was configured, Exim
6086 crashed, because it could not find a port number from the (non-existent)
6087 transport. It now assumes port 25 in this circumstance.
6089 21. Added the -t option to exigrep.
6091 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6092 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6093 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6094 them is set, LOOKUP_LSEARCH is forced.
6096 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6098 24. Added sender and host information to the "rejected by local_scan()" log
6099 line; previously there was no indication of these.
6101 25. Added .include_if_exists.
6103 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6104 Linux. It turns out that not all file systems support this. Apparently some
6105 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6106 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6107 not supported on the file descriptor, is now ignored when Exim is trying to
6108 sync a directory. This applies only to Linux.
6110 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6112 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6113 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6116 29. Added support for maildirsize files from supplied patch (modified a bit).
6118 30. The use of :fail: followed by an empty string could lead Exim to respond to
6119 sender verification failures with (e.g.):
6121 550 Verification failed for <xxx>
6122 550 Sender verify failed
6124 where the first response line was missing the '-' that indicates it is not
6125 the final line of the response.
6127 31. The loop for finding the name of the user that called Exim had a hardwired
6128 limit of 10; it now uses the value of finduser_retries, which is used for
6129 all other user lookups.
6131 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6134 33. Exim was neglecting to zero errno before one call of strtol() when
6135 expanding a string and expecting an integer value. On some systems this
6136 resulted in spurious "integer overflow" errors. Also, it was casting the
6137 result into an int without checking.
6139 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6140 did not work. The code looks as if it has *never* worked, though it appears
6141 to have been documented since at least release 1.62. I have made it work.
6143 35. The "timeout_DNS" error in retry rules, also documented since at least
6144 1.62, also never worked. As it isn't clear exactly what this means, and
6145 clearly it isn't a major issue, I have abolished the feature by treating it
6146 as "timeout", and writing a warning to the main and panic logs.
6148 36. The display of retry rules for -brt wasn't always showing the error code
6151 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6152 timeout_connect_A, timeout_connect_MX.
6154 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6155 to the empty sender.
6157 39. The daemon was not analysing the content of -oX till after it had closed
6158 stderr and disconnected from the controlling terminal. This meant that any
6159 syntax errors were only noted on the panic log, and the return code from
6160 the command was 0. By re-arranging the code a little, I've made the
6161 decoding happen first, so such errors now appear on stderr, and the return
6162 code is 1. However, the actual setting up of the sockets still happens in
6163 the disconnected process, so errors there are still only recorded on the
6166 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6167 connections (as happens on some IP stacks) was logged at start up time as
6168 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6169 it from "IPv6 and IPv4", which means that two separate sockets are being
6172 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6173 says whether AF_INET or AF_INET6 was passed as an argument.
6175 42. Exiwhat output was messed up when time zones were included in log
6178 43. Exiwhat now gives more information about the daemon's listening ports,
6179 and whether -tls-on-connect was used.
6181 44. The "port" option of the smtp transport is now expanded.
6183 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6184 silently ignored. Now an error message is written to the main and panic
6187 46. There's a new ACL modifier called "logwrite" which writes to a log file
6188 as soon as it is encountered.
6190 47. Added $local_user_uid and $local_user_gid at routing time.
6192 48. Exim crashed when trying to verify a sender address that was being
6195 49. Exim was recognizing only a space character after ".include". It now also
6196 recognizes a tab character.
6198 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6199 extracting the relevant information from the specification. The man page no
6200 longer contains scrambled data for the -d option, and I've added a section
6201 at the front about calling Exim under different names.
6203 51. Added "extra_headers" argument to the "mail" command in filter files.
6205 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6208 53. Installed eximstats 1.29.
6210 54. Added transport_filter_timeout as a generic transport option.
6212 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6213 Cc: header lines. This was required by RFC 822, but it not required by RFC
6216 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6217 incoming messages that did not have them. Now it does so only if the
6218 message originates locally, that is, if there is no associated remote host
6219 address. When Resent- header lines are present, this applies to the Resent-
6220 lines rather than the non-Resent- lines.
6222 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6223 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6225 58. Messages for configuration errors now include the name of the main
6226 configuration file - useful now that there may be more than one file in a
6227 list (.included file names were always shown).
6229 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6230 for those rare installations that do not start the daemon as root or run it
6231 setuid root. I've cut out the call to initgroups() if the daemon is not
6234 60. The Exim user and group can now be bound into the binary as text strings
6235 that are looked up at the start of Exim's processing.
6237 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6239 62. Added $mailstore_basename variable.
6241 63. Installed patch to sieve.c from Michael Haardt.
6243 64. When Exim failed to open the panic log after failing to open the main log,
6244 the original message it was trying to log was written to stderr and debug
6245 output, but if they were not available (the usual case in production), it
6246 was lost. Now it is written to syslog before the two lines that record the
6247 failures to open the logs.
6249 65. Users' Exim filters run in subprocesses under the user's uid. It is
6250 possible for a "deliver" command or an alias in a "personal" command to
6251 provoke an address rewrite. If logging of address rewriting is configured,
6252 this fails because the process is not running as root or exim. There may be
6253 a better way of dealing with this, but for the moment (because 4.30 needs
6254 to be released), I have disabled address rewrite logging when running a
6255 filter in a non-root, non-exim process.
6261 1. The buildconfig auxiliary program wasn't quoting the value set for
6262 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6263 not defined. This bug was masked in 4.22 by the effect that was fixed in
6266 2. Some messages that were rejected after a message id was allocated were
6267 shown as "incomplete" by exigrep. It no longer does this for messages that
6268 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6270 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6271 have allow_domain_literals set, the ID did not get logged in the <= line.
6272 Domain literals are now always recognized in Message-ID: header lines.
6274 4. The first argument for a ${extract expansion item is the key name or field
6275 number. Leading and trailing spaces in this item were not being ignored,
6276 causing some misleading effects.
6278 5. When deliver_drop_privilege was set, single queue runner processes started
6279 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6280 same command in the process it spins off) were not dropping privilege.
6282 6. When the daemon running as "exim" started a queue runner, it always
6283 re-executed Exim in the spun-off process. This is a waste of effort when
6284 deliver_drop_privilege is set. The new process now just calls the
6285 queue-runner function directly.
6291 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6294 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6295 ignored. Though the use of -oP was forcing the writing of a pid file, it
6296 was always written to the default place.
6298 3. If the message "no IP address found for host xxxx" is generated during
6299 incoming verification, it is now followed by identification of the incoming
6300 connection (so you can more easily find what provoked it).
6302 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6304 5. Added some features to "harden" Exim a bit more against certain attacks:
6306 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6307 be put in Local/Makefile. This is like the never_users runtime option,
6308 but it cannot be overridden. The default setting is "root".
6310 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6311 prefix string with which any file named in a -C command line option
6314 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6315 is retained for -C and -D only if the caller of Exim is root. Without
6316 it, the exim user may also use -C and -D and retain privilege.
6318 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6319 command line option is disabled.
6321 6. Macro names set by the -D option must start with an upper case letter, just
6322 like macro names defined in the configuration file.
6324 7. Added "dereference=" facility to LDAP.
6326 8. Two instances of the typo "uknown" in the source files are fixed.
6328 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6329 the Configure-Makefile script screwed up while processing it.
6331 10. Incorporated PCRE 4.4.
6333 11. The SMTP synchronization check was not operating right at the start of an
6334 SMTP session. For example, it could not catch a HELO sent before the client
6335 waited for the greeting. There is now a check for outstanding input at the
6336 point when the greeting is written. Because of the duplex, asynchronous
6337 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6338 way, but not yet received, when the check is performed.
6340 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6341 on TCP/IP sockets, because this apparently causes some broken clients to
6344 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6345 unchanged) from the Cygwin maintainer.
6347 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6348 was supported instead of testing for mbx. Effectively a typo.
6350 15. The spa authenticator server code was not checking that the input it
6351 received was valid base64.
6353 16. The debug output line for the "set" modifier in ACLs was not showing the
6354 name of the variable that was being set.
6356 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6358 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6359 lookup on its own. Something else was needed to trigger the lookup. For
6360 example, a match in host_lookup or the need for a host name in a host list.
6361 Now, if $sender_host_name is referenced and the host name has not yet been
6362 looked up, a lookup is performed. If the lookup fails, the variable remains
6363 empty, and $host_lookup_failed is set to "1".
6365 19. Added "eqi" as a case-independent comparison operator.
6367 20. The saslauthd authentication condition could segfault if neither service
6368 nor realm was specified.
6370 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6371 error message that was logged was misleading, and incoming SMTP
6372 connections were dropped. The message is now more accurate, and temporary
6373 errors are given to SMTP connections.
6375 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6376 (except RSET) with a 421 error, until QUIT is received. However, it was
6377 failing to send a response to QUIT.
6379 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6380 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6381 ACL is now run after the helo_try_verify_hosts code.
6383 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6384 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6385 case-independent, so other case variants are also recognized). Apparently
6386 some systems use these upper case variants.
6388 25. If more than two messages were waiting for the same host, and a transport
6389 filter was specified for the transport, Exim sent two messages over the
6390 same TCP/IP connection, and then failed with "socket operation on non-
6391 socket" when it tried to send the third.
6393 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6395 27. The extern definition of crypt16() in expand.c was not being excluded when
6396 the OS had its own crypt16() function.
6398 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6399 as a preferred synonym for return_size_limit, both as an option and as an
6402 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6404 30. Changed the default configuration ACL to relax the local part checking rule
6405 for addresses that are not in any local domains. For these addresses,
6406 slashes and pipe symbols are allowed within local parts, but the sequence
6407 /../ is explicitly forbidden.
6409 31. SPA server authentication was not clearing the challenge buffer before
6412 32. log_message in a "warn" ACL statement was writing to the reject log as
6413 well as to the main log, which contradicts the documentation and doesn't
6414 seem right (because no rejection is happening). So I have stopped it.
6416 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6417 However, I am unable to do any testing of this.
6419 34. Fixed an infelicity in the appendfile transport. When checking directories
6420 for a mailbox, to see if any needed to be created, it was accidentally
6421 using path names with one or more superfluous leading slashes; tracing
6422 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6424 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6425 contents are added to the log line that is written for every discarded
6426 recipient. (Previously a log_message setting was ignored.)
6428 36. The ${quote: operator now quotes the string if it is empty.
6430 37. The install script runs exim in order to find its version number. If for
6431 some reason other than non-existence or emptiness, which it checks, it
6432 could not run './exim', it was installing it with an empty version number,
6433 i.e. as "exim-". This error state is now caught, and the installation is
6436 38. An argument was missing from the function that creates an error message
6437 when Exim fails to connect to the socket for saslauthd authentication.
6438 This could cause Exim to crash, or give a corrupted message.
6440 39. Added isip, isip4, and isip6 to ${if conditions.
6442 40. The ACL variables $acl_xx are now saved with the message, and can be
6443 accessed later in routers, transports, and filters.
6445 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6446 strings in the file are not string-expanded.
6448 42. If a MAIL command specified a SIZE value that was too large to fit into an
6449 int variable, the check against message_size_limit failed. Such values are
6450 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6451 day this will have to be increased, but I don't think I want to be around
6452 when emails are that large.
6459 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6460 iconv() is not standard in FreeBSD.
6462 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6463 IPv6 enabled. The observed symptom was a segmentation fault on return from
6464 the function os_common_find_running_interfaces() in src/os.c.
6466 3. In the check_special_case() function in daemon.c I had used "errno" as an
6467 argument name, which causes warnings on some systems. This was basically a
6468 typo, since it was named "eno" in the comments!
6470 4. The code that waits for the clock to tick (at a resolution of some fraction
6471 of a second) so as to ensure message-id uniqueness was always waiting for
6472 at least one whole tick, when it could have waited for less. [This is
6473 almost certainly not relevant at current processor speeds, where it is
6474 unlikely to ever wait at all. But we try to future-proof.]
6476 5. The function that sleeps for a time interval that includes fractions of a
6477 second contained a race. It did not block SIGALRM between setting the
6478 timer, and suspending (a couple of lines later). If the interval was short
6479 and the sigsuspend() was delayed until after it had expired, the suspension
6480 never ended. On busy systems this could lead to processes getting stuck for
6483 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6484 process, before it forks any delivery processes. The open lookup caching
6485 mechanism meant that the open file or database connection was passed into
6486 the delivery process. The problem was that delivery processes always tidy
6487 up cached lookup data. This could cause a problem for the next delivery
6488 process started by the queue runner, because the external queue runner
6489 process does not know about the closure. So the next delivery process
6490 still has data in the lookup cache. In the case of a file lookup, there was
6491 no problem because closing a file descriptor in a subprocess doesn't affect
6492 the parent. However, if the lookup was caching a connection to a database,
6493 the connection was closed, and the second delivery process was likely to
6494 see errors such as "PGSQL: query failed: server closed the connection
6495 unexpectedly". The problem has been fixed by closing all cached lookups
6496 in a queue runner before running a delivery process.
6498 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6499 seem to have the "const" qualifier which it has on other OS. I've
6502 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6503 *of the same type* (client or server) with the same public name that an
6504 error should be diagnosed.
6506 9. When Exim looked up a host name for an IP address, but failed to find the
6507 original IP address when looking up the host name (a safety check), it
6508 output the message "<ip address> does not match any IP for NULL", which was
6509 confusing, to say the least. The bug was that the host name should have
6510 appeared instead of "NULL".
6512 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6513 user that is built into the binary, and the -C or -D options is used, root
6514 privilege is dropped before the configuration file is read. In addition,
6515 logging is switched to stderr instead of the normal log files. If the
6516 configuration then re-defines the Exim user, the unprivileged environment
6517 is probably not what is expected, so Exim logs a panic warning message (but
6520 However, if deliver_drop_privilege is set, the unprivileged state may well
6521 be exactly what is intended, so the warning has been cut out in that case,
6522 and Exim is allowed to try to write to its normal log files.
6528 1. smtp_return_error_details was not giving details for temporary sender
6529 or receiver verification errors.
6531 2. Diagnose a configuration error if two authenticators have the same public
6534 3. Exim used not to create the message log file for a message until the first
6535 delivery attempt. This could be confusing when incoming messages were held
6536 for policy or load reasons. The message log file is now created at the time
6537 the message is received, and an initial "Received" line is written to it.
6539 4. The automatically generated man page for command line options had a minor
6540 bug that caused no ill effects; however, a more serious problem was that
6541 the procedure for building the man page automatically didn't always
6542 operate. Consequently, release 4.20 contains an out-of-date version. This
6543 shouldn't happen again.
6545 5. When building Exim with embedded Perl support, the script that builds the
6546 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6547 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6549 6. The freeze_tell option was not being used for messages that were frozen on
6550 arrival, either by an ACL or by local_scan().
6552 7. Added the smtp_incomplete_transaction log selector.
6554 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6555 was accepting AUTH without a new EHLO.
6557 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6558 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6561 10. Logging of TCP/IP connections (when configured) now happens in the main
6562 daemon process instead of the child process, so that the TCP/IP connection
6563 count is more accurate (but it can never be perfect).
6565 11. The use of "drop" in a nested ACL was not being handled correctly in the
6566 outer ACL. Now, if condition failure induced by the nested "drop" causes
6567 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6568 or "require"), the connection is dropped.
6570 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6571 that yield "discard" can now be used with an "accept" or a "discard" verb,
6572 but an error is generated for any others (because I can't see a useful way
6573 to define what should happen).
6575 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6576 are now processed in the same way as lines in the Exim configuration file.
6577 In particular, continuation lines are supported.
6579 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6581 15. Added -ti meaning -t -i.
6583 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6584 domains, and warn by logging if others are found.
6586 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6587 returned from GIFCONF when Exim is trying to find the list of interfaces on
6588 a host. The code in os.c has been modified to copy each ifreq to an aligned
6589 structure in all cases.
6591 Also, in some cases, the returned ifreq's were being copied to a 'struct
6592 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6593 means the last couple of bytes of an IPv6 address could be chopped if the
6594 ifreq contained only a normal sockaddr (14 bytes storage).
6596 18. Named domain lists were not supported in the hosts_treat_as_local option.
6597 An entry such as +xxxx was not recognized, and was treated as a literal
6600 19. Ensure that header lines added by a DATA ACL are included in the reject log
6601 if the ACL subsequently rejects the message.
6603 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6604 MD5 (which is deprecated).
6606 21. When testing a filter file using -bf, Exim was writing a message when it
6607 took the sender from a "From " line in the message, but it was not doing so
6608 when it took $return_path from a Return-Path: header line. It now does.
6610 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6611 with a valid header line field name (a series of printing characters
6612 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6614 23. Changed "disc" in the source to "disk" to conform to the documentation and
6615 the book and for uniformity.
6617 24. Ignore Sendmail's -Ooption=value command line item.
6619 25. When execve() failed while trying to run a command in a pipe transport,
6620 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6621 could be confused with a return value of 69 from the command itself. This
6622 has been changed to 127, the value the shell returns if it is asked to run
6623 a non-existent command. The wording for the related log line suggests a
6624 non-existent command as the problem.
6626 26. If received_header_text expands to an empty string, do not add a Received:
6627 header line to the message. (Well, it adds a token one on the spool, but
6628 marks it "old" so that it doesn't get used or transmitted.)
6630 27. Installed eximstats 1.28 (addition of -nt option).
6632 28. There was no check for failure on the call to getsockname() in the daemon
6633 code. This can fail if there is a shortage of resources on the system, with
6634 ENOMEM, for example. A temporary error is now given on failure.
6636 29. Contrary to the C standard, it seems that in some environments, the
6637 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6638 program. Exim now does this explicitly; it affects the formatting of
6639 timestamps using strftime().
6641 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6642 complaints. I've now initialized all the variables, to avoid this.
6644 32. Header lines added by a system filter were not being "seen" during
6645 transport-time rewrites.
6647 33. The info_callback() function passed to OpenSSL is set up with type void
6648 (*)(SSL *, int, int), as described somewhere. However, when calling the
6649 function (actually a macro) that sets it up, the type void(*)() is
6650 expected. I've put in a cast to prevent warnings from picky compilers.
6652 34. If a DNS black list lookup found a CNAME record, but there were no A
6653 records associated with the domain it pointed at, Exim crashed.
6655 35. If a DNS black list lookup returned more than one A record, Exim ignored
6656 all but the first. It now scans all returned addresses if a particular IP
6657 value is being sought. In this situation, the contents of the
6658 $dnslist_value variable are a list of all the addresses, separated by a
6661 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6662 to accept a host name and all its aliases if the forward lookup for any of
6663 them yielded the IP address of the incoming connection. Now it accepts only
6664 those names whose forward lookup yields the correct IP address. Any other
6665 names are discarded. This closes a loophole whereby a rogue DNS
6666 administrator could create reverse DNS records to break through a
6667 wildcarded host restriction in an ACL.
6669 37. If a user filter or a system filter that ran in a subprocess used any of
6670 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6671 the wrong values were passed to the pipe command ($thisaddress had the
6672 value of $0, $0 had the value of $1, etc). This bug was introduced by
6673 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6675 38. Improved the line breaking for long SMTP error messages from ACLs.
6676 Previously, if there was no break point between 40 and 75 characters, Exim
6677 left the rest of the message alone. Two changes have been made: (a) I've
6678 reduced the minimum length to 35 characters; (b) if it can't find a break
6679 point between 35 and 75 characters, it looks ahead and uses the first one
6680 that it finds. This may give the occasional overlong line, but at least the
6681 remaining text gets split now.
6683 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6684 file descriptors might be low, and that setting 1000 would always raise it.
6685 It turns out that in some environments, the limit is already over 1000 and
6686 that lowering it causes trouble. So now Exim takes care not to decrease it.
6688 40. When delivering a message, the value of $return_path is set to $sender_
6689 address at the start of routing (routers may change the value). By an
6690 oversight, this default was not being set up when an address was tested by
6691 -bt or -bv, which affected the outcome if any router or filter referred to
6694 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6695 log only when "message" or "log_message" are set. However, if one of the
6696 conditions was an address verification, or a call to a nested ACL, the
6697 messages generated by the underlying test were being passed through. This
6698 no longer happens. The underlying message is available in $acl_verify_
6699 message for both "message" and "log_message" expansions, so it can be
6700 passed through if needed.
6702 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6703 new expansion $bh_ to give the encoded byte string without charset
6704 translation. Translation happens only if iconv() is available; HAVE_ICONV
6705 indicates this at build time. HEADERS_CHARSET gives the charset to
6706 translate to; headers_charset can change it in the configuration, and
6707 "headers charset" can change it in an individual filter file.
6709 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6710 that creates RFC 2047 encoded "words" labels them as that charset instead
6711 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6712 expansion operator; (ii) when Exim creates a From: line for a local
6713 message; (iii) when a header line is rewritten to include a "phrase" part.
6715 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6716 buggy, causing it to skip the first lines of messages whose message ID
6717 ended in 'D'. This would not have bitten before Exim release 4.14, because
6718 message IDs were unlikely to end in 'D' before then. The effect was to have
6719 incorrect size information for certain domains.
6721 45. #include "config.h" was missing at the start of the crypt16.c module. This
6722 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6725 46. If there was a timeout during a "random" callout check, Exim treated it as
6726 a failure of the random address, and carried on sending RSET and the real
6727 address. If the delay was just some slowness somewhere, the response to the
6728 original RCPT would be taken as a response to RSET and so on, causing
6729 mayhem of various kinds.
6731 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6732 when I implemented it. It didn't allow for the fact that some option values
6733 may legitimately be negative (e.g. size_addition), and it didn't even do
6734 the right test for positive values.
6736 48. Domain names in DNS records are case-independent. Exim always looks them up
6737 in lower case. Some resolvers return domain names in exactly the case they
6738 appear in the zone file, that is, they may contain uppercase letters. Not
6739 all resolvers do this - some return always lower case. Exim was treating a
6740 change of case by a resolver as a change of domain, similar to a widening
6741 of a domain abbreviation. This triggered its re-routing code and so it was
6742 trying to route what was effectively the same domain again. This normally
6743 caused routing to fail (because the router wouldn't handle the domain
6744 twice). Now Exim checks for this case specially, and just changes the
6745 casing of the domain that it ultimately uses when it transmits the message
6748 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6751 50. If a filter generated a file delivery with a non-absolute name (possible if
6752 no home directory exists for the router), the forbid_file option was not
6755 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6756 the existing equality matching.
6758 52. Exim was using ints instead of ino_t variables in some places where it was
6759 dealing with inode numbers.
6761 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6762 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6763 called TMPDIR, and if it finds it is different, it changes its value.
6765 54. The smtp_printf() function is now made available to local_scan() so
6766 additional output lines can be written before returning. There is also an
6767 smtp_fflush() function to enable the detection of a dropped connection.
6768 The variables smtp_input and smtp_batched_input are exported to
6771 55. Changed the default runtime configuration: the message "Unknown user"
6772 has been removed from the ACL, and instead placed on the localuser router,
6773 using the cannot_route_message feature. This means that any verification
6774 failures that generate their own messages won't get overridden. Similarly,
6775 the "Unrouteable address" message that was in the ACL for unverifiable
6776 relay addresses has also been removed.
6778 56. Added hosts_avoid_esmtp to the smtp transport.
6780 57. The exicyclog script was not checking for the esoteric option
6781 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6782 will work only if exicyclog is run under the appropriate euid.
6784 58. Following a discussion on the list, the rules by which Exim recognises line
6785 endings on incoming messages have been changed. The -dropcr and drop_cr
6786 options are now no-ops, retained only for backwards compatibility. The
6787 following line terminators are recognized: LF CRLF CR. However, special
6788 processing applies to CR:
6790 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6791 nor a local message in the state where . is a terminator.
6793 (ii) If a bare CR is encountered in a header line, an extra space is added
6794 after the line terminator so as not to end the header. The reasoning
6795 behind this is that bare CRs in header lines are most likely either
6796 to be mistakes, or people trying to play silly games.
6798 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6799 was being incorrectly given as 18 bytes larger than it should have been.
6800 This is a VOB (very old bug).
6802 60. This may never have affected anything current, but just in case it has:
6803 When the local host is found other than at the start of a list of hosts,
6804 the local host, those with the same MX, and any that follow, are discarded.
6805 When the list in question was part of a longer list of hosts, the following
6806 hosts (not currently being processed) were also being discarded. This no
6807 longer happens. I'm not sure if this situation could ever has previously
6810 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6813 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6814 header. This is something that is recommended in a new Internet Draft, and
6815 is something that is documented as being done by Sendmail. There are two
6816 possible values. For messages generated by the autoreply transport, Exim
6819 Auto-Submitted: auto-replied
6821 whereas for all other generated messages (e.g. bounces) it adds
6823 Auto-Submitted: auto-generated
6825 63. The "personal" condition in filters now includes a test for the
6826 Auto-Submitted: header. If it contains the string "auto-" the message it
6827 not considered personal.
6829 64. Added rcpt_include_affixes as a generic transport option.
6831 65. Added queue_only_override (default true).
6833 66. Added the syslog_duplication option.
6835 67. If what should have been the first header line of a message consisted of
6836 a space followed by a colon, Exim was mis-interpreting it as a header line.
6837 It isn't of course - it is syntactically invalid and should therefore be
6838 treated as the start of the message body. The misbehaviour could have
6839 caused a number of strange effects, including loss of data in subsequent
6840 header lines, and spool format errors.
6842 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6843 client host had authenticated. This control can now be exercised by an ACL
6844 for more flexibility.
6846 69. By default, callouts do not happen when testing with -bh. There is now a
6847 variant, -bhc, which does actually run the callout code, including
6848 consulting and updating the callout cache.
6850 70. Added support for saslauthd authentication, courtesy of Alexander
6853 71. If statvfs() failed on the spool or log directories while checking their
6854 size for availability, Exim confusingly gave the error "space shortage".
6855 Furthermore, in debugging mode it crashed with a floating point exception.
6856 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6857 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6858 serious problem, Exim now writes to the main and panic logs when this
6859 happens, with details of the failure. It then refuses to accept the
6860 incoming message, giving the message "spool directory problem" or "log
6861 directory problem" with a 421 code for SMTP messages.
6863 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6864 0, 1, and 2 exist, because some OS complain for execs without them (see
6865 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6866 descriptors. However, the code omitted to check that the open succeeded,
6867 causing mysterious errors if for some reason the permissions on /dev/null
6868 got screwed. Now Exim writes a message to the main and panic logs, and
6869 bombs out if it can't open /dev/null.
6871 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6872 interact so that it is all more flexible. It is supposed to remain
6873 backwards compatible. Also added extra_local_interfaces.
6875 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6876 to bomb out with an assertion failure - to the client this appears as a
6877 connection drop. This problem occurs in the part of the code that was taken
6878 from the Samba project. Fortunately, the assertion is in a very simple
6879 function, so I have fixed this by reproducing the function inline in the
6880 one place where it is called, and arranging for authentication to fail
6881 instead of killing the process with assert().
6883 75. The SPA client code was not working when the server requested OEM rather
6884 than Unicode encoding.
6886 76. Added code to make require_files with a specific uid setting more usable in
6887 the case where statting the file as root fails - usually a non-root-mounted
6888 NFS file system. When this happens and the failure is EACCES, Exim now
6889 forks a subprocess and does the per-uid checking as the relevant uid.
6891 77. Added process_log_path.
6893 78. If log_file_path was not explicitly set, a setting of check_log_space or
6894 check_log_inodes was ignored.
6896 79. If a space check for the spool or log partitions fails, the incident is now
6897 logged. Of course, in the latter case the data may get lost...
6899 80. Added the %p formatting code to string_format() so that it can be used to
6900 print addresses in debug_print(). Adjusted all the address printing in the
6901 debugging in store.c to use %p rather than %d.
6903 81. There was a concern that a line of code in smtp_in.c could overflow a
6904 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6905 initially expressed, the concern was not well-founded, because trailing
6906 spaces are removed early. However, if the trailing spaces were followed by
6907 a NULL, they did not get removed, so the overflow was possible. Two fixes
6910 (a) I re-wrote the offending code in a cleaner fashion.
6911 (b) If an incoming SMTP command contains a NULL character, it is rejected
6914 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6915 runs initgroups(), so that if the Exim user is in any additional groups,
6916 they will be used during message reception.
6922 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob