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
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".
145 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
146 than -2003 did; needs libidn2 in addition to libidn.
148 JH/02 The path option on a pipe transport is now expanded before use.
150 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
151 Patch provided by "Björn", documentation fix added too.
153 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
154 missing a wire-to-host endian conversion.
156 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
157 close after a BDAT command line could be taken as a following command,
158 giving a synch failure. Fix by only checking for synch immediately
159 before acknowledging the chunk.
161 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
162 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
163 macro. Patches provided by Josh Soref.
165 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
166 Previously we did not; the RFC seems ambiguous and VRFY is not listed
167 by IANA as a service extension. However, John Klensin suggests that we
170 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
171 the dkim code may be unix-mode line endings rather than smtp wire-format
172 CRLF, so prepend a CR to any bare LF.
174 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
175 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
177 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
178 the first were themselves being wrongly included in the feed into dkim
179 processing; with most chunk sizes in use this resulted in an incorrect
180 body hash calculated value.
182 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
183 DKIM signature block, for verification. Although advised against by
184 standards it is specifically not ruled illegal.
186 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
188 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
189 missing a body hash (the bh= tag).
191 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
192 It seems that HAProxy sends the Proxy Protocol information in clear and
193 only then does a TLS startup, so do the same.
195 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
196 TCP connections (such as for Spamd) unless the daemon successfully set
197 Fast Open mode on its listening sockets. This fixes breakage seen on
198 too-old kernels or those not configured for Fast Open, at the cost of
199 requiring both directions being enabled for TFO, and TFO never being used
200 by non-daemon-related Exim processes.
202 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
203 endings, at least on the first header line. Try to canonify any that get
204 past that check, despite the cost.
206 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
207 now limited to an arbitrary five deep, while parsing addresses with the
208 strip_excess_angle_brackets option enabled.
210 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
211 instead leave the unprompted TLS handshake in socket buffer for the
212 TLS library to consume.
214 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
216 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
218 JH/16 Drop variables when they go out of scope. Memory management drops a whole
219 region in one operation, for speed, and this leaves assigned pointers
220 dangling. Add checks run only under the testsuite which checks all
221 variables at a store-reset and panics on a dangling pointer; add code
222 explicitly nulling out all the variables discovered. Fixes one known
223 bug: a transport crash, where a dangling pointer for $sending_ip_address
224 originally assigned in a verify callout, is re-used.
226 PP/06 Drop '.' from @INC in various Perl scripts.
228 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
230 PP/08 Reduce a number of compilation warnings under clang; building with
231 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
232 should be warning-free.
234 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
236 HS/01 Fix portability problems introduced by PP/08 for platforms where
237 realloc(NULL) is not equivalent to malloc() [SunOS et al].
239 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
240 chunk. This allows us to accept broken chunked messages. We need a more
241 general solution here.
243 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
244 already-broken messages in the queue.
246 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
248 JH/19 Fix reference counting bug in routing-generated-address tracking.
254 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
255 supports it and a size is available (ie. the sending peer gave us one).
257 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
258 years of being deprecated). The replacements are the ACLs
259 acl_smtp_mime and acl_not_smtp_mime.
261 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
262 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
263 or even an in-clear connection were permitted. Now, if the host lookup
264 was dnssec and dane was requested then the host is only used if the
265 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
266 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
267 if one fails this test.
268 This means that a poorly-configured remote DNS will make it incommunicado;
269 but it protects against a DNS-interception attack on it.
271 JH/04 Bug 1810: make continued-use of an open smtp transport connection
272 non-noisy when a race steals the message being considered.
274 JH/05 If main configuration option tls_certificate is unset, generate a
275 self-signed certificate for inbound TLS connections.
277 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
278 in rewrites and routers.
280 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
281 and logged a warning sing 4.83; now they are a configuration file error.
283 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
284 (lacking @domain). Apply the same qualification processing as RCPT.
286 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
288 JH/10 Support ${sha256:} applied to a string (as well as the previous
291 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
292 a cutthrough deliver is pending, as we always want to make a connection.
293 This also avoids re-routing the message when later placing the cutthrough
294 connection after a verify cache hit.
295 Do not update it with the verify result either.
297 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
298 when routing results in more than one destination address.
300 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
301 signing (which inhibits the cutthrough capability). Previously only
302 the presence of an option was tested; now an expansion evaluating as
303 empty is permissible (obviously it should depend only on data available
304 when the cutthrough connection is made).
306 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
307 the relevant preceding SMTP command did not note the pipelining mode.
309 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
310 Previously they were not counted.
312 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
313 as one having no matching records. Previously we deferred the message
314 that needed the lookup.
316 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
317 distinguished as "(=".
319 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
320 for missing MX records. Previously it only worked for missing A records.
322 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
324 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
325 after the data-go-ahead and data-ack. Patch from Jason Betts.
327 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
328 even for a "none" policy. Patch from Tony Meyer.
330 JH/22 Fix continued use of a connection for further deliveries. If a port was
331 specified by a router, it must also match for the delivery to be
334 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
335 When one of the recipients of a message was unsuitable for the connection
336 (has no matching addresses), we lost track of needing to mark it
337 deferred. As a result mail would be lost.
339 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
341 JH/25 Decoding ACL controls is now done using a binary search; the source code
342 takes up less space and should be simpler to maintain. Merge the ACL
343 condition decode tables also, with similar effect.
345 JH/26 Fix problem with one_time used on a redirect router which returned the
346 parent address unchanged. A retry would see the parent address marked as
347 delivered, so not attempt the (identical) child. As a result mail would
350 JH/27 Fix a possible security hole, wherein a process operating with the Exim
351 UID can gain a root shell. Credit to http://www.halfdog.net/ for
352 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
355 JH/28 Enable {spool,log} filesystem space and inode checks as default.
356 Main config options check_{log,spool}_{inodes,space} are now
357 100 inodes, 10MB unless set otherwise in the configuration.
359 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
360 Previously it only applied to the main-section connection policy
363 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
365 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
366 by me. Added RFC7919 DH primes as an alternative.
368 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
369 are not in the system include path.
371 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
372 GnuTLS, when a session startup failed (eg because the client disconnected)
373 Exim did stdio operations after fclose. This was exposed by a recent
374 change which nulled out the file handle after the fclose.
376 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
377 signed directly by the cert-signing cert, rather than an intermediate
378 OCSP-signing cert. This is the model used by LetsEncrypt.
380 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
382 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
383 an incoming connection.
385 HS/02 Bug 1802: Do not half-close the connection after sending a request
388 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
389 fallback to "prime256v1".
391 JH/34 SECURITY: Use proper copy of DATA command in error message.
392 Could leak key material. Remotely exploitable. CVE-2016-9963.
398 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
399 and 3.4.4 - once the server is enabled to respond to an OCSP request
400 it does even when not requested, resulting in a stapling non-aware
401 client dropping the TLS connection.
403 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
404 support variable-length bit vectors. No functional change.
406 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
407 The I= interface field on outgoing lines is now after the H= remote
408 host field, same as incoming lines. There is a separate
409 outgoing_interface log selector which allows you to disable the
412 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
413 If not running log_selector +smtp_connection the mainlog would be held
414 open indefinitely after a "too many connections" event, including to a
415 deleted file after a log rotate. Leave the per net connection logging
416 leaving it open for efficiency as that will be quickly detected by the
417 check on the next write.
419 HS/01 Bug 1671: Fix post transport crash.
420 Processing the wait-<transport> messages could crash the delivery
421 process if the message IDs didn't exist for some reason. When
422 using 'split_spool_directory=yes' the construction of the spool
423 file name failed already, exposing the same netto behaviour.
425 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
426 mime_regex ACL conditions.
428 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
429 to DSN fail messages (bounces): remote IP, remote greeting, remote response
430 to HELO, local diagnostic string.
432 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
433 log line to debug. Even when configured with a tls authenticator many
434 client connections are expected to not authenticate in this way, so
435 an authenticate fail is not an error.
437 HS/02 Add the Exim version string to the process info. This way exiwhat
438 gives some more detail about the running daemon.
440 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
441 matter for fast-change records such as DNSBLs.
443 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
444 retry record, even if constant. There may be multiple transports with
445 different interface settings and the retry behaviour needs to be kept
448 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
450 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
452 JH/10 Bug 840: fix log_defer_output option of pipe transport
454 JH/11 Bug 830: use same host for all RCPTS of a message, even under
455 hosts_randomize. This matters a lot when combined with mua_wrapper.
457 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
458 ${quote_pgsql:<string>} operator.
460 JH/13 Bug 1708: avoid misaligned access in cached lookup.
462 JH/14 Change header file name for freeradius-client. Relevant if compiling
463 with Radius support; from the Gentoo tree and checked under Fedora.
465 JH/15 Bug 1712: Introduce $prdr_requested flag variable
467 JH/16 Bug 1714: Permit an empty string as expansion result for transport
468 option transport_filter, meaning no filtering.
470 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
472 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
473 defaults to "*" (all hosts). The variable is now available when not built
474 with TLS, default unset, mainly to enable keeping the testsuite sane.
475 If a server certificate is not supplied (via tls_certificate) an error is
476 logged, and clients will find TLS connections fail on startup. Presumably
477 they will retry in-clear.
478 Packagers of Exim are strongly encouraged to create a server certificate
479 at installation time.
481 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
482 with the $config_file variable.
484 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
485 in transport context, after the attempt, and per-recipient. The latter type
486 is per host attempted. The event data is the error message, and the errno
487 information encodes the lookup type (A vs. MX) used for the (first) host,
488 and the trailing two digits of the smtp 4xx response.
490 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
491 to write to mainlog (or rejectlog, paniclog) in the window between file
492 creation and permissions/ownership being changed. Particularly affects
493 installations where exicyclog is run as root, rather than exim user;
494 result is that the running daemon panics and dies.
496 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
498 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
499 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
500 "pri" and "weight". Note that the previous implicit priority given by the
501 list order is no longer honoured.
503 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
506 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
507 by defining SUPPORT_SOCKS.
509 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
510 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
511 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
512 variables are renamed to proxy_{local,external}_{address,port}.
514 JH/27 Move Internationalisation support from Experimental to mainline, enabled
515 for a build by defining SUPPORT_I18N
517 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
518 of the query string, and make ${quote_redis:} do that quoting.
520 JH/29 Move Events support from Experimental to mainline, enabled by default
521 and removable for a build by defining DISABLE_EVENT.
523 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
525 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
526 cached by the daemon.
528 JH/32 Move Redis support from Experimental to mainline, enabled for a build
529 by defining LOOKUP_REDIS. The libhiredis library is required.
531 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
532 keys are given for lookup.
534 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
535 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
536 only supported when built with TLS support. The PolarSSL SHA routines
537 are still used when the TLS library is too old for convenient support.
539 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
540 openssl_options), for security. OpenSSL forces this from version 1.1.0
541 server-side so match that on older versions.
543 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
544 allocation for $value could be released as the expansion processing
545 concluded, but leaving the global pointer active for it.
547 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
548 and to use the domains and local_parts ACL conditions.
550 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
551 incorrectly not doubled on cutthrough transmission, hence seen as a
552 body-termination at the receiving system - resulting in truncated mails.
553 Commonly the sender saw a TCP-level error, and retransmitted the message
554 via the normal store-and-forward channel. This could result in duplicates
555 received - but deduplicating mailstores were liable to retain only the
556 initial truncated version.
558 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
560 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
562 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
563 we're in there, support oversigning also; bug 1309.
565 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
567 HS/04 Add support for keep_environment and add_environment options.
569 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
570 either intentional arithmetic overflow during PRNG, or testing config-
573 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
574 delivery resulted in actual delivery. Cancel cutthrough before DATA
577 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
578 rejects a recipient: pass the reject to the originator.
580 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
581 Many were false-positives and ignorable, but it's worth fixing the
584 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
585 for the new environment-manipulation done at startup. Move the routines
586 from being local to tls.c to being global via the os.c file.
588 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
589 an extract embedded as result-arg for a map, the first arg for extract
590 is unavailable so we cannot tell if this is a numbered or keyed
591 extraction. Accept either.
597 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
600 JH/02 The smtp transport option "multi_domain" is now expanded.
602 JH/03 The smtp transport now requests PRDR by default, if the server offers
605 JH/04 Certificate name checking on server certificates, when exim is a client,
606 is now done by default. The transport option tls_verify_cert_hostnames
607 can be used to disable this per-host. The build option
608 EXPERIMENTAL_CERTNAMES is withdrawn.
610 JH/05 The value of the tls_verify_certificates smtp transport and main options
611 default to the word "system" to access the system default CA bundle.
612 For GnuTLS, only version 3.0.20 or later.
614 JH/06 Verification of the server certificate for a TLS connection is now tried
615 (but not required) by default. The verification status is now logged by
616 default, for both outbound TLS and client-certificate supplying inbound
619 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
622 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
623 Status Notification (bounce) messages are now MIME format per RFC 3464.
624 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
625 under the control of the dsn_advertise_hosts option, and routers may
626 have a dsn_lasthop option.
628 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
629 default, modifiable by a malware= option. The list separator for
630 the options can now be changed in the usual way. Bug 68.
632 JH/10 The smtp_receive_timeout main option is now expanded before use.
634 JH/11 The incoming_interface log option now also enables logging of the
635 local interface on delivery outgoing connections.
637 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
638 if the interface and destination host and port all match.
640 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
643 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
644 Patch from Andrew Lewis.
646 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
647 now supports optional time-restrictions, weighting, and priority
648 modifiers per server. Patch originally by <rommer@active.by>.
650 JH/16 The spamd_address main option now supports a mixed list of local
651 and remote servers. Remote servers can be IPv6 addresses, and
652 specify a port-range.
654 JH/17 Bug 68: The spamd_address main option now supports an optional
655 timeout value per server.
657 JH/18 Bug 1581: Router and transport options headers_add/remove can
658 now have the list separator specified.
660 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
663 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
666 JH/21 Support for the A6 type of dns record is withdrawn.
668 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
669 rather than the verbs used.
671 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
672 from 255 to 1024 chars.
674 JH/24 Verification callouts now attempt to use TLS by default.
676 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
677 are generic router options now. The defaults didn't change.
679 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
680 Original patch from Alexander Shikoff, worked over by JH.
682 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
683 files if ZCAT_COMMAND is not executable.
685 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
687 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
689 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
690 Normally benign, it bites when the pair was led to by a CNAME;
691 modern usage is to not canonicalize the domain to a CNAME target
692 (and we were inconsistent anyway for A-only vs AAAA+A).
694 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
696 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
697 when evaluating $sender_host_dnssec.
699 JH/31 Check the HELO verification lookup for DNSSEC, adding new
700 $sender_helo_dnssec variable.
702 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
704 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
706 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
708 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
709 documented as working, but never had. Support all but $spam_report.
711 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
712 added for tls authenticator.
714 HS/03 Add perl_taintmode main config option
720 TL/01 When running the test suite, the README says that variables such as
721 no_msglog_check are global and can be placed anywhere in a specific
722 test's script, however it was observed that placement needed to be near
723 the beginning for it to behave that way. Changed the runtest perl
724 script to read through the entire script once to detect and set these
725 variables, reset to the beginning of the script, and then run through
726 the script parsing/test process like normal.
728 TL/02 The BSD's have an arc4random API. One of the functions to induce
729 adding randomness was arc4random_stir(), but it has been removed in
730 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
731 function when detected.
733 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
734 cause callback expansion.
736 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
737 syntax errors in an expansion can be treated as a string instead of
738 logging or causing an error, due to the internal use of bool_lax
739 instead of bool when processing it.
741 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
742 server certificates when making smtp deliveries.
744 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
746 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
748 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
750 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
751 Merged patch from Sebastian Wiedenroth.
753 JH/05 Fix results-pipe from transport process. Several recipients, combined
754 with certificate use, exposed issues where response data items split
755 over buffer boundaries were not parsed properly. This eventually
756 resulted in duplicates being sent. This issue only became common enough
757 to notice due to the introduction of connection certificate information,
758 the item size being so much larger. Found and fixed by Wolfgang Breyha.
760 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
761 size buffer was used, resulting in syntax errors when an expansion
764 JH/07 Add support for directories of certificates when compiled with a GnuTLS
765 version 3.3.6 or later.
767 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
768 is EXPERIMENTAL_EVENT, the main-configuration and transport options
769 both become "event_action", the variables become $event_name, $event_data
770 and $event_defer_errno. There is a new variable $verify_mode, usable in
771 routers, transports and related events. The tls:cert event is now also
772 raised for inbound connections, if the main configuration event_action
775 TL/06 In test suite, disable OCSP for old versions of openssl which contained
776 early OCSP support, but no stapling (appears to be less than 1.0.0).
778 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
779 server certificate names available under the smtp transport option
780 "tls_verify_cert_hostname" now do not permit multi-component wildcard
783 JH/10 Time-related extraction expansions from certificates now use the main
784 option "timezone" setting for output formatting, and are consistent
785 between OpenSSL and GnuTLS compilations. Bug 1541.
787 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
788 encoded parameter in the incoming message. Bug 1558.
790 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
791 include certificate info, eximon was claiming there were spoolfile
794 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
796 JH/14 Log delivery-related information more consistently, using the sequence
797 "H=<name> [<ip>]" wherever possible.
799 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
800 are problematic for Debian distribution, omit them from the release
803 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
805 JH/16 Fix string representation of time values on 64bit time_t architectures.
808 JH/17 Fix a null-indirection in certextract expansions when a nondefault
809 output list separator was used.
814 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
815 checkers that were complaining about end of non-void function with no
818 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
819 This was a regression introduced in 4.83 by another bugfix.
821 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
823 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
824 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
830 TF/01 Correctly close the server side of TLS when forking for delivery.
832 When a message was received over SMTP with TLS, Exim failed to clear up
833 the incoming connection properly after forking off the child process to
834 deliver the message. In some situations the subsequent outgoing
835 delivery connection happened to have the same fd number as the incoming
836 connection previously had. Exim would try to use TLS and fail, logging
837 a "Bad file descriptor" error.
839 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
840 utilities have not been installed.
842 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
843 temporary space as the ACL may create new global variables.
845 TL/01 LDAP support uses per connection or global context settings, depending
846 upon the detected version of the libraries at build time.
848 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
849 to extract and use the src ip:port in logging and expansions as if it
850 were a direct connection from the outside internet. PPv2 support was
851 updated based on HAProxy spec change in May 2014.
853 JH/02 Add ${listextract {number}{list}{success}{fail}}.
855 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
856 Properly escape header and check for NULL return.
858 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
861 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
863 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
864 characters in header names, implemented as a verify condition.
865 Contributed by Michael Fischer v. Mollard.
867 TL/05 Rename SPF condition results err_perm and err_temp to standardized
868 results permerror and temperror. Previous values are deprecated but
869 still accepted. In a future release, err_perm and err_temp will be
870 completely removed, which will be a backward incompatibility if the
871 ACL tests for either of these two old results. Patch contributed by
872 user bes-internal on the mailing list.
874 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
876 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
877 selectors, in both main and reject logs.
879 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
882 JH/07 Add malware type "sock" for talking to simple daemon.
884 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
886 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
887 routers/transports under cutthrough routing.
889 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
890 numbers. Touch up "bool" conditional to keep the same definition.
892 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
894 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
896 JH/12 Expand items in router/transport headers_add or headers_remove lists
897 individually rather than the list as a whole. Bug 1452.
899 Required for reasonable handling of multiple headers_ options when
900 they may be empty; requires that headers_remove items with embedded
901 colons must have them doubled (or the list-separator changed).
903 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
904 view the policy declared in the DMARC record. Currently, $dmarc_status
905 is a combined value of both the record presence and the result of the
908 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
910 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
911 dnslookup router and the smtp transport (applying to the forward
914 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
915 of ldap servers used for a specific lookup. Patch provided by Heiko
918 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
919 New variable $lookup_dnssec_authenticated for observability.
921 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
922 Patch submitted by Lars Timman.
924 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
926 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
927 Requires trusted mode and valid format message id, aborts otherwise.
928 Patch contributed by Heiko Schlichting.
930 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
931 certextract with support for various fields. Bug 1358.
933 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
934 is requested by default, modifiable by smtp transport option
937 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
938 operate on certificate variables to give certificate fingerprints
939 Also new ${sha256:cert_variable}.
941 JH/23 The PRDR feature is moved from being Experimental into the mainline.
943 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
944 Christian Aistleitner.
946 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
948 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
949 file. Patch from Wolfgang Breyha.
951 JH/25 Expand the coverage of the delivery $host and $host_address to
952 client authenticators run in verify callout. Bug 1476.
954 JH/26 Port service names are now accepted for tls_on_connect_ports, to
955 align with daemon_smtp_ports. Bug 72.
957 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
958 support and error reporting did not work properly.
960 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
961 and is readable. Patch from Andrew Colin Kissa.
963 TL/14 Enhance documentation of ${run expansion and how it parses the
964 commandline after expansion, particularly in the case when an
965 unquoted variable expansion results in an empty value.
967 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
969 PP/02 Fix internal collision of T_APL on systems which support RFC3123
970 by renaming away from it. Addresses GH issue 15, reported by
973 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
975 TL/15 SECURITY: prevent double expansion in math comparison functions
976 (can expand unsanitized data). Not remotely exploitable.
983 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
985 PP/02 Make -n do something, by making it not do something.
986 When combined with -bP, the name of an option is not output.
988 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
991 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
992 $sender_host_name and config options to manage this, and basic check
995 PP/05 DSCP support for outbound connections and control modifier for inbound.
997 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
998 (Only plugin which currently uses this is kerberos4, which nobody should
999 be using, but we should make it available and other future plugins might
1000 conceivably use it, even though it would break NAT; stuff *should* be
1001 using channel bindings instead).
1003 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1004 name; added for Sendmail compatibility; requires admin caller.
1005 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1006 just ignore it); requires trusted caller.
1007 Also parse but ignore: -Ac -Am -X<logfile>
1010 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1012 TL/02 Add +smtp_confirmation as a default logging option.
1014 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1015 Patch by Magnus Holmgren from 2007-02-20.
1017 TL/04 Bugzilla 1281 - Spec typo.
1018 Bugzilla 1283 - Spec typo.
1019 Bugzilla 1290 - Spec grammar fixes.
1021 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1023 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1025 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1026 mailing list by by Dmitry Isaikin.
1028 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1030 JH/02 Support "G" suffix to numbers in ${if comparisons.
1032 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1034 NM/01 Bugzilla 1197 - Spec typo
1035 Bugzilla 1196 - Spec examples corrections
1037 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1039 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1040 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1043 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1044 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1046 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1047 "acl {{name}{arg}...}", and optional args on acl condition
1050 JH/05 Permit multiple router/transport headers_add/remove lines.
1052 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1054 JH/07 Avoid using a waiting database for a single-message-only transport.
1055 Performance patch from Paul Fisher. Bugzilla 1262.
1057 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1060 JH/09 Add $headers_added variable, with content from use of ACL modifier
1061 add_header (but not yet added to the message). Bugzilla 199.
1063 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1064 Pulled from Bugzilla 817 by Wolfgang Breyha.
1066 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1068 (nb: this is the same fix as in Exim 4.80.1)
1070 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1073 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1076 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1078 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1079 advertise SMTP AUTH mechanism to us, instead of a generic
1080 protocol violation error. Also, make Exim more robust to bad
1081 data from the Dovecot auth socket.
1083 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1085 When a queue runner is handling a message, Exim first routes the
1086 recipient addresses, during which it prunes them based on the retry
1087 hints database. After that it attempts to deliver the message to
1088 any remaining recipients. It then updates the hints database using
1091 So if a recipient address works intermittently, it can get repeatedly
1092 deferred at routing time. The retry hints record remains fresh so the
1093 address never reaches the final cutoff time.
1095 This is a fairly common occurrence when a user is bumping up against
1096 their storage quota. Exim had some logic in its local delivery code
1097 to deal with this. However it did not apply to per-recipient defers
1098 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1100 This change adds a proper retry rule check during routing so that the
1101 final cutoff time is checked against the message's age. We only do
1102 this check if there is an address retry record and there is not a
1103 domain retry record; this implies that previous attempts to handle
1104 the address had the retry_use_local_parts option turned on. We use
1105 this as an approximation for the destination being like a local
1106 delivery, as in LMTP.
1108 I suspect this new check makes the old local delivery cutoff check
1109 redundant, but I have not verified this so I left the code in place.
1111 TF/02 Correct gecos expansion when From: is a prefix of the username.
1113 Test 0254 submits a message to Exim with the header
1117 When I ran the test suite under the user fanf2, Exim expanded
1118 the header to contain my full name, whereas it should have added
1119 a Resent-Sender: header. It erroneously treats any prefix of the
1120 username as equal to the username.
1122 This change corrects that bug.
1124 GF/01 DCC debug and logging tidyup
1125 Error conditions log to paniclog rather than rejectlog.
1126 Debug lines prefixed by "DCC: " to remove any ambiguity.
1128 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1130 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1131 Bug spotted by Jeremy Harris; was flawed since initial commit.
1132 Would have resulted in OCSP responses post-SNI triggering an Exim
1133 NULL dereference and crash.
1135 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1137 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1138 Bug detection, analysis and fix by Samuel Thibault.
1139 Bugzilla 1331, Debian bug #698092.
1141 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1143 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1144 Server implementation by Todd Lyons, client by JH.
1145 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1146 config variable "prdr_enable" controls whether the server
1147 advertises the facility. If the client requests PRDR a new
1148 acl_data_smtp_prdr ACL is called once for each recipient, after
1149 the body content is received and before the acl_smtp_data ACL.
1150 The client is controlled by both of: a hosts_try_prdr option
1151 on the smtp transport, and the server advertisement.
1152 Default client logging of deliveries and rejections involving
1153 PRDR are flagged with the string "PRDR".
1155 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1156 fclose(). Diagnosis by Todd Lyons.
1158 PP/17 Update configure.default to handle IPv6 localhost better.
1159 Patch by Alain Williams (plus minor tweaks).
1162 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1163 This is now consistent with GnuTLS, and is now documented: the
1164 previous undocumented portable approach to treating the option as
1165 unset was to force an expansion failure. That still works, and
1166 an empty string is now equivalent.
1168 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1169 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1170 not performing validation itself.
1172 PP/20 Added force_command boolean option to pipe transport.
1173 Patch from Nick Koston, of cPanel Inc.
1175 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1178 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1180 PP/21 Fix eximon continuous updating with timestamped log-files.
1181 Broken in a format-string cleanup in 4.80, missed when I repaired the
1182 other false fix of the same issue.
1183 Report and fix from Heiko Schlichting.
1186 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1187 Report from Prashanth Katuri.
1189 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1190 It's SecureTransport, so affects any MacOS clients which use the
1191 system-integrated TLS libraries, including email clients.
1193 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1194 using a MIME ACL for non-SMTP local injection.
1195 Report and assistance in diagnosis by Warren Baker.
1197 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1199 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1201 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1202 last id that failed so it may be referenced in subsequent ACL's.
1204 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1207 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1208 ldap library initialization, allowing self-signed CA's to be
1209 used. Also properly sets require_cert option later in code by
1210 using NULL (global ldap config) instead of ldap handle (per
1211 session). Bug diagnosis and testing by alxgomz.
1213 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1214 the src/util/ subdirectory.
1216 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1217 renamed to Transport Post Delivery Action by Jeremy Harris, as
1220 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1221 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1222 redis_servers = needs to be configured which will be used by the redis
1223 lookup. Patch from Warren Baker, of The Packet Hub.
1225 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1227 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1228 hostname or reverse DNS when processing a host list. Used suggestions
1229 from multiple comments on this bug.
1231 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1233 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1234 Missed a few lines, added it to make the runtest require no keyboard
1237 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1238 contains upper case chars. Make router use caseful_local_part.
1240 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1241 support when GnuTLS has been built with p11-kit.
1247 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1249 This, or similar/improved, will also be change PP/11 of 4.82.
1255 PP/01 Handle short writes when writing local log-files.
1256 In practice, only affects FreeBSD (8 onwards).
1257 Bugzilla 1053, with thanks to Dmitry Isaikin.
1259 NM/01 Bugzilla 949 - Documentation tweak
1261 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1264 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1266 PP/02 Implemented gsasl authenticator.
1268 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1270 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1271 `pkg-config foo` for cflags/libs.
1273 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1274 with rest of GSASL and with heimdal_gssapi.
1276 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1277 `pkg-config foo` for cflags/libs for the TLS implementation.
1279 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1280 properties get this fed in as external SSF. A number of robustness
1281 and debugging improvements to the cyrus_sasl authenticator.
1283 PP/08 cyrus_sasl server now expands the server_realm option.
1285 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1286 Patch by Jeremy Harris.
1288 PP/10 Added dbmjz lookup type.
1290 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1292 PP/12 MAIL args handles TAB as well as SP, for better interop with
1293 non-compliant senders.
1294 Analysis and variant patch by Todd Lyons.
1296 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1297 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1298 Patch from Dirk Mueller <dmueller@suse.com>
1300 PP/13 tls_peerdn now print-escaped for spool files.
1301 Observed some $tls_peerdn in wild which contained \n, which resulted
1302 in spool file corruption.
1304 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1305 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1306 or write after TLS renegotiation, which otherwise led to messages
1309 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1310 as a tracking header (ie: a signed header comes before the signature).
1311 Patch from Wolfgang Breyha.
1313 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1314 comma-sep list; embedded commas doubled.
1316 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1318 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1320 Report and patch from Dmitry Banschikov.
1322 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1323 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1324 failures. We appear to now support TLS1.1+ with Exim.
1326 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1327 lets Exim select keys and certificates based upon TLS SNI from client.
1328 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1329 before an outbound SMTP session. New log_selector, +tls_sni.
1331 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1332 NULL dereference. Report and patch from Alun Jones.
1334 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1335 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1336 Not seeing resolver debug output on NetBSD, but suspect this is a
1337 resolver implementation change.
1339 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1340 Left warnings. Added "eximon gdb" invocation mode.
1342 PP/21 Defaulting "accept_8bitmime" to true, not false.
1344 PP/22 Added -bw for inetd wait mode support.
1346 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1347 locate the relevant includes and libraries. Made this the default.
1349 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1350 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1352 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1353 This may cause build issues on older platforms.
1355 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1356 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1357 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1358 Added SNI support via GnuTLS too.
1359 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1361 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1363 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1364 Applied second patch from Janne, implementing suggestion to default
1365 multiple-strings-in-record handling to match SPF spec.
1367 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1369 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1370 read-only, out of scope).
1371 Patch from Wolfgang Breyha, report from Stuart Northfield.
1373 PP/29 Fix three issues highlighted by clang analyser static analysis.
1374 Only crash-plausible issue would require the Cambridge-specific
1375 iplookup router and a misconfiguration.
1376 Report from Marcin Mirosław.
1378 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1380 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1381 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1382 As part of this, removing so much warning spew let me fix some minor
1383 real issues in debug logging.
1385 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1386 assignment on my part. Fixed.
1388 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1389 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1390 Janne Snabb (who went above and beyond: thank you).
1392 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1393 string otherwise requires a connection and a bunch more work and it's
1394 relatively easy to get wrong. Should also expose TLS library linkage
1397 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1398 64-bit ${eval} (JH/03).
1400 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1401 GNU libc to support some of the 64-bit stuff, should not lead to
1402 conflicts. Defined before os.h is pulled in, so if a given platform
1403 needs to override this, it can.
1405 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1406 protection layer was required, which is not implemented.
1407 Bugzilla 1254, patch from Wolfgang Breyha.
1409 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1410 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1411 tls_dhparam take prime identifiers. Also unbreak combination of
1412 OpenSSL+DH_params+TLSSNI.
1414 PP/39 Disable SSLv2 by default in OpenSSL support.
1420 PP/01 Solaris build fix for Oracle's LDAP libraries.
1421 Bugzilla 1109, patch from Stephen Usher.
1423 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1425 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1428 TF/02 Fix a couple more cases where we did not log the error message
1429 when unlink() failed. See also change 4.74-TF/03.
1431 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1432 lock up or crash if it happened to be inside a call to libc when it
1433 got a SIGUSR1 from exiwhat.
1435 The SIGUSR1 handler appends the current process status to the process
1436 log which is later printed by exiwhat. It used to use the general
1437 purpose logging code to do this, but several functions it calls are
1438 not safe for signals.
1440 The new output code in the SIGUSR1 handler is specific to the process
1441 log, and simple enough that it's easy to inspect for signal safety.
1442 Removing some special cases also simplifies the general logging code.
1443 Removing the spurious timestamps from the process log simplifies
1446 TF/04 Improved ratelimit ACL condition.
1448 The /noupdate option has been deprecated in favour of /readonly which
1449 has clearer semantics. The /leaky, /strict, and /readonly update modes
1450 are mutually exclusive. The update mode is no longer included in the
1451 database key; it just determines when the database is updated. (This
1452 means that when you upgrade Exim will forget old rate measurements.)
1454 Exim now checks that the per_* options are used with an update mode that
1455 makes sense for the current ACL. For example, when Exim is processing a
1456 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1457 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1458 must specify per_mail/readonly. If you omit the update mode it defaults to
1459 /leaky where that makes sense (as before) or /readonly where required.
1461 The /noupdate option is now undocumented but still supported for
1462 backwards compatibility. It is equivalent to /readonly except that in
1463 ACLs where /readonly is required you may specify /leaky/noupdate or
1464 /strict/noupdate which are treated the same as /readonly.
1466 A useful new feature is the /count= option. This is a generalization
1467 of the per_byte option, so that you can measure the throughput of other
1468 aggregate values. For example, the per_byte option is now equivalent
1469 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1471 The per_rcpt option has been generalized using the /count= mechanism
1472 (though it's more complicated than the per_byte equivalence). When it is
1473 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1474 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1475 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1476 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1477 Note that using per_rcpt with a non-readonly update mode in more than
1478 one ACL will cause the recipients to be double-counted. (The per_mail
1479 and per_byte options don't have this problem.)
1481 The handling of very low rates has changed slightly. If the computed rate
1482 is less than the event's count (usually one) then this event is the first
1483 after a long gap. In this case the rate is set to the same as this event's
1484 count, so that the first message of a spam run is counted properly.
1486 The major new feature is a mechanism for counting the rate of unique
1487 events. The new per_addr option counts the number of different
1488 recipients that someone has sent messages to in the last time period. It
1489 behaves like per_rcpt if all the recipient addresses are different, but
1490 duplicate recipient addresses do not increase the measured rate. Like
1491 the /count= option this is a general mechanism, so the per_addr option
1492 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1493 example, measure the rate that a client uses different sender addresses
1494 with the options per_mail/unique=$sender_address. There are further
1495 details in the main documentation.
1497 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1499 TF/06 Removed a few PCRE remnants.
1501 TF/07 Automatically extract Exim's version number from tags in the git
1502 repository when doing development or release builds.
1504 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1505 Bugzilla 879. Patch from Paul Fisher.
1507 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1508 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1511 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1513 PP/05 Variable $av_failed, true if the AV scanner deferred.
1514 Bugzilla 1078. Patch from John Horne.
1516 PP/06 Stop make process more reliably on build failure.
1517 Bugzilla 1087. Patch from Heiko Schlittermann.
1519 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1520 Bugzilla 1089. Patch from Heiko Schlittermann.
1522 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1523 Bugzilla 1131. Patch from Holger Weiß.
1525 PP/09 Handle IPv6 addresses with SPF.
1526 Bugzilla 860. Patch from Wolfgang Breyha.
1528 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1530 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1533 PP/11 match_* no longer expand right-hand-side by default.
1534 New compile-time build option, EXPAND_LISTMATCH_RHS.
1535 New expansion conditions, "inlist", "inlisti".
1537 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1539 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1541 PP/14 fix log_write() format string regression from TF/03.
1542 Bugzilla 1152. Patch from Dmitry Isaikin.
1548 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1550 PP/02 Harmonised TLS library version reporting; only show if debugging.
1551 Layout now matches that introduced for other libraries in 4.74 PP/03.
1553 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1555 PP/04 New "dns_use_edns0" global option.
1557 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1560 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1561 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1563 TK/01 Updated PolarSSL code to 0.14.2.
1564 Bugzilla 1097. Patch from Andreas Metzler.
1566 PP/07 Catch divide-by-zero in ${eval:...}.
1567 Fixes bugzilla 1102.
1569 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1572 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1573 format-string attack -- SECURITY: remote arbitrary code execution.
1575 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1576 time unintentionally subject to list matching rules, letting the header
1577 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1578 arbitrary string expansion). This allowed for information disclosure.
1580 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1581 INT_MIN/-1 -- value coerced to INT_MAX.
1587 NM/01 Workaround for PCRE version dependency in version reporting
1590 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1591 This fixes portability to compilers other than gcc, notably
1592 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1594 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1595 makefiles for portability to HP-UX and POSIX correctness.
1597 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1598 Also via indented variable definition in the Makefile.
1599 (Debugging by Oliver Heesakkers).
1601 PP/02 Restore caching of spamd results with expanded spamd_address.
1602 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1604 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1605 Improves build reliability. Fix from: Frank Elsner
1607 NM/02 Fix wide character breakage in the rfc2047 coding
1608 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1610 NM/03 Allow underscore in dnslist lookups
1611 Fixes bug 1026. Patch from Graeme Fowler
1613 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1614 Code patches from Adam Ciarcinski of NetBSD.
1616 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1619 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1620 is logged, to avoid truncation. Patch from John Horne.
1622 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1623 Patch from Jakob Hirsch.
1625 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1626 SQL string expansion failure details.
1627 Patch from Andrey Oktyabrski.
1629 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1630 Patch from Simon Arlott.
1632 PP/09 New lookups functionality failed to compile on old gcc which rejects
1633 extern declarations in function scope.
1634 Patch from Oliver Fleischmann
1636 PP/10 Use sig_atomic_t for flags set from signal handlers.
1637 Check getgroups() return and improve debugging.
1638 Fixed developed for diagnosis in bug 927 (which turned out to be
1641 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1642 Patch from Mark Zealey.
1644 PP/12 Bugzilla 1056: Improved spamd server selection.
1645 Patch from Mark Zealey.
1647 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1648 Based on patch from Heiko Schlittermann.
1650 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1651 Patch from Uwe Doering, sign-off by Michael Haardt.
1653 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1654 variable declaration deep within a block. Bug and patch from
1657 PP/15 lookups-Makefile IRIX compatibility coercion.
1659 PP/16 Make DISABLE_DKIM build knob functional.
1661 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1662 Patch by Simon Arlott
1664 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1665 variable argument macros. Our copy now differs from upstream.
1671 TF/01 Failure to get a lock on a hints database can have serious
1672 consequences so log it to the panic log.
1674 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1675 controlled using the smtp_confirmation log selector.
1677 TF/03 Include the error message when we fail to unlink a spool file.
1679 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1680 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1681 for maintaining out-of-tree patches for some time.
1683 PP/01 Bugzilla 139: Documentation and portability issues.
1684 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1685 Handle per-OS dynamic-module compilation flags.
1687 PP/02 Let /dev/null have normal permissions.
1688 The 4.73 fixes were a little too stringent and complained about the
1689 permissions on /dev/null. Exempt it from some checks.
1690 Reported by Andreas M. Kirchwitz.
1692 PP/03 Report version information for many libraries, including
1693 Exim version information for dynamically loaded libraries. Created
1694 version.h, now support a version extension string for distributors
1695 who patch heavily. Dynamic module ABI change.
1697 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1698 privilege escalation vulnerability whereby the Exim run-time user
1699 can cause root to append content of the attacker's choosing to
1702 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1705 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1706 If dropping privileges for untrusted macros, we disabled normal logging
1707 on the basis that it would fail; for the Exim run-time user, this is not
1708 the case, and it resulted in successful deliveries going unlogged.
1709 Fixed. Reported by Andreas Metzler.
1715 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1716 only prepend for the Resent-* case. Fixes regression introduced in
1717 Exim 4.70 by NM/22 for Bugzilla 607.
1719 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1720 increasing numbers of administrators be bitten by this.
1722 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1724 PP/03 Bugzilla 994: added openssl_options main configuration option.
1726 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1728 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1730 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1732 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1733 without a peer certificate, leading to a segfault because of an
1734 assumption that peers always have certificates. Be a little more
1735 paranoid. Problem reported by Martin Tscholak.
1737 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1738 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1739 NB: ClamAV planning to remove STREAM in "middle of 2010".
1740 CL also introduces -bmalware, various -d+acl logging additions and
1741 more caution in buffer sizes.
1743 PP/09 Implemented reverse_ip expansion operator.
1745 PP/10 Bugzilla 937: provide a "debug" ACL control.
1747 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1749 PP/12 Bugzilla 973: Implement --version.
1751 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1753 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1755 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1757 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1758 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1759 ignore trailing whitespace.
1761 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1763 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1766 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1767 Notification from Dr Andrew Aitchison.
1769 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1770 ExtendedDetectionInfo response format.
1771 Notification from John Horne.
1773 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1776 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1777 XSL and documented dependency on system catalogs, with examples of how
1780 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1783 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1784 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1785 configuration file which is writeable by the Exim user or group.
1787 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1788 of configuration files to cover files specified with the -C option if
1789 they are going to be used with root privileges, not just the default
1792 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1793 option (effectively making it always true).
1795 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1796 files to be used while preserving root privileges.
1798 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1799 that rogue child processes cannot use them.
1801 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1802 run-time user, instead of root.
1804 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1805 Exim run-time user without dropping privileges.
1807 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1808 result string, instead of calling string_vformat() twice with the same
1811 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1812 for other users. Others should always drop root privileges if they use
1813 -C on the command line, even for a whitelisted configure file.
1815 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1817 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1823 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1824 $data_path, and $header_path variables; fixed documentation bugs and
1827 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1828 exipick to access non-standard spools, including the "frozen" queue
1831 NM/01 Bugzilla 965: Support mysql stored procedures.
1832 Patch from Alain Williams
1834 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1836 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1837 Patch from Andreas Metzler
1839 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1840 Patch from Kirill Miazine
1842 NM/05 Bugzilla 671: Added umask to procmail example.
1844 JJ/03 installed exipick 20100323.0, fixing doc bug
1846 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1847 directory. Notification and patch from Dan Rosenberg.
1849 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1851 TK/02 Improve log output when DKIM signing operation fails.
1853 MH/01 Treat the transport option dkim_domain as a colon separated
1854 list, not as a single string, and sign the message with each element,
1855 omitting multiple occurences of the same signer.
1857 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1858 Bugzilla 985, 986. Patch by Simon Arlott
1860 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1861 Patch by Simon Arlott
1863 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1864 MBX locking. Notification from Dan Rosenberg.
1870 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1872 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1874 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1876 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1878 NM/04 Bugzilla 915: Flag broken perl installation during build.
1884 TK/01 Added patch by Johannes Berg that expands the main option
1885 "spamd_address" if it starts with a dollar sign.
1887 TK/02 Write list of recipients to X-Envelope-Sender header when building
1888 the mbox-format spool file for content scanning (suggested by Jakob
1891 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1892 (http://www.dcc-servers.net/) support via dccifd. Activated by
1893 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1895 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1896 by Mark Daniel Reidel <mr@df.eu>.
1898 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1899 When building exim an external PCRE library is now needed -
1900 PCRE is a system library on the majority of modern systems.
1901 See entry on PCRE_LIBS in EDITME file.
1903 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1904 conversation. Added nologin parameter to request.
1905 Patch contributed by Kirill Miazine.
1907 TF/01 Do not log submission mode rewrites if they do not change the address.
1909 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1911 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1912 log files in place. Contributed by Roberto Lima.
1914 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1916 TF/03 Bugzilla 615: When checking the local_parts router precondition
1917 after a local_part_suffix or local_part_prefix option, Exim now
1918 does not use the address's named list lookup cache, since this
1919 contains cached lookups for the whole local part.
1921 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1922 Robert Millan. Documentation is in experimental-spec.txt.
1924 TF/04 Bugzilla 668: Fix parallel build (make -j).
1926 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1928 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1929 Patch provided by Jan Srzednicki.
1931 TF/05 Leading white space used to be stripped from $spam_report which
1932 wrecked the formatting. Now it is preserved.
1934 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1935 that they are available at delivery time.
1937 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1939 TF/08 TLS error reporting now respects the incoming_interface and
1940 incoming_port log selectors.
1942 TF/09 Produce a more useful error message if an SMTP transport's hosts
1943 setting expands to an empty string.
1945 NM/06 Bugzilla 744: EXPN did not work under TLS.
1946 Patch provided by Phil Pennock.
1948 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1949 Patch provided by Richard Godbee.
1951 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1952 acl_smtp_notquit, added index entry.
1954 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1955 Patch provided by Eugene Bujak.
1957 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1958 accept(). Patch provided by Maxim Dounin.
1960 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1961 Patch provided by Phil Pennock.
1963 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1965 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1966 Patch provided by Brad "anomie" Jorsch.
1968 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1969 Patch provided by Dean Brooks.
1971 TK/05 Add native DKIM support (does not depend on external libraries).
1973 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1974 Patch provided by Graeme Fowler.
1976 NM/16 Bugzilla 851: Documentation example syntax fix.
1978 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1980 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1983 NM/19 Bugzilla 745: TLS version reporting.
1984 Patch provided by Phil Pennock.
1986 NM/20 Bugzilla 167: bool: condition support.
1987 Patch provided by Phil Pennock.
1989 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1990 clients. Patch provided by Phil Pennock.
1992 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1993 Patch provided by Brad "anomie" Jorsch.
1995 NM/23 Bugzilla 687: Fix misparses in eximstats.
1996 Patch provided by Heiko Schlittermann.
1998 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1999 Patch provided by Heiko Schlittermann.
2001 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2002 plus update to original patch.
2004 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2006 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2007 Patch provided by David Brownlee.
2009 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2011 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2013 NM/30 Bugzilla 888: TLS documentation bugfixes.
2015 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2017 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2018 Unlike the original bugzilla I have changed all shell scripts in src tree.
2020 NM/33 Bugzilla 898: Transport filter timeout fix.
2021 Patch by Todd Rinaldo.
2023 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2024 Patch by Serge Demonchaux.
2026 NM/35 Bugzilla 39: Base64 decode bug fixes.
2027 Patch by Jakob Hirsch.
2029 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2031 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2033 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2035 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2041 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2042 ALT-N's libdkim that I have put here:
2043 http://duncanthrax.net/exim-experimental/
2045 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2046 were called 'true' and it seems that C99 defines that as a reserved
2047 keyword to be used with 'bool' variable types. That means you could
2048 not include C99-style headers which use bools without triggering
2049 build errors in sieve.c.
2051 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2052 as mailq or other aliases. Changed the --help handling significantly
2053 to do whats expected. exim_usage() emits usage/help information.
2055 SC/01 Added the -bylocaldomain option to eximstats.
2057 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2059 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2061 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2067 PH/01 Another patch from the Sieve maintainer.
2069 PH/02 When an IPv6 address is converted to a string for single-key lookup
2070 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2071 dots are used instead of colons so that keys in lsearch files need not
2072 contain colons. This was done some time before quoting was made available
2073 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2074 (notated using the quote facility) so as to distinguish them from IPv4
2075 keys. This meant that lookups for IP addresses in host lists did not work
2076 for iplsearch lookups.
2078 This has been fixed by arranging for IPv6 addresses to be expressed with
2079 colons if the lookup type is iplsearch. This is not incompatible, because
2080 previously such lookups could never work.
2082 The situation is now rather anomalous, since one *can* have colons in
2083 ordinary lsearch keys. However, making the change in all cases is
2084 incompatible and would probably break a number of configurations.
2086 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2089 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2090 conversion specification without a maximum field width, thereby enabling
2091 a rogue spamd server to cause a buffer overflow. While nobody in their
2092 right mind would setup Exim to query an untrusted spamd server, an
2093 attacker that gains access to a server running spamd could potentially
2094 exploit this vulnerability to run arbitrary code as the Exim user.
2096 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2097 $primary_hostname instead of what libspf2 thinks the hosts name is.
2099 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2100 a directory entry by the name of the lookup key. Previously, if a
2101 symlink pointed to a non-existing file or a file in a directory that
2102 Exim lacked permissions to read, a lookup for a key matching that
2103 symlink would fail. Now it is enough that a matching directory entry
2104 exists, symlink or not. (Bugzilla 503.)
2106 PH/03 The body_linecount and body_zerocount variables are now exported in the
2109 PH/04 Added the $dnslist_matched variable.
2111 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2112 This means they are set thereafter only if the connection becomes
2115 PH/06 Added the client_condition to authenticators so that some can be skipped
2116 by clients under certain conditions.
2118 PH/07 The error message for a badly-placed control=no_multiline_responses left
2119 "_responses" off the end of the name.
2121 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2123 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2124 (without spaces) instead of just copying the configuration text.
2126 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2128 PH/11 Added $max_received_linelength.
2130 PH/12 Added +ignore_defer and +include_defer to host lists.
2132 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2133 way in which PCRE > 7.0 is built.
2135 PH/14 Implemented queue_only_load_latch.
2137 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2138 MAIL command. The effect was to mangle the value on 64-bit systems.
2140 PH/16 Another patch from the Sieve maintainer.
2142 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2144 PH/18 If a system quota error occurred while trying to create the file for
2145 a maildir delivery, the message "Mailbox is full" was not appended to the
2146 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2147 only to a quota excession during the actual writing of the file.
2149 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2150 characters?) which causes problems in log lines. The DN values are now
2151 passed through string_printing() before being added to log lines.
2153 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2154 and InterBase are left for another time.)
2156 PH/21 Added message_body_newlines option.
2158 PH/22 Guard against possible overflow in moan_check_errorcopy().
2160 PH/23 POSIX allows open() to be a macro; guard against that.
2162 PH/24 If the recipient of an error message contained an @ in the local part
2163 (suitably quoted, of course), incorrect values were put in $domain and
2164 $local_part during the evaluation of errors_copy.
2170 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2171 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2174 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2175 issue a MAIL command.
2177 PH/02 In an ACL statement such as
2179 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2181 if a client was not listed at all, or was listed with a value other than
2182 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2183 the condition was not true (as it should be), so access was not denied.
2184 The bug was that the ! inversion was incorrectly passed on to the second
2185 item. This has been fixed.
2187 PH/03 Added additional dnslists conditions == and =& which are different from
2188 = and & when the dns lookup returns more than one IP address.
2190 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2191 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2193 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2194 FSYNC, which compiles an option called disable_fsync that allows for
2195 bypassing fsync(). The documentation is heavily laced with warnings.
2197 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2199 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2200 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2201 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2202 including adding "make clean"; (3) Added -fPIC when compiling the test
2203 dynamically loaded module, to get rid of a warning.
2205 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2206 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2207 = 0s. The bug is otherwise harmless.
2209 PH/07 There was a bug in the dovecot authenticator such that the value of
2210 $auth1 could be overwritten, and so not correctly preserved, after a
2211 successful authentication. This usually meant that the value preserved by
2212 the server_setid option was incorrect.
2214 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2216 PH/09 Installed PCRE release 7.0.
2218 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2219 run for batched SMTP input. It is now run at the start of every message
2220 in the batch. While fixing this I discovered that the process information
2221 (output by running exiwhat) was not always getting set for -bs and -bS
2222 input. This is fixed, and it now also says "batched" for BSMTP.
2224 PH/11 Added control=no_pipelining.
2226 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2227 patch, slightly modified), and move the expansion of helo_data till after
2228 the connection is made in the smtp transport (so it can use these
2231 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2233 PH/14 Added log_selector = +pid.
2235 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2237 PH/16 Add ${if forany and ${if forall.
2239 PH/17 Added dsn_from option to vary the From: line in DSNs.
2241 PH/18 Flush SMTP output before performing a callout, unless control =
2242 no_callout_flush is set.
2244 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2245 was true (the default) a successful delivery failed to delete the retry
2246 item, thus causing premature timeout of the address. The bug is now
2249 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2251 PH/21 Long custom messages for fakedefer and fakereject are now split up
2252 into multiline responses in the same way that messages for "deny" and
2253 other ACL rejections are.
2255 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2256 with slight modification.
2258 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2259 draft, changing the syntax and factoring some duplicate code".
2261 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2262 for deliveries of the second and subsequent messages over the same SMTP
2265 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2266 ${reduce, with only minor "tidies".
2268 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2270 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2271 expansion side effects.
2273 PH/27 When a message times out after an over-quota error from an Exim-imposed
2274 quota, the bounce message says "mailbox is full". This message was not
2275 being given when it was a system quota that was exceeded. It now should
2278 MH/03 Made $recipients available in local_scan(). local_scan() already has
2279 better access to the recipient list through recipients_list[], but
2280 $recipients can be useful in postmaster-provided expansion strings.
2282 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2283 in the case of a MAIL command with additional options following the
2284 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2285 were accidentally chopped off.
2287 PH/29 SMTP synchronization checks are implemented when a command is read -
2288 there is a check that no more input is waiting when there shouldn't be
2289 any. However, for some commands, a delay in an ACL can mean that it is
2290 some time before the response is written. In this time, more input might
2291 arrive, invalidly. So now there are extra checks after an ACL has run for
2292 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2293 pipelining has not been advertised.
2295 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2297 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2298 returned in the SMTP response when smtp_return_error_details was false.
2299 This has been fixed.
2301 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2302 instead of the C I/O that was originally supplied, because problems were
2303 reported on Solaris.
2305 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2306 Exim which did not show up earlier: it was assuming that a call to
2307 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2308 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2309 was a macro that became an assignment, so it seemed to work. This has
2310 changed to a proper function call with a void return, hence the compile
2311 error. Exim's code has been fixed.
2313 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2316 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2318 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2320 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2321 because it turns out to be too big or there is a timeout) there is no
2322 "Completed" line in the log. When some messages of this type were
2323 selected by exigrep, they were listed as "not completed". Others were
2324 picked up by some special patterns. I have improved the selection
2325 criteria to be more general.
2327 PH/38 The host_find_failed option in the manualroute router can now be set
2328 to "ignore", to completely ignore a host whose IP address cannot be
2329 found. If all hosts are ignored, the behaviour is controlled by the new
2330 host_all_ignored option.
2332 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2333 homing or because of multiple MX records with /mx) generated more than
2334 one IP address, and the following item turned out to be the local host,
2335 all the secondary addresses of the first item were incorrectly removed
2336 from the list, along with the local host and any following hosts (which
2337 is what is supposed to happen).
2339 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2340 whoever called Exim into the -H file. In the case of the daemon it was
2341 behaving confusingly. When first started, it used values for whoever
2342 started the daemon, but after a SIGHUP it used the Exim user (because it
2343 calls itself on a restart). I have changed the code so that it now always
2346 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2347 message are rejected with the same error (e.g. no authentication or bad
2348 sender address), and a DATA command is nevertheless sent (as can happen
2349 with PIPELINING or a stupid MUA), the error message that was given to the
2350 RCPT commands is included in the rejection of the DATA command. This is
2351 intended to be helpful for MUAs that show only the final error to their
2354 PH/42 Another patch from the Sieve maintainer.
2356 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2357 Eximstats - Fixed some broken HTML links and added missing column headers
2359 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2360 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2362 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2364 PH/43 Yet another patch from the Sieve maintainer.
2366 PH/44 I found a way to check for a TCP/IP connection going away before sending
2367 the response to the final '.' that terminates a message, but only in the
2368 case where the client has not sent further data following the '.'
2369 (unfortunately, this is allowed). However, in many cases there won't be
2370 any further data because there won't be any more messages to send. A call
2371 to select() can be used: if it shows that the input is "ready", there is
2372 either input waiting, or the socket has been closed. An attempt to read
2373 the next input character can distinguish the two cases. Previously, Exim
2374 would have sent an OK response which the client would never have see.
2375 This could lead to message repetition. This fix should cure that, at
2376 least in a lot of common cases.
2378 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2379 advertised in response to EHLO.
2385 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2386 fixed by 4.65/MH/01 (is this a record?) are fixed:
2388 (i) An empty string was always treated as zero by the numeric comparison
2389 operators. This behaviour has been restored.
2391 (ii) It is documented that the numeric comparison operators always treat
2392 their arguments as decimal numbers. This was broken in that numbers
2393 starting with 0 were being interpreted as octal.
2395 While fixing these problems I realized that there was another issue that
2396 hadn't been noticed. Values of message_size_limit (both the global option
2397 and the transport option) were treated as octal if they started with 0.
2398 The documentation was vague. These values are now always treated as
2399 decimal, and I will make that clear in the documentation.
2405 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2406 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2409 MH/01 Don't check that the operands of numeric comparison operators are
2410 integers when their expansion is in "skipping" mode (fixes bug
2411 introduced by 4.64-PH/07).
2413 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2414 child addresses, Exim now panics and dies. Previously, because the count
2415 is held in a short int, deliveries were likely to be lost. As such a
2416 large number of recipients for a single message is ridiculous
2417 (performance will be very, very poor), I have chosen to impose a limit
2418 rather than extend the field.
2424 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2425 leftover -K file (the existence of which was triggered by #402).
2426 While we were at it, introduced process PID as part of the -K
2427 filename. This should rule out race conditions when creating
2430 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2431 processing considerably. Previous code took too long for large mails,
2432 triggering a timeout which in turn triggers #401.
2434 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2435 in the DK code in transports.c. sendfile() is not really portable,
2436 hence the _LINUX specificness.
2438 TF/01 In the add_headers option to the mail command in an Exim filter,
2439 there was a bug that Exim would claim a syntax error in any
2440 header after the first one which had an odd number of characters
2443 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2444 callout verification, Exim cached a "reject" for the entire domain. This
2445 is correct for most verifications, but it is not correct for a recipient
2446 verification with use_sender or use_postmaster set, because in that case
2447 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2448 case of MAIL FROM:<> rejection from other early rejections (e.g.
2449 rejection of HELO). When verifying a recipient using a non-null MAIL
2450 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2451 Whatever the result of the callout, the value of the domain cache is
2452 left unchanged (for any other kind of callout, getting as far as trying
2453 RCPT means that the domain itself is ok).
2455 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2458 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2459 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2460 session, the daemon ignores ECONNECT errors and logs others; it now
2461 ignores EPIPE as well.
2463 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2464 (quoted-printable decoding).
2466 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2467 later the small subsequent patch to fix an introduced bug.
2469 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2471 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2473 PH/08 An error is now given if message_size_limit is specified negative.
2475 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2476 to be given (somewhat) arbitrary names.
2478 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2481 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2482 miscellaneous code fixes
2484 PH/10 Added the log_reject_target ACL modifier to specify where to log
2487 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2488 hostname. This is wrong, because it relates to the incoming message (and
2489 probably the interface on which it is arriving) and not to the outgoing
2490 callout (which could be using a different interface). This has been
2491 changed to use the value of the helo_data option from the smtp transport
2492 instead - this is what is used when a message is actually being sent. If
2493 there is no remote transport (possible with a router that sets up host
2494 addresses), $smtp_active_hostname is used.
2496 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2497 tweaks were necessary in order to get it to work (see also 21 below):
2498 (a) The code assumed that strncpy() returns a negative number on buffer
2499 overflow, which isn't the case. Replaced with Exim's string_format()
2501 (b) There were several signed/unsigned issues. I just did the minimum
2502 hacking in of casts. There is scope for a larger refactoring.
2503 (c) The code used strcasecmp() which is not a standard C function.
2504 Replaced with Exim's strcmpic() function.
2505 (d) The code set only $1; it now sets $auth1 as well.
2506 (e) A simple test gave the error "authentication client didn't specify
2507 service in request". It would seem that Dovecot has changed its
2508 interface. Fortunately there's a specification; I followed it and
2509 changed what the client sends and it appears to be working now.
2511 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2514 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2515 address A is aliased to B and C, where B exists and C does not. Without
2516 -v the output is "A verified" because verification stops after a
2517 successful redirection if more than one address is generated. However,
2518 with -v the child addresses are also verified. Exim was outputting "A
2519 failed to verify" and then showing the successful verification for C,
2520 with its parentage. It now outputs "B failed to verify", showing B's
2521 parentage before showing the successful verification of C.
2523 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2524 look up a TXT record in a specific list after matching in a combined
2527 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2528 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2529 they consult the DNS. I had assumed they would set it the way they
2530 wanted; and indeed my experiments on Linux seem to show that in some
2531 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2532 To be on the safe side, however, I have now made the interface to
2533 host_find_byname() similar to host_find_bydns(), with an argument
2534 containing the DNS resolver options. The host_find_byname() function now
2535 sets these options at its start, just as host_find_bydns() does. The smtp
2536 transport options dns_qualify_single and dns_search_parents are passed to
2537 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2538 of host_find_byname() use the default settings of RES_DEFNAMES
2539 (qualify_single) but not RES_DNSRCH (search_parents).
2541 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2542 spool_read_header() do less string testing, by means of a preliminary
2543 switch on the second character of optional "-foo" lines. (This is
2544 overdue, caused by the large number of possibilities that now exist.
2545 Originally there were few.) While I was there, I also converted the
2546 str(n)cmp tests so they don't re-test the leading "-" and the first
2547 character, in the hope this might squeeze out yet more improvement.
2549 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2550 flag allowing group syntax was set by the header_syntax check but not
2551 turned off, possible causing trouble later; (2) The flag was not being
2552 set at all for the header_verify test, causing "group"-style headers to
2553 be rejected. I have now set it in this case, and also caused header_
2554 verify to ignore an empty address taken from a group. While doing this, I
2555 came across some other cases where the code for allowing group syntax
2556 while scanning a header line wasn't quite right (mostly, not resetting
2557 the flag correctly in the right place). These bugs could have caused
2558 trouble for malformed header lines. I hope it is now all correct.
2560 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2561 with the "reply" argument non-NULL. The code, however (which originally
2562 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2563 but it didn't always do it. This confused somebody who was copying the
2564 code for some other use. I have removed all the tests.
2566 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2567 feature that was used to support insecure browsers during the U.S. crypto
2568 embargo. It requires special client support, and Exim is probably the
2569 only MTA that supported it -- and would never use it because real RSA is
2570 always available. This code has been removed, because it had the bad
2571 effect of slowing Exim down by computing (never used) parameters for the
2572 RSA_EXPORT functionality.
2574 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2575 authenticator to fail if there's a tab character in the incoming data
2576 (there should never be unless someone is messing about, as it's supposed
2577 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2578 if the connection is using TLS or if the remote IP is the same as the
2579 local IP, and the "valid-client-cert option" if a client certificate has
2582 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2583 authenticators. This can be used for authorization after authentication
2584 succeeds. (In the case of plaintext, it servers for both authentication
2587 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2588 if any retry times were supplied.
2590 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2591 connection, where there is no client IP address to check. In this
2592 situation, the verify now always succeeds.
2594 PH/25 Applied John Jetmore's -Mset patch.
2596 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2598 PH/27 In a string expansion for a processed (not raw) header when multiple
2599 headers of the same name were present, leading whitespace was being
2600 removed from all of them, but trailing whitespace was being removed only
2601 from the last one. Now trailing whitespace is removed from each header
2602 before concatenation. Completely empty headers in a concatenation (as
2603 before) are ignored.
2605 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2606 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2608 PH/29 [Removed. This was a change that I later backed out, and forgot to
2609 correct the ChangeLog entry (that I had efficiently created) before
2610 committing the later change.]
2612 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2613 address errors (4xx response to RCPT) over the same connection as other
2614 messages routed to the same hosts. Such deliveries are always "forced",
2615 so retry times are not inspected. This resulted in far too many retries
2616 for the affected addresses. The effect occurred only when there were more
2617 hosts than the hosts_max_try setting in the smtp transport when it had
2618 the 4xx errors. Those hosts that it had tried were not added to the list
2619 of hosts for which the message was waiting, so if all were tried, there
2620 was no problem. Two fixes have been applied:
2622 (i) If there are any address or message errors in an SMTP delivery, none
2623 of the hosts (tried or untried) are now added to the list of hosts
2624 for which the message is waiting, so the message should not be a
2625 candidate for sending over the same connection that was used for a
2626 successful delivery of some other message. This seems entirely
2627 reasonable: after all the message is NOT "waiting for some host".
2628 This is so "obvious" that I'm not sure why it wasn't done
2629 previously. Hope I haven't missed anything, but it can't do any
2630 harm, as the worst effect is to miss an optimization.
2632 (ii) If, despite (i), such a delivery is accidentally attempted, the
2633 routing retry time is respected, so at least it doesn't keep
2634 hammering the server.
2636 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2637 in ${readsocket because some servers need this prod.
2639 PH/32 Added some extra debug output when updating a wait-xxx database.
2641 PH/33 The hint "could be header name not terminated by colon", which has been
2642 given for certain expansion errors for a long time, was not being given
2643 for the ${if def:h_colon_omitted{... case.
2645 PH/34 The spec says: "With one important exception, whenever a domain list is
2646 being scanned, $domain contains the subject domain." There was at least
2647 one case where this was not true.
2649 PH/35 The error "getsockname() failed: connection reset by peer" was being
2650 written to the panic log as well as the main log, but it isn't really
2651 panic-worthy as it just means the connection died rather early on. I have
2652 removed the panic log writing for the ECONNRESET error when getsockname()
2655 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2656 runs only) independently of the message's sender address. This meant
2657 that, if the 4xx error was in fact related to the sender, a different
2658 message to the same recipient with a different sender could confuse
2659 things. In particular, this can happen when sending to a greylisting
2660 server, but other circumstances could also provoke similar problems.
2661 I have changed the default so that the retry time for these errors is now
2662 based a combination of the sender and recipient addresses. This change
2663 can be overridden by setting address_retry_include_sender=false in the
2666 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2667 remote server are returned as part of bounce messages. This was not
2668 happening for LMTP over a pipe (the lmtp transport), but now it is the
2669 same for both kinds of LMTP.
2671 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2672 in header lines that were in fact CNAMEs. This is no longer the case.
2674 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2675 and queue runs started by the daemon processed all messages. This has
2676 been fixed so that -R and -S can now usefully be given with -q<time>.
2678 PH/40 Import PCRE release 6.7 (fixes some bugs).
2680 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2682 PH/42 Give an error if -q is specified more than once.
2684 PH/43 Renamed the variables $interface_address and $interface_port as
2685 $received_ip_address and $received_port, to make it clear that these
2686 values apply to message reception, and not to the outgoing interface when
2687 a message is delivered. (The old names remain recognized, of course.)
2689 PH/44 There was no timeout on the connect() call when using a Unix domain
2690 socket in the ${readsocket expansion. There now is.
2692 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2693 be meaningful with "accept".
2695 SC/01 Eximstats V1.43
2696 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2698 SC/02 Eximstats V1.44
2699 Use a glob alias rather than an array ref in the generated
2700 parser. This improves both readability and performance.
2702 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2703 Collect SpamAssassin and rejection statistics.
2704 Don't display local sender or destination tables unless
2705 there is data to show.
2706 Added average volumes into the top table text output.
2708 SC/04 Eximstats V1.46
2709 Collect data on the number of addresses (recipients)
2710 as well as the number of messages.
2712 SC/05 Eximstats V1.47
2713 Added 'Message too big' to the list of mail rejection
2714 reasons (thanks to Marco Gaiarin).
2716 SC/06 Eximstats V1.48
2717 Mainlog lines which have GMT offsets and are too short to
2718 have a flag are now skipped.
2720 SC/07 Eximstats V1.49 (Alain Williams)
2721 Added the -emptyok flag.
2723 SC/08 Eximstats V1.50
2724 Fixes for obtaining the IP address from reject messages.
2726 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2727 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2728 whitespace changes from 4.64-PH/27
2730 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2733 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2734 are found, allow negative numbers in numeric criteria)
2736 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2738 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2739 to match changes made in 4.64-PH/43
2741 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2743 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2744 "domain" is actually an IP address, and force a failure. This locks out
2745 those revolvers/nameservers that support "A-for-A" lookups, in
2746 contravention of the specifications.
2748 PH/48 When a host name was looked up from an IP address, and the subsequent
2749 forward lookup of the name timed out, the host name was left in
2750 $sender_host_name, contrary to the specification.
2752 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2753 restricted to single-key lookups, Exim was not diagnosing an error if
2754 * or *@ was used with a query-style lookup.
2756 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2758 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2759 long ago, but noone interested enough thought of it. Let's just say that
2760 the "1.1" means that there are some new functions that weren't there at
2761 some point in the past.
2763 PH/51 Error processing for expansion failure of helo_data from an smtp
2764 transport during callout processing was broken.
2766 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2767 tested/used via the -bh/-bhc/-bs options.
2769 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2770 bug, fixed in subsequent PCRE releases).
2772 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2773 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2775 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2781 SC/01 Use a glob alias rather than an array ref in eximstats generated
2782 parser. This improves both readability and performance.
2784 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2785 Don't display local sender or destination tables in eximstats unless
2786 there is data to show.
2787 Added average volumes into the eximstats top table text output.
2789 SC/03 Collect data on the number of addresses (recipients) as well
2790 as the number of messages in eximstats.
2792 TF/01 Correct an error in the documentation for the redirect router. Exim
2793 does (usually) call initgroups() when daemonizing.
2795 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2796 with consistent privilege compared to when running as a daemon.
2798 TF/03 Note in the spec that $authenticated_id is not set for local
2799 submissions from trusted users.
2801 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2802 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2804 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2805 by adding some example configuration directives to the default
2806 configuration file. A little bit of work is required to uncomment the
2807 directives and define how usernames and passwords are checked, but
2808 there is now a framework to start from.
2810 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2811 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2812 without this. I don't know how relevant this is to other LDAP libraries.
2814 PH/02 Add the verb name to the "unknown ACL verb" error.
2816 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2818 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2820 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2821 directory not expanded when it should be if an expanded home directory
2822 was set for the address (which is overridden by the transport).
2824 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2827 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2828 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2829 because it is too late at that time, and has no effect.
2831 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2832 security issue with \' (bugzilla #107). I could not use the
2833 PQescapeStringConn() function, because it needs a PGconn value as one of
2836 PH/08 When testing addresses using -bt, indicate those final addresses that
2837 are duplicates that would not cause an additional delivery. At least one
2838 person was confused, thinking that -bt output corresponded to deliveries.
2839 (Suppressing duplicates isn't a good idea as you lose the information
2840 about possibly different redirections that led to the duplicates.)
2842 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2843 systems where poll() doesn't work, in particular OS X.
2845 PH/10 Added more information to debugging output for retry time not reached.
2847 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2848 operations in malware.c.
2850 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2853 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2854 syslog_duplication set false, log lines that would normally be written
2855 both the the main log and to the reject log were not written to syslog at
2858 PH/14 In the default configuration, change the use of "message" in ACL warn
2859 statements to "add_header".
2861 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2862 not followed by a command (e.g. "seen endif").
2864 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2865 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2868 PH/17 Added extra conditions to the default value of delay_warning_condition
2872 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2873 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2874 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2877 The Auto-Submitted: and various List- headers are standardised, whereas I
2878 don't think Precedence: ever was.
2880 PH/18 Refactored debugging code in route_finduser() to show more information,
2881 in particular, the error code if getpwnam() issues one.
2883 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2884 This is apparently needed in addition to the PH/07 change above to avoid
2885 any possible encoding problems.
2887 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2888 but not after initializing Perl.
2890 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2891 output them only if debugging. By default they are written stderr,
2892 apparently, which is not desirable.
2894 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2897 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2900 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2902 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2903 authenticated or an ident call has been made. Suppress the default
2904 values for $authenticated_id and $authenticated_sender (but permit -oMai
2905 and -oMas) when testing with -bh.
2907 PH/24 Re-jigged the order of the tests in the default configuration so that the
2908 tests for valid domains and recipients precede the DNS black list and CSA
2909 tests, on the grounds that those ones are more expensive.
2911 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2912 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2913 This bug exists in every version of Exim that I still have, right back to
2916 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2917 However, an attempt to turn on case-sensitivity in a regex key by
2918 including (?-i) didn't work because the subject string was already
2919 lowercased, and the effects were non-intuitive. It turns out that a
2920 one-line patch can be used to allow (?-i) to work as expected.
2926 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2927 other effects) broke the use of negated acl sub-conditions.
2929 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2932 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2933 "Deny" causes Exim to reject the incoming connection with a 554 error.
2934 Unfortunately, if there is a major crisis, such as a disk failure,
2935 tcp-wrappers gives "deny", whereas what one would like would be some
2936 kind of temporary error. A kludge has been added to help with this.
2937 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2938 554 error is used if errno is still zero or contains ENOENT (which occurs
2939 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2942 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2944 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2945 errors. Otherwise a message that provokes a temporary error (when other
2946 messages do not) can cause a whole host to time out.
2948 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2949 addresses were routed directly to files or pipes from a redirect router.
2950 File deliveries just didn't batch; pipe deliveries might have suffered
2953 PH/06 A failure to get a lock for a hints database would erroneously always say
2954 "Failed to get write lock", even when it was really a read lock.
2956 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2957 of 0600. This has been changed to use the value of the lockfile_mode
2958 option (which defaults to 0600).
2960 PH/08 Applied small patch from the Sieve maintainer.
2962 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2963 folder from quota calculations, a direct delivery into this folder messed
2964 up the contents of the maildirsize file. This was because the regex was
2965 used only to exclude .Trash (or whatever) when the size of the mailbox
2966 was calculated. There was no check that a delivery was happening into an
2967 excluded directory. This bug has been fixed by ignoring all quota
2968 processing for deliveries into excluded directories.
2970 PH/10 Added the maildirfolder_create_regex option to appendfile.
2976 PH/01 The code for finding all the local interface addresses on a FreeBSD
2977 system running IPv6 was broken. This may well have applied to all BSD
2978 systems, as well as to others that have similar system calls. The broken
2979 code found IPv4 interfaces correctly, but gave incorrect values for the
2980 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2981 that it would not match correctly against @[] and not recognize the IPv6
2984 PH/02 The ipliteral router was not recognizing addresses of the form user@
2985 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2987 PH/03 Added disable_ipv6.
2989 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2990 decoded form, because it is most often used to construct To: headers
2991 lines in autoreplies, and the decoded form may well be syntactically
2992 invalid. However, $reply_address has leading white space removed, and all
2993 newlines turned into spaces so that the autoreply transport does not
2996 PH/05 If group was specified without a user on a router, and no group or user
2997 was specified on a transport, the group from the router was ignored.
2999 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3000 for visible compile-time settings that can be used to change these
3001 numbers, for those that want even more. Backwards compatibility with old
3002 spool files has been maintained. However, going back to a previous Exim
3003 release will lost any variables that are in spool files.
3005 PH/07 Two small changes when running in the test harness: increase delay when
3006 passing a TCP/IP connection to a new process, in case the original
3007 process has to generate a bounce, and remove special handling of
3008 127.0.0.2 (sic), which is no longer necessary.
3010 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3011 be the same on different OS.
3013 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3016 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3017 whether --show-vars was specified or not
3019 JJ/02 exipick: Added support for new ACL variable spool format introduced
3022 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3023 syntactically invalid From: or Reply-to: line, and a filter used this to
3024 generate an autoreply, and therefore failed to obtain an address for the
3025 autoreply, Exim could try to deliver to a non-existent relative file
3026 name, causing unrelated and misleading errors. What now happens is that
3027 it logs this as a hard delivery error, but does not attempt to create a
3030 PH/11 The exinext utility has a -C option for testing purposes, but although
3031 the given file was scanned by exinext itself; it wasn't being passed on
3032 when Exim was called.
3034 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3035 an end-of-file indication when reading a command response.
3037 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3038 compiled. In many other places in Exim, IPv6 addresses are always
3039 recognized, so I have changed this. It also means that IPv4 domain
3040 literals of the form [IPV4:n.n.n.n] are now always recognized.
3042 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3043 used if the router is not running as root, for example, when verifying at
3044 ACL time, or when using -bh. The debugging output from this situation was
3045 non-existent - all you got was a failure to exec. I have made two
3048 (a) Failures to set uid/gid, the current directory, or a process leader
3049 in a subprocess such as that created by queryprogram now generate
3050 suitable debugging output when -d is set.
3052 (b) The queryprogram router detects when it is not running as root,
3053 outputs suitable debugging information if -d is set, and then runs
3054 the subprocess without attempting to change uid/gid.
3056 PH/15 Minor change to Makefile for building test_host (undocumented testing
3059 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3060 additional section of a DNS packet that returns MX or SRV records.
3061 Instead, it always explicitly searches for A/AAAA records. This avoids
3062 major problems that occur when a DNS server includes only records of one
3063 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3064 fixed another bug: if SRV records were looked up and the corresponding
3065 address records were *not* found in the additional section, the port
3066 values from the SRV records were lost.
3068 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3069 using the correct key (the original address) when searching the retry
3070 rules in order to find which one to use for generating the retry hint.
3072 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3073 adding the default one. Similarly, if it contains a Reply-To: header, the
3074 errors_reply_to option, if set, is not used.
3076 PH/19 When calculating a retry time, Exim used to measure the "time since
3077 failure" by looking at the "first failed" field in the retry record. Now
3078 it does not use this if it is later than than the arrival time of the
3079 message. Instead it uses the arrival time. This makes for better
3080 behaviour in cases where some deliveries succeed, thus re-setting the
3081 "first failed" field. An example is a quota failure for a huge message
3082 when small messages continue to be delivered. Without this change, the
3083 "time since failure" will always be short, possible causing more frequent
3084 delivery attempts for the huge message than are intended.
3085 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3087 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3088 $1, $2, $3) because the numerical variables can be reset during some
3089 expansion items (e.g. "match"), thereby losing the authentication data.
3091 PH/21 Make -bV show the size of off_t variables so that the test suite can
3092 decide whether to run tests for quotas > 2G.
3094 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3095 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3096 filecount value is greater than 2G or if a quota value is greater than 2G
3097 on a system where the size of off_t is not greater than 4, a panic error
3100 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3101 never match. The debug and -bh output now contains an explicit error
3102 message indicating a malformed IPv4 address or mask.
3104 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3105 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3106 PH/23 above applies.
3108 PH/25 Do not write to syslog when running in the test harness. The only
3109 occasion when this arises is a failure to open the main or panic logs
3110 (for which there is an explicit test).
3112 PH/26 Added the /no_tell option to "control=freeze".
3114 PH/27 If a host name lookup failed very early in a connection, for example, if
3115 the IP address matched host_lookup and the reverse lookup yielded a name
3116 that did not have a forward lookup, an error message of the form "no IP
3117 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3118 could be logged. Now it outputs the IP address instead of "NULL".
3120 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3121 allows the sender and the authenticated sender to be set when
3122 submitting a message from within Exim. Since child_open_exim() is
3123 documented for local_scan(), the new function should be too.
3125 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3126 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3127 results in an empty string is now treated as unset.
3129 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3131 PH/31 Added sender_verify_fail logging option.
3133 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3134 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3135 tidied the source and removed it altogether.
3137 PH/33 When a queue run was abandoned because the load average was too high, a
3138 log line was always written; now it is written only if the queue_run log
3139 selector is set. In addition, the log line for abandonment now contains
3140 information about the queue run such as the pid. This is always present
3141 in "start" and "stop" lines but was omitted from the "abandon" line.
3143 PH/34 Omit spaces between a header name and the colon in the error message that
3144 is given when verify = headers_syntax fails (if there are lots of them,
3145 the message gets confusing).
3147 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3148 names, as there are now some PTR records that contain slashes. This check
3149 is only to protect against broken name servers that fall over on strange
3150 characters, so the fact that it applies to all lookups doesn't matter.
3152 PH/36 Now that the new test suite is complete, we can remove some of the
3153 special code in Exim that was needed for the old test suite. For example,
3154 sorting DNS records because real resolvers return them in an arbitrary
3155 order. The new test suite's fake resolver always returns records in the
3158 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3159 bounces) except when queue_only is set, to avoid logging races between
3160 the different processes.
3162 PH/38 Panic-die if .include specifies a non-absolute path.
3164 PH/39 A tweak to the "H" retry rule from its user.
3166 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3167 a label. They prevented compilation on older perls.
3169 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3170 a warning to be raised on newish perls.
3172 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3173 on queue. Changes to match documented behaviour of showing count of
3174 messages matching specified criteria.
3176 PH/40 Changed the default ident timeout from 30s to 5s.
3178 PH/41 Added support for the use of login_cap features, on those BSD systems
3179 that have them, for controlling the resources used by pipe deliveries.
3181 PH/42 The content-scanning code uses fopen() to create files in which to put
3182 message data. Previously it was not paying any attention to the mode of
3183 the files. Exim runs with umask(0) because the rest of the code creates
3184 files with open(), and sets the required mode explicitly. Thus, these
3185 files were ending up world-writeable. This was not a big issue, because,
3186 being within the spool directory, they were not world-accessible. I have
3187 created a function called modefopen, which takes an additional mode
3188 argument. It sets umask(777), creates the file, chmods it to the required
3189 mode, then resets the umask. All the relevant calls to fopen() in the
3190 content scanning code have been changed to use this function.
3192 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3193 to 24 hours. This avoids potential overflow problems when processing G
3194 and H retry rules. I suspect nobody ever tinkers with this value.
3196 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3198 PH/45 When the plaintext authenticator is running as a client, the server's
3199 challenges are checked to ensure they are valid base64 strings. By
3200 default, the authentication attempt is cancelled if an invalid string is
3201 received. Setting client_ignore_invalid_base64 true ignores these errors.
3202 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3203 they are received. Thus, the responses can be made to depend on the
3204 challenges. If an invalid string is ignored, an empty string is placed in
3207 PH/46 Messages that are created by the autoreply transport now contains a
3208 References: header, in accordance with RFCs 2822 and 3834.
3210 PH/47 Added authenticated_sender_force to the smtp transport.
3212 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3214 PH/49 Installed latest patch from the Sieve maintainer.
3216 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3217 was also set, the appendfile transport was unnecessarily scanning a
3218 directory of message files (e.g. for maildir delivery) to find the count
3219 of files (along with the size), even though it did not need this
3220 information. It now does the scan only if it needs to find either the
3221 size of the count of files.
3223 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3225 PH/52 Two bugs concerned with error handling when the smtp transport is
3228 (i) Exim was not creating retry information for temporary errors given
3229 for individual recipients after the DATA command when the smtp transport
3230 was used in LMTP mode. This meant that they could be retried too
3231 frequently, and not timed out correctly.
3233 (ii) Exim was setting the flag that allows error details to be returned
3234 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3235 recipients that were returned after the DATA command.
3237 PH/53 This is related to PH/52, but is more general: for any failing address,
3238 when detailed error information was permitted to be returned to the
3239 sender, but the error was temporary, then after the final timeout, only
3240 "retry timeout exceeded" was returned. Now it returns the full error as
3241 well as "retry timeout exceeded".
3243 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3244 do this, and (what is worse) MTAs that accept it.
3246 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3247 will now be deprecated.
3249 PH/56 New os.c-cygwin from the Cygwin maintainer.
3251 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3252 formats (previously only available in exim formats via -bpr, -bpru,
3253 and -bpra. Now also available in native and exiqgrep formats)
3255 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3256 with very large, slow to parse queues
3258 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3260 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3262 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3263 responses to authentication challenges, though it was showing the
3264 challenges; (ii) I've removed the CR characters from the debug output for
3267 PH/58 Allow for the insertion of a newline as well as a space when a string
3268 is turned into more than one encoded-word during RFC 2047 encoding. The
3269 Sieve code now uses this.
3271 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3272 data_4xx, lost_connection, tls_required.
3274 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3275 message was being sent as an SMTP response.
3277 PH/61 Add -l and -k options to exicyclog.
3279 PH/62 When verifying, if an address was redirected to one new address, so that
3280 verification continued, and the new address failed or deferred after
3281 having set something in $address_data, the value of $address_data was not
3282 passed back to the ACL. This was different to the case when no
3283 redirection occurred. The value is now passed back in both cases.
3285 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3286 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3287 use login_cap.h, so on its own it isn't the distinguishing feature. The
3288 new name refers directly to the setclassresources() function.
3290 PH/65 Added configuration files for NetBSD3.
3292 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3294 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3295 is preferred over IPv4.
3297 PH/68 The bounce_return_message and bounce_return_body options were not being
3298 honoured for bounces generated during the reception of non-SMTP messages.
3299 In particular, this applied to messages rejected by the ACL. This bug has
3300 been fixed. However, if bounce_return_message is true and bounce_return_
3301 body is false, the headers that are returned for a non-SMTP message
3302 include only those that have been read before the error was detected.
3303 (In the case of an ACL rejection, they have all been read.)
3305 PH/69 The HTML version of the specification is now built in a directory called
3306 spec_html instead of spec.html, because the latter looks like a path with
3307 a MIME-type, and this confuses some software.
3309 PH/70 Catch two compiler warnings in sieve.c.
3311 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3312 function verify_get_ident() calls ip_connect() to connect a socket, but
3313 if the "connect()" function timed out, ip_connect() used to close the
3314 socket. However, verify_get_ident() also closes the socket later, and in
3315 between Exim writes to the log, which may get opened at this point. When
3316 the socket was closed in ip_connect(), the log could get the same file
3317 descriptor number as the socket. This naturally causes chaos. The fix is
3318 not to close the socket in ip_connect(); the socket should be closed by
3319 the function that creates it. There was only one place in the code where
3320 this was missing, in the iplookup router, which I don't think anybody now
3321 uses, but I've fixed it anyway.
3323 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3324 well as to direct DNS lookups. Otherwise the handling of names in host
3325 lists is inconsistent and therefore confusing.
3331 PH/01 Two changes to the default runtime configuration:
3333 (1) Move the checks for relay_from_hosts and authenticated clients from
3334 after to before the (commented out) DNS black list checks.
3336 (2) Add control=submission to the relay_from_hosts and authenticated
3337 clients checks, on the grounds that messages accepted by these
3338 statements are most likely to be submissions.
3340 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3342 (1) Generate an error if the third argument for the ${prvs expansion is
3345 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3348 (3) Reset the variables that are obtained from the first argument of
3349 ${prvscheck and used in the second argument before leaving the code,
3350 because their memory is reclaimed, so using them afterwards may do
3353 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3354 one (it's much easier than Tom thought :-).
3356 (5) Because of (4), we can now allow for the use of $prvscheck_result
3357 inside the third argument.
3359 PH/03 For some reason, the default setting of PATH when running a command from
3360 a pipe transport was just "/usr/bin". I have changed it to
3363 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3364 anything to be listed in the output from -bV.
3366 PH/05 When a filter generated an autoreply, the entire To: header line was
3367 quoted in the delivery log line, like this:
3369 => >A.N.Other <ano@some.domain> <original@ddress> ...
3371 This has been changed so that it extracts the operative address. There
3372 may be more than one such address. If so, they are comma-separated, like
3375 => >ano@some.domain,ona@other.domain <original@ddress> ...
3377 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3378 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3379 not being looked up in the rDNS to get a host name, Exim was showing the
3380 IP address twice in Received: lines, even though the IP addresses were
3381 identical. For example:
3383 Received: from [1.2.3.4] (helo=[1.2.3.4])
3385 However, if the real host name was known, it was omitting the HELO data
3386 if it matched the actual IP address. This has been tidied up so that it
3387 doesn't show the same IP address twice.
3389 PH/07 When both +timestamp and +memory debugging was on, the value given by
3390 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3391 called by the debug printing, thereby overwriting the timestamp buffer.
3392 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3394 PH/08 When the original message was included in an autoreply transport, it
3395 always said "this is a copy of the message, including all the headers",
3396 even if body_only or headers_only was set. It now gives an appropriate
3399 PH/09 Applied a patch from the Sieve maintainer which:
3401 o fixes some comments
3402 o adds the (disabled) notify extension core
3403 o adds some debug output for the result of if/elsif tests
3404 o points to the current vacation draft in the documentation
3405 and documents the missing references header update
3409 o fixes a bug in processing the envelope test (when testing
3410 multiple envelope elements, the last element determined the
3413 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3414 Electronic Mail") by including:
3416 Auto-submitted: auto-generated
3418 in the messages that it generates (bounce messages and others, such as
3419 warnings). In the case of bounce messages for non-SMTP messages, there was
3420 also a typo: it was using "Auto_submitted" (underscore instead of
3421 hyphen). Since every message generated by Exim is necessarily in response
3422 to another message, thes have all been changed to:
3424 Auto-Submitted: auto-replied
3426 in accordance with these statements in the RFC:
3428 The auto-replied keyword:
3430 - SHOULD be used on messages sent in direct response to another
3431 message by an automatic process,
3433 - MUST NOT be used on manually-generated messages,
3435 - MAY be used on Delivery Status Notifications (DSNs) and Message
3436 Disposition Notifications (MDNs),
3438 - MUST NOT be used on messages generated by automatic or periodic
3439 processes, except for messages which are automatic responses to
3442 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3443 to the default Received: header definition.
3445 PH/12 Added log selector acl_warn_skipped (default on).
3447 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3448 variables because (a) they are in the wrong storage pool and (b) even if
3449 they were copied, it wouldn't work properly because of the caching.
3451 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3452 checking when decoding. Apparently there are clients that generate
3453 overlong encoded strings. Why am I not surprised?
3455 PH/15 If the first argument of "${if match_address" was not empty, but did not
3456 contain an "@" character, Exim crashed. Now it writes a panic log message
3457 and treats the condition as false.
3459 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3461 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3462 extension "envelope-auth". The code is finished and in agreement with
3463 other implementations, but there is no documentation so far and in fact,
3464 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3465 not changing the active code.
3467 Print executed "if" and "elsif" statements when debugging is used. This
3468 helps a great deal to understand what a filter does.
3470 Document more things not specified clearly in RFC3028. I had all this
3471 sorted out, when out of a sudden new issues came to my mind. Oops."
3473 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3476 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3477 canonical form (as documented). However, after a host name lookup from
3478 the IP address, check_host() was doing a simple string comparison with
3479 addresses acquired from the DNS when checking that the found name did
3480 have the original IP as one of its addresses. Since any found IPv6
3481 addresses are likely to be in abbreviated form, the comparison could
3482 fail. Luckily, there already exists a function for doing the comparison
3483 by converting both addresses to binary, so now that is used instead of
3484 the text comparison.
3486 PH/20 There was another similar case to PH/19, when a complete host name was
3487 given in a host list; looking up its IP address could give an abbreviated
3488 form, whereas the current host's name might or might not be abbreviated.
3489 The same fix has been applied.
3495 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3496 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3499 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3500 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3502 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3504 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3505 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3506 Also, for "Auto-Submitted", treat anything other than "no" as
3507 non-personal, in accordance with RFC 3834. (Previously it treated
3508 anything starting "auto-" as non-personal.)
3510 TF/01 The control=submission/name=... option had a problem with syntax
3511 errors if the name included a slash character. The /name= option
3512 now slurps the rest of the string, so it can include any characters
3513 but it must come last in the list of options (after /sender_retain
3516 PH/05 Some modifications to the interface to the fake nameserver for the new
3524 TK/01 Added the "success_on_redirect" address verification option. See
3525 NewStuff for rationale and an example.
3527 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3529 PH/02 Patch to exigrep to allow it to work on syslog lines.
3531 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3532 fread() to skip over the body file's header line, because in Cygwin the
3533 header line is locked and is inaccessible.
3535 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3536 co-exist for some time) to make it clear that it is the Exim ID that is
3537 referenced, not the Message-ID: header line.
3539 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3540 string_format() function, because snprintf() does not exist on all
3543 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3544 ${stat: expansion item.
3546 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3547 protocol synchronization error", to keep the pedants happy.
3549 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3550 well as for IRIX systems, when gcc is being used. See the host.c source
3553 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3555 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3558 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3559 transport and to the smtp transport in LMTP mode.
3561 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3563 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3564 run a filter in a subprocess. This could lead to confusion in subsequent
3565 lookups in the parent process. There should also be a search_tidyup() at
3566 the end of the subprocess.
3568 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3569 only if the host matched helo_try_verify_hosts, which caused the
3570 verification to occur when the EHLO/HELO command was issued. The ACL just
3571 tested the remembered result. Now, if a previous verification attempt has
3572 not happened, "verify = helo" does it there and then.
3574 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3576 TK/03 Fix log output including CR from clamd.
3578 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3579 exist provoked a memory error which could cause a segfault.
3581 PH/15 Installed PCRE 6.2
3583 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3585 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3586 of the problem. Specifically, suggested +O2 rather than +O1 for the
3589 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3591 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3592 record, the port was not correctly logged when the outgoing_port log
3593 selector was set (it logged the transort's default port).
3595 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3596 fallback_hosts, and "hosts" in the smtp transport.
3598 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3599 host errors such as "Connection refused".
3601 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3602 authentication with radiusclient 0.4.9:
3604 - Error returned from rc_read_config was caught wrongly
3605 - Username/password not passed on to radius server due to wrong length.
3607 The presumption is that some radiusclient API changes for 4.51/PH/17
3608 were not taken care of correctly. The code is still untested by me (my
3609 Linux distribution still has 0.3.2 of radiusclient), but it was
3610 contributed by a Radius user.
3612 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3613 expanding the "port" option of the smtp transport.
3615 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3616 while reading a MIME header. Thanks to Tom Hughes for a patch.
3618 PH/24 Include config.h inside local_scan.h so that configuration settings are
3621 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3622 that in an ACL for RCPT (for example), you can examine exactly what was
3625 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3626 commands, but it was not correctly comparing the address with the actual
3627 client host address. Thus, it would show the EHLO address in Received:
3628 header lines when this was not necessary.
3630 PH/27 Added the % operator to ${eval:}.
3632 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3633 it should be ignoring failures (because with -C, for example, it has lost
3634 privilege). It wasn't ignoring creation failures other than "already
3637 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3640 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3641 because an input line was too long, either on its own, or by virtue of
3642 too many continuations, the temporary file was not being removed, and the
3643 return code was incorrect.
3645 PH/31 Missing "BOOL" in function definition in filtertest.c.
3647 PH/32 Applied Sieve patches from the maintainer.
3649 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3651 PH/33 Added "verify = not_blind".
3653 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3654 Local/Makefile (with some defaults set). These are used in built scripts
3655 such as exicyclog, but they have never been used in the exim_install
3656 script (though there are many overriding facilities there). I have
3657 arranged that the exim_install script now takes note of these two
3660 PH/35 Installed configuration files for Dragonfly.
3662 PH/36 When a locally submitted message by a trusted user did not contain a
3663 From: header, and the sender address was obtained from -f or from an SMTP
3664 MAIL command, and the trusted user did not use -F to supply a sender
3665 name, $originator_name was incorrectly used when constructing a From:
3666 header. Furthermore, $originator_name was used for submission mode
3667 messages from external hosts without From: headers in a similar way,
3668 which is clearly wrong.
3670 PH/37 Added control=suppress_local_fixups.
3672 PH/38 When log_selector = +received_sender was set, and the addition of the
3673 sender made the log line's construction buffer exactly full, or one byte
3674 less than full, an overflow happened when the terminating "\n" was
3677 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3678 when the result of a list match is failure because a DNS lookup failed.
3680 PH/40 RM_COMMAND is now used in the building process.
3682 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3683 the "build-* directories that it finds.
3685 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3686 address in a domain literal was a prefix of an interface address.
3688 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3689 when verifying a sender address, unless rewrite_headers is false.
3691 PH/44 Wrote a long comment about why errors_to addresses are verified as
3692 recipients, not senders.
3694 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3695 the ratelimit ACL was added.
3697 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3699 PH/46 Added extra information about PostgreSQL errors to the error string.
3701 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3702 avoiding the need to install special zones in a real server. This is
3703 backwards compatible; if it can't find the fake resolver, it drops back.
3704 Thus, both old and new test suites can be run.
3706 TF/02 Added util/ratelimit.pl
3708 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3709 clock is set back in time.
3711 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3712 Candler <B.Candler@pobox.com>.
3714 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3715 OK for addresses that are the result of redirections.
3717 PH/48 A number of further additions for the benefit of the new test suite,
3718 including a fake gethostbyname() that interfaces to the fake DNS resolver
3721 TF/06 The fix for widen_domains has also been applied to qualify_single and
3722 search_parents which are the other dnslookup options that can cause
3725 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3728 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3730 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3731 service" if it encountered an MX record with an empty target hostname.
3732 The message is now "an MX or SRV record indicated no SMTP service".
3734 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3735 if the DNS of the sending site is misconfigured. This is quite a
3736 common situation. This change restores the behaviour of treating a
3737 helo verification defer as a failure.
3739 PH/51 If self=fail was set on a router, the bounce message did not include the
3740 actual error message.
3746 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3748 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3749 command itself ended in error, the underlying message about the transport
3750 filter timeout was being overwritten with the pipe command error. Now the
3751 underlying error message should be appended to the second error message.
3753 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3755 PH/02 Reduce the amount of output that "make" produces by default. Full output
3756 can still be requested.
3758 PH/03 The warning log line about a condition test deferring for a "warn" verb
3759 was being output only once per connection, rather than after each
3760 occurrence (because it was using the same function as for successful
3761 "warn" verbs). This seems wrong, so I have changed it.
3763 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3764 it should not have, which might have caused a crash in the right
3765 circumstances, but probably never did.
3767 PH/04 Installed a modified version of Tony Finch's patch to make submission
3768 mode fix the return path as well as the Sender: header line, and to
3769 add a /name= option so that you can make the user's friendly name appear
3772 TF/03 Added the control = fakedefer ACL modifier.
3774 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3775 Mark Lowes for thorough testing.
3777 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3779 TK/03 Merged latest SRS patch from Miles Wilton.
3781 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3782 with the definition in sysexits.h (which is #included earlier).
3783 Fortunately, Exim does not actually use EX_OK. The code used to try to
3784 preserve the sysexits.h value, by assuming that macro definitions were
3785 scanned for macro replacements. I have been disabused of this notion,
3786 so now the code just undefines EX_OK before #including unistd.h.
3788 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3789 in the smtp transport. When a block could not be written in a single
3790 write() function, the timeout was being re-applied to each part-write.
3791 This seems wrong - if the receiver was accepting one byte at a time it
3792 would take for ever. The timeout is now adjusted when this happens. It
3793 doesn't have to be particularly precise.
3795 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3796 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3798 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3799 without a domain if the check to <postmaster@domain> fails.
3801 SC/01 Eximstats: added -xls and the ability to specify output files
3802 (patch written by Frank Heydlauf).
3804 SC/02 Eximstats: use FileHandles for outputting results.
3806 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3808 SC/04 Eximstats: fixed display of large numbers with -nvr option
3810 SC/05 Eximstats: fixed merging of reports with empty tables.
3812 SC/06 Eximstats: added the -include_original_destination flag
3814 SC/07 Eximstats: removed tabs and trailing whitespace.
3816 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3818 TK/06 MBOX spool code: Add real "From " MBOX separator line
3819 so the .eml file is really in mbox format (even though
3820 most programs do not really care). Patch from Alex Miller.
3822 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3823 The latter is generated from $received_to and is only set if the
3824 message has one envelope recipient. SA can use these headers,
3825 obviously out-of-the-box. Patch from Alex Miller.
3827 PH/08 The ${def test on a variable was returning false if the variable's
3828 value was "0", contrary to what the specification has always said!
3829 The result should be true unless the variable is empty.
3831 PH/09 The syntax error of a character other than { following "${if
3832 def:variable_name" (after optional whitespace) was not being diagnosed.
3833 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3834 accidental colon was present, for example, could give incorrect results.
3836 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3837 result is used (not including appendfile, where other changes are about
3840 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3841 This involved changing a lot of size variables from int to off_t. It
3842 should work with maildirs and everything.
3844 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3845 spamd dying while we are connected to it.
3847 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3850 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3851 function for BDB 4.3.
3853 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3855 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3856 the sources for PCRE have been re-arranged and more files are now
3859 PH/15 The code I had for printing potentially long long variables in PH/11
3860 above was not the best (it lost precision). The length of off_t variables
3861 is now inspected at build time, and an appropriate printing format (%ld
3862 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3863 to be "long long int" or "long int". This is needed for the internal
3864 formatting function string_vformat().
3866 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3867 the configuration file to be ":syslog", then the script "guesses" where
3868 the logs files are, rather than using the compiled in default. In our
3869 case the guess is not the same as the compiled default, so the script
3870 suddenly stopped working when I started to use syslog. The patch checks
3871 to see if log_file_path is "". If so, it attempts to read it from exim
3872 with no configuration file to get the compiled in version, before it
3873 falls back to the previous guessing code."
3875 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3876 implementing BATV in an Exim configuration. See NewStuff for the gory
3879 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3880 Makefile that are specific to HP-UX.
3882 PH/18 If the "use_postmaster" option was set for a recipient callout together
3883 with the "random" option, the postmaster address was used as the MAIL
3884 FROM address for the random test, but not for the subsequent recipient
3885 test. It is now used for both.
3887 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3888 patch removes a few documentation additions to RFC 3028, because the
3889 latest draft now contains them. It adds the new en;ascii-case comparator
3890 and a new error check for 8bit text in MIME parts. Comparator and
3891 require names are now matched exactly. I enabled the subaddress
3892 extension, but it is not well tested yet (read: it works for me)."
3894 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3895 rework some of the code of TK/09 above to avoid the hardwired use of
3896 "%lld" and "long long". Replaced the call to snprintf() with a call to
3899 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3900 records point to non-existent hosts", "retry timeout exceeded", and
3901 "retry time not reached for any host after a long failure period".
3903 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3904 experimental DomainKeys support:
3906 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3907 (2) On an error such as an illegally used "control", the wrong name for
3908 the control was given.
3910 These problems did NOT occur unless DomainKeys support was compiled.
3912 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3914 PH/24 Added ${if match_ip condition.
3916 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3917 databases so that it will be absolutely obvious if a crash occurs in the
3918 DB library. This is a regular occurrence (often caused by mis-matched
3921 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3922 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3923 picked up on a user's system that detects such things. There doesn't seem
3924 to be a gcc warning option for this - only an attribute that has to be
3925 put on the function's prototype. It seems that in Fedora Core 4 they have
3926 set this on a number of new functions. No doubt there will be more in due
3929 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3930 specify a transport. However, if an address that was verified by such a
3931 router was the subject of a callout, Exim crashed because it tried to
3932 read the rcpt_include_affixes from the non-existent transport. Now it
3933 just assumes that the setting of that option is false. This bug was
3934 introduced by 4.51/PH/31.
3936 PH/28 Changed -d+all to exclude +memory, because that information is very
3937 rarely of interest, but it makes the output a lot bigger. People tend to
3938 do -d+all out of habit.
3940 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3941 code in os-type was giving problems when libc.so lives in lib64, like on
3944 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3945 aren't the modern standard, and it seems that some systems' include files
3946 don't always have them. Exim was already checking for some of the newer
3947 ones like T_AAAA, and defining it itself. I've added checks for all the
3948 record types that Exim uses.
3950 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3951 not automatically generating a new one, as it is supposed to. This
3952 prevented TLS from working. If the file did exist, but contained invalid
3953 data, a new version was generated, as expected. It was only the case of a
3954 non-existent file that was broken.
3956 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3957 with a change in libdomainkeys > 0.64.
3959 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3960 from DNS. If the selector record carries the flag, it now has
3961 precedence over the domain-wide flag.
3963 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3965 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3966 the use of a port name that isn't defined in /etc/services) occurred, the
3967 message was deferred as in a normal delivery, and thus remained on the
3968 spool, instead of being failed because of the mua_wrapper setting. This
3969 is now fixed, and I tidied up some of the mua_wrapper messages at the
3972 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3973 the messages in a hash of arrays rather than using individual hashes.
3974 This is a bit cleaner and results in dramatic memory savings, albeit
3975 at a slight CPU cost.
3977 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3978 as requested by Marc Sherman.
3980 SC/10 Eximstats: added histograms for user specified patterns as requested
3983 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3985 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3986 fopen() in the content-scanning modules that did not already have it.
3992 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3993 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3995 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3997 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3999 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4000 to test Sieve filters that use "vacation".
4002 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4003 that changes the way the GnuTLS parameters are stored in the cache file.
4004 The new format can be generated externally. For backward compatibility,
4005 if the data in the cache doesn't make sense, Exim assumes it has read an
4006 old-format file, and it generates new data and writes a new file. This
4007 means that you can't go back to an older release without removing the
4010 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4011 work if there are any delivery delays because "one_time" forces the
4012 parent to be marked "delivered", so its unseen clone is never tried
4013 again. For this reason, Exim now forbids the simultaneous setting of
4016 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4017 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4018 entry if you want to know the details. The fix, however, neglected to
4019 consider the case when local delivery batching is involved. The test for
4020 "previously delivered" was not happening when checking to see if an
4021 address could be batched with a previous (undelivered) one; under
4022 certain circumstances this could lead to multiple deliveries to the same
4025 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4026 in its include files, and this causes problems building Exim.
4028 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4029 header_syntax) but Exim was just ignoring anything given after a slash.
4030 In particular, this caused confusion with an attempt to use "verify =
4031 reverse_host_lookup/defer_ok". An error is now given when options are
4032 supplied for verify items that do not have them. (Maybe reverse_host_
4033 lookup should have a defer_ok option, but that's a different point.)
4035 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4036 defined by RFC 821) to 2048, because there were problems with some AUTH
4037 commands, and RFC 1869 says the size should be increased for extended
4038 SMTP commands that take arguments.
4040 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4043 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4044 "unknown" error; now it says that the functionality isn't in the binary.
4046 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4047 an address' error message when a string expansion fails (syntax or
4048 whatever). Otherwise the password may appear in the log. Following change
4049 PH/42 below, there is no longer a chance of it appearing in a bounce
4052 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4054 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4055 of its addresses was ever tried. (Bugzilla bug #2.)
4057 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4058 the result incorrectly in the debug output. (It correctly added a newline
4059 to what was transported.)
4061 TF/01 Added $received_time.
4063 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4064 commented out examples of how to interface to a virus scanner and to
4065 SpamAssassin. Also added commented examples of av_scanner and
4066 spamd_address settings.
4068 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4069 and controls are allowed in which ACLs. There were a couple of minor
4070 errors. Some of the entries in the conditions table (which is a table of
4071 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4072 negation of where the condition IS allowed.
4074 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4076 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4077 header file does not have a version number, so I've had to invent a new
4078 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4079 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4080 radiusclient), but it was contributed by a Radius user.
4082 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4083 files or directories, for OpenSSL.
4085 PH/19 When an Exim process that is running as root has to create an Exim log
4086 file, it does so in a subprocess that runs as exim:exim so as to get the
4087 ownership right at creation (otherwise, other Exim processes might see
4088 the file with the wrong ownership). There was no test for failure of this
4089 fork() call, which would lead to the process getting stuck as it waited
4090 for a non-existent subprocess. Forks do occasionally fail when resources
4091 run out. I reviewed all the other calls to fork(); they all seem to check
4094 PH/20 When checking for unexpected SMTP input at connect time (before writing
4095 the banner), Exim was not dealing correctly with a non-positive return
4096 from the read() function. If the client had disconnected by this time,
4097 the result was a log entry for a synchronization error with an empty
4098 string after "input=" when read() returned zero. If read() returned -1
4099 (an event I could not check), uninitialized data bytes were printed.
4100 There were reports of junk text (parts of files, etc) appearing after
4103 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4105 PH/22 Added support for macro redefinition, and (re)definition in between
4106 driver and ACL definitions.
4108 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4109 forgetting to use the resulting value; it was using the unexpanded value.
4111 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4112 hadn't been configured. The fix is from Juergen Kreileder, who
4113 understands it better than I do:
4115 "Here's what I see happening with three configured cyrus_sasl
4116 authenticators configured (plain, login, cram-md5):
4118 On startup auth_cyrus_sasl_init() gets called for each of these.
4119 This means three calls to sasl_listmech() without a specified mech_list.
4120 => SASL tests which mechs of all available mechs actually work
4121 => three warnings about OTP not working
4122 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4124 With the patch, sasl_listmech() also gets called three times. But now
4125 SASL's mech_list option is set to the server_mech specified in the the
4126 authenticator. Or in other words, the answer from sasl_listmech()
4127 gets limited to just the mech you're testing for (which is different
4129 => the return list contains just 'plain' or 'login', 'cram-md5' or
4130 nothing depending on the value of ob->server_mech.
4132 I've just tested the patch: Authentication still works fine,
4133 unavailable mechs specified in the exim configuration are still
4134 caught, and the auth.log warnings about OTP are gone."
4136 PH/25 When debugging is enabled, the contents of the command line are added
4137 to the debugging output, even when log_selector=+arguments is not
4140 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4141 answer is "GNU", and only if the return is "GNU/something" is the answer
4144 PH/27 $acl_verify_message is now set immediately after the failure of a
4145 verification in an ACL, and so is available in subsequent modifiers. In
4146 particular, the message can be preserved by coding like this:
4148 warn !verify = sender
4149 set acl_m0 = $acl_verify_message
4151 Previously, $acl_verify_message was set only while expanding "message"
4152 and "log_message" when a very denied access.
4154 PH/28 Modified OS/os.c-Linux with
4156 -#ifndef OS_LOAD_AVERAGE
4157 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4159 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4160 nomenclature these days.)
4162 PH/29 Installed patch from the Sieve maintainer that adds the options
4163 sieve_useraddress and sieve_subaddress to the redirect router.
4165 PH/30 In these circumstances:
4166 . Two addresses routed to the same list of hosts;
4167 . First host does not offer TLS;
4168 . First host accepts first address;
4169 . First host gives temporary error to second address;
4170 . Second host offers TLS and a TLS session is established;
4171 . Second host accepts second address.
4172 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4173 and peerdn, if requested) that were in fact used only for the second
4176 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4177 attention to any local part prefix or suffix that was matched by the
4178 router that accepted the address. It now behaves in the same way as it
4179 does for delivery: the affixes are removed from the local part unless
4180 rcpt_include_affixes is set on the transport.
4182 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4183 timeout during the DATA phase of an incoming message.
4185 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4186 applied a patch sanctioned by the Sieve maintainer.
4188 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4189 the uid or gid is negative. A case of a negative gid caused this to be
4190 noticed. The fix allows for either to be negative.
4192 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4193 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4196 PH/36 The change PH/12 above was broken. Fixed it.
4198 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4199 the grounds that routing the same address twice would always produce the
4200 same answer. This might have been true once, but it is certainly no
4201 longer true now. Routing a child address may depend on the previous
4202 routing that produced that child. Some complicated redirection strategies
4203 went wrong when messages had multiple recipients, and made Exim's
4204 behaviour dependent on the order in which the addresses were given.
4206 I have moved the duplicate checking until after the routing is complete.
4207 Exim scans the addresses that are assigned to local and remote
4208 transports, and removes any duplicates. This means that more work will be
4209 done, as duplicates will always all be routed, but duplicates are
4210 presumably rare, so I don't expect this is of any significance.
4212 For deliveries to pipes, files, and autoreplies, the duplicate checking
4213 still happens during the routing process, since they are not going to be
4216 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4217 It corrects a timeout issue with spamd. This is Ian's comment: "The
4218 background is that sometimes spamd either never reads data from a
4219 connection it has accepted, or it never writes response data. The exiscan
4220 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4221 blindly assumes that writes won't block so it may never time out."
4223 PH/39 Allow G after quota size as well as K and M.
4225 PH/40 The value set for $authenticated_id in an authenticator may not contain
4226 binary zeroes or newlines because the value is written to log lines and
4227 to spool files. There was no check on this. Now the value is run through
4228 the string_printing() function so that such characters are converted to
4229 printable escape sequences.
4231 PH/41 $message_linecount is a new variable that contains the total number of
4232 lines in the message. Compare $body_linecount, which is the count for the
4235 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4236 bounce and delay warning messages, except in certain special cases, which
4239 (a) An SMTP error message from a remote host;
4240 (b) A message specified in a :fail: redirection;
4241 (c) A message specified in a "fail" command in a system filter;
4242 (d) A message specified in a FAIL return from the queryprogram router;
4243 (e) A message specified by the cannot_route_message router option.
4245 In these cases only, Exim does include the error details in bounce and
4246 warning messages. There are also a few cases where bland messages such
4247 as "unrouteable address" or "local delivery error" are given.
4249 PH/43 $value is now also set for the "else" part of a ${run expansion.
4251 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4252 being worked on, but at least Exim now implements the latest version to
4255 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4256 process to complete was treated as a delivery failure, a timeout while
4257 writing the message to the pipe was logged, but erroneously treated as a
4258 successful delivery. Such timeouts include transport filter timeouts. For
4259 consistency with the overall process timeout, these timeouts are now
4260 treated as errors, giving rise to delivery failures by default. However,
4261 there is now a new Boolean option for the pipe transport called
4262 timeout_defer, which, if set TRUE, converts the failures into defers for
4263 both kinds of timeout. A transport filter timeout is now identified in
4266 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4267 systems where "make" and "gmake" are different, calling "gmake" at top
4268 level broke things. I've arranged for the value of $(MAKE) to be passed
4269 from the Makefile to this script so that it can call the same version of
4273 A note about Exim versions 4.44 and 4.50
4274 ----------------------------------------
4276 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4277 changes of various kinds. As a consequence, a big documentation update was
4278 needed. This delayed the release for rather longer than seemed good, especially
4279 in the light of a couple of (minor) security issues. Therefore, the changes
4280 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4281 release. So 4.44 and 4.50 are in effect two different branches that both start
4284 I have left the 4.50 change log unchanged; it contains all the changes since
4285 4.43. The change log for 4.44 is below; many of its items are identical to
4286 those for 4.50. This seems to be the most sensible way to preserve the
4287 historical information.
4293 1. Minor wording change to the doc/README.SIEVE file.
4295 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4296 computation of the current number of files was incorrect.
4298 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4299 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4302 4. Give more explanation in the error message when the command for a transport
4303 filter fails to execute.
4305 5. There are several places where Exim runs a non-Exim command in a
4306 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4307 was being done only for the command run by the queryprogram router. It is
4308 now done for all such subprocesses. The other cases are: ${run, transport
4309 filters, and the commands run by the lmtp and pipe transports.
4311 6. Added CONFIGURE_GROUP build-time option.
4313 7. Some older OS have a limit of 256 on the maximum number of file
4314 descriptors. Exim was using setrlimit() to set 1000 as a large value
4315 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4316 systems. I've change it so that if it can't get 1000, it tries for 256.
4318 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4319 was an oversight, and furthermore, ever since the addition of extra
4320 controls (e.g. 4.43/32), the checks on when to allow different forms of
4321 "control" were broken. There should now be diagnostics for all cases when a
4322 control that does not make sense is encountered.
4324 9. Added the /retain_sender option to "control=submission".
4326 10. $recipients is now available in the predata ACL (oversight).
4328 11. Tidy the search cache before the fork to do a delivery from a message
4329 received from the command line. Otherwise the child will trigger a lookup
4330 failure and thereby defer the delivery if it tries to use (for example) a
4331 cached ldap connection that the parent has called unbind on.
4333 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4334 of $address_data from the recipient verification was clobbered by the
4335 sender verification.
4337 13. The value of address_data from a sender verification is now available in
4338 $sender_address_data in subsequent conditions in the ACL statement.
4340 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4342 15. Added a new option "connect=<time>" to callout options, to set a different
4345 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4346 was its contents. (It was OK if the option was not defined at all.)
4348 17. A "Completed" log line is now written for messages that are removed from
4349 the spool by the -Mrm option.
4351 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4352 information about exactly what failed.
4354 19. Added -dd to debug only the daemon process.
4356 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4357 handles timeouts, both on the server side and network timeouts. Renamed the
4358 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4360 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4361 It is now set to "smtps".
4363 22. $host_address is now set to the target address during the checking of
4364 ignore_target_hosts.
4366 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4367 being passed; this would have caused $sender_host_name to have been used if
4368 matching the list had actually called for a host name (not very likely,
4369 since this list is usually IP addresses). A host name is now passed as
4372 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4373 code that specifies a non-restarting handler (typically sigaction() in
4374 modern systems) in an attempt to fix a rare and obscure crash bug.
4376 25. Narrowed the window for a race in the daemon that could cause it to ignore
4377 SIGCHLD signals. This is not a major problem, because they are used only to
4378 wake it up if nothing else does.
4380 26. A malformed maildirsize file could cause Exim to calculate negative values
4381 for the mailbox size or file count. Odd effects could occur as a result.
4382 The maildirsize information is now recalculated if the size or filecount
4385 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4386 support for a long time. Removed HAVE_SYS_VFS_H.
4388 28. Installed the latest version of exipick from John Jetmore.
4390 29. In an address list, if the pattern was not a regular expression, an empty
4391 subject address (from a bounce message) matched only if the pattern was an
4392 empty string. Non-empty patterns were not even tested. This was the wrong
4393 because it is perfectly reasonable to use an empty address as part of a
4394 database query. An empty address is now tested by patterns that are
4395 lookups. However, all the other forms of pattern expect the subject to
4396 contain a local part and a domain, and therefore, for them, an empty
4397 address still always fails if the pattern is not itself empty.
4399 30. Exim went into a mad DNS loop when attempting to do a callout where the
4400 host was specified on an smtp transport, and looking it up yielded more
4401 than one IP address.
4403 31. Re-factored the code for checking spool and log partition space into a
4404 function that finds that data and another that does the check. The former
4405 is then used to implement four new variables: $spool_space, $log_space,
4406 $spool_inodes, and $log_inodes.
4408 32. The RFC2047 encoding function was originally intended for short strings
4409 such as real names; it was not keeping to the 75-character limit for
4410 encoded words that the RFC imposes. It now respects the limit, and
4411 generates multiple encoded words if necessary. To be on the safe side, I
4412 have increased the buffer size for the ${rfc2047: expansion operator from
4415 33. It is now permitted to omit both strings after an "if" condition; if the
4416 condition is true, the result is "true". As before, when the second string
4417 is omitted, a false condition yields an empty string. This makes it less
4418 cumbersome to write custom ACL and router conditions.
4420 34. Failure to deliver a bounce message always caused it to be frozen, even if
4421 there was an errors_to setting on the router. The errors_to setting is now
4424 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4425 canonical form (fully expanded) before being placed in
4426 $sender_host_address.
4428 36. The table in the code that translates DNS record types into text (T_A to
4429 "A" for instance) was missing entries for NS and CNAME. It is just possible
4430 that this could have caused confusion if both these types were looked up
4431 for the same domain, because the text type is used as part of Exim's
4432 per-process caching. But the chance of anyone hitting this buglet seems
4435 37. The dnsdb lookup has been extended in a number of ways.
4437 (1) There is a new type, "zns", which walks up the domain tree until it
4438 finds some nameserver records. It should be used with care.
4440 (2) There is a new type, "mxh", which is like "mx" except that it returns
4441 just the host names, not the priorities.
4443 (3) It is now possible to give a list of domains (or IP addresses) to be
4444 looked up. The behaviour when one of the lookups defers can be
4445 controlled by a keyword.
4447 (4) It is now possible to specify the separator character for use when
4448 multiple records are returned.
4450 38. The dnslists ACL condition has been extended: it is now possible to supply
4451 a list of IP addresses and/or domains to be looked up in a particular DNS
4454 39. Added log_selector=+queue_time_overall.
4456 40. When running the queue in the test harness, wait just a tad after forking a
4457 delivery process, to get repeatability of debugging output.
4459 41. Include certificate and key file names in error message when GnuTLS fails
4460 to set them up, because the GnuTLS error message doesn't include the name
4461 of the failing file when there is a problem reading it.
4463 42. Allow both -bf and -bF in the same test run.
4465 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4467 44. The "Exiscan patch" is now merged into the mainline Exim source.
4469 45. Sometimes the final signoff response after QUIT could fail to get
4470 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4471 < 0 before doing a fflush(). This bug looks as though it goes back to the
4472 introduction of TLS in release 3.20, but "sometimes" must have been rare
4473 because the tests only now provoked it.
4475 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4476 (this can affect the format of dates).
4478 47. exim_tidydb, when checking for the continued existence of a message for
4479 which it has found a message-specific retry record, was not finding
4480 messages that were in split spool directories. Consequently, it was
4481 deleting retry records that should have stayed in existence.
4483 48. Steve fixed some bugs in eximstats.
4485 49. The SPA authentication driver was not abandoning authentication and moving
4486 on to the next authenticator when an expansion was forced to fail,
4487 contradicting the general specification for all authenticators. Instead it
4488 was generating a temporary error. It now behaves as specified.
4490 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4491 (the order specifies the preference for clients). The order is now AES256,
4492 AES128, 3DES, ARCFOUR128.
4494 51. Small patch to Sieve code - explicitly set From: when generating an
4497 52. Exim crashed if a remote delivery caused a very long error message to be
4498 recorded - for instance if somebody sent an entire SpamAssassin report back
4499 as a large number of 550 error lines. This bug was coincidentally fixed by
4500 increasing the size of one of Exim's internal buffers (big_buffer) that
4501 happened as part of the Exiscan merge. However, to be on the safe side, I
4502 have made the code more robust (and fixed the comments that describe what
4505 53. Now that there can be additional text after "Completed" in log lines (if
4506 the queue_time_overall log selector is set), a one-byte patch to exigrep
4507 was needed to allow it to recognize "Completed" as not the last thing in
4510 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4511 patch that reportedly fixes this has been added. I am not expert enough to
4512 create a test for it. This is what the patch creator wrote:
4514 "I found a little strange behaviour of ldap code when working with
4515 Windows 2003 AD Domain, where users was placed in more than one
4516 Organization Units. When I tried to give exim partial DN, the exit code
4517 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4518 But simultaneously result of request was absolutely normal ldap result,
4519 so I produce this patch..."
4521 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4522 REFERENCE, so I have modified the code to exclude the patch when that macro
4525 55. Some experimental protocols are using DNS PTR records for new purposes. The
4526 keys for these records are domain names, not reversed IP addresses. The
4527 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4528 leaves it alone. Component reversal etc. now happens only for IP addresses.
4531 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4533 57. Double the size of the debug message buffer (to 2048) so that more of very
4534 long debug lines gets shown.
4536 58. The exicyclog utility now does better if the number of log files to keep
4537 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4539 59. Two changes related to the smtp_active_hostname option:
4541 (1) $smtp_active_hostname is now available as a variable.
4542 (2) The default for smtp_banner uses $smtp_active_hostname instead
4543 of $primary_hostname.
4545 60. The host_aton() function is supposed to be passed a string that is known
4546 to be a valid IP address. However, in the case of IPv6 addresses, it was
4547 not checking this. This is a hostage to fortune. Exim now panics and dies
4548 if the condition is not met. A case was found where this could be provoked
4549 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4550 components; fortuitously, this particular loophole had already been fixed
4551 by change 4.50/55 above.
4553 If there are any other similar loopholes, the new check in host_aton()
4554 itself should stop them being exploited. The report I received stated that
4555 data on the command line could provoke the exploit when Exim was running as
4556 exim, but did not say which command line option was involved. All I could
4557 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4558 running as the user.
4561 61. There was a buffer overflow vulnerability in the SPA authentication code
4562 (which came originally from the Samba project). I have added a test to the
4563 spa_base64_to_bits() function which I hope fixes it.
4566 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4567 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4569 63. The daemon start-up calls getloadavg() while still root for those OS that
4570 need the first call to be done as root, but it missed one case: when
4571 deliver_queue_load_max is set with deliver_drop_privilege. This is
4572 necessary for the benefit of the queue runner, because there is no re-exec
4573 when deliver_drop_privilege is set.
4575 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4576 This has been fixed.
4578 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4579 was in use, was not putting the data itself into the right store pool;
4580 consequently, it could be overwritten for a subsequent message in the same
4581 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4584 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4586 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4587 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4588 the function were treating the return as a boolean value, which happened to
4589 work because 0=false and not-0=true, but is not correct code.
4591 68. The host_aton() function was not handling scoped IPv6 addresses (those
4592 with, for example, "%eth0" on the end) correctly.
4594 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4595 negated items (that is, ~something) in unsigned ints. Some compilers
4596 apparently mutter when there is no cast.
4598 70. If an address verification called from an ACL failed, and did not produce a
4599 user-specific message (i.e. there was only a "system" message), nothing was
4600 put in $acl_verify_message. In this situation, it now puts the system
4603 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4604 session; change 4.31/43 added the unwanted input to the log line - except
4605 that it did not do this in the start of session case. It now does.
4607 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4608 This is wrong and can cause the other end to generate a synchronization
4609 error if it is another Exim or anything else that does the synchronization
4610 check. A QUIT command is no longer sent after a timeout.
4612 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4613 during host lookups.
4615 74. The defer_ok option of callout verification was not working if it was used
4616 when verifying addresses in header lines, that is, for this case:
4618 verify = header_sender/callout=defer_ok
4620 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4621 those file descriptors could be used for SMTP connections. If anything
4622 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4623 could be sent to the SMTP client, causing chaos. The daemon now opens
4624 stdin, stdout, and stderr to /dev/null when it puts itself into the
4627 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4628 log by default. The user can override this with suitable Perl magic.
4630 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4631 the log message when discard triggers, was not working for the DATA ACL or
4632 for the non-SMTP ACL.
4634 78. Error message wording change in sieve.c.
4636 79. If smtp_accept_max_per_host was set, the number of connections could be
4637 restricted to fewer than expected, because the daemon was trying to set up
4638 a new connection before checking whether the processes handling previous
4639 connections had finished. The check for completed processes is now done
4640 earlier. On busy systems, this bug wouldn't be noticed because something
4641 else would have woken the daemon, and it would have reaped the completed
4644 80. If a message was submitted locally by a user whose login name contained one
4645 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4646 It caused a spool format error. I have fixed the spool reading code. A
4647 related problem was that the "from" clause in the Received: line became
4648 illegal because of the space(s). It is now covered by ${quote_local_part.
4650 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4653 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4655 83. Patch from Sieve maintainer to fix unterminated string problem in
4656 "vacation" handling.
4658 84. Some minor changes to the Linux configuration files to help with other
4659 OS variants using glibc.
4661 85. One more patch for Sieve to update vacation handling to latest spec.
4664 ----------------------------------------------------
4665 See the note above about the 4.44 and 4.50 releases.
4666 ----------------------------------------------------
4672 1. Change 4.43/35 introduced a bug that caused file counts to be
4673 incorrectly computed when quota_filecount was set in an appendfile
4676 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4677 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4680 3. Give more explanation in the error message when the command for a transport
4681 filter fails to execute.
4683 4. There are several places where Exim runs a non-Exim command in a
4684 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4685 was being done only for the command run by the queryprogram router. It is
4686 now done for all such subprocesses. The other cases are: ${run, transport
4687 filters, and the commands run by the lmtp and pipe transports.
4689 5. Some older OS have a limit of 256 on the maximum number of file
4690 descriptors. Exim was using setrlimit() to set 1000 as a large value
4691 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4692 systems. I've change it so that if it can't get 1000, it tries for 256.
4694 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4695 was an oversight, and furthermore, ever since the addition of extra
4696 controls (e.g. 4.43/32), the checks on when to allow different forms of
4697 "control" were broken. There should now be diagnostics for all cases when a
4698 control that does not make sense is encountered.
4700 7. $recipients is now available in the predata ACL (oversight).
4702 8. Tidy the search cache before the fork to do a delivery from a message
4703 received from the command line. Otherwise the child will trigger a lookup
4704 failure and thereby defer the delivery if it tries to use (for example) a
4705 cached ldap connection that the parent has called unbind on.
4707 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4708 of $address_data from the recipient verification was clobbered by the
4709 sender verification.
4711 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4712 was its contents. (It was OK if the option was not defined at all.)
4714 11. A "Completed" log line is now written for messages that are removed from
4715 the spool by the -Mrm option.
4717 12. $host_address is now set to the target address during the checking of
4718 ignore_target_hosts.
4720 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4721 being passed; this would have caused $sender_host_name to have been used if
4722 matching the list had actually called for a host name (not very likely,
4723 since this list is usually IP addresses). A host name is now passed as
4726 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4727 code that specifies a non-restarting handler (typically sigaction() in
4728 modern systems) in an attempt to fix a rare and obscure crash bug.
4730 15. Narrowed the window for a race in the daemon that could cause it to ignore
4731 SIGCHLD signals. This is not a major problem, because they are used only to
4732 wake it up if nothing else does.
4734 16. A malformed maildirsize file could cause Exim to calculate negative values
4735 for the mailbox size or file count. Odd effects could occur as a result.
4736 The maildirsize information is now recalculated if the size or filecount
4739 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4740 support for a long time. Removed HAVE_SYS_VFS_H.
4742 18. Updated exipick to current release from John Jetmore.
4744 19. Allow an empty sender to be matched against a lookup in an address list.
4745 Previously the only cases considered were a regular expression, or an
4748 20. Exim went into a mad DNS lookup loop when doing a callout where the
4749 host was specified on the transport, if the DNS lookup yielded more than
4752 21. The RFC2047 encoding function was originally intended for short strings
4753 such as real names; it was not keeping to the 75-character limit for
4754 encoded words that the RFC imposes. It now respects the limit, and
4755 generates multiple encoded words if necessary. To be on the safe side, I
4756 have increased the buffer size for the ${rfc2047: expansion operator from
4759 22. Failure to deliver a bounce message always caused it to be frozen, even if
4760 there was an errors_to setting on the router. The errors_to setting is now
4763 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4764 canonical form (fully expanded) before being placed in
4765 $sender_host_address.
4767 24. Updated eximstats to version 1.33
4769 25. Include certificate and key file names in error message when GnuTLS fails
4770 to set them up, because the GnuTLS error message doesn't include the name
4771 of the failing file when there is a problem reading it.
4773 26. Expand error message when OpenSSL has problems setting up cert/key files.
4776 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4777 (this can affect the format of dates).
4779 28. exim_tidydb, when checking for the continued existence of a message for
4780 which it has found a message-specific retry record, was not finding
4781 messages that were in split spool directories. Consequently, it was
4782 deleting retry records that should have stayed in existence.
4784 29. eximstats updated to version 1.35
4785 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4786 1.35 - bugfix such that pie charts by volume are generated correctly
4788 30. The SPA authentication driver was not abandoning authentication and moving
4789 on to the next authenticator when an expansion was forced to fail,
4790 contradicting the general specification for all authenticators. Instead it
4791 was generating a temporary error. It now behaves as specified.
4793 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4794 (the order specifies the preference for clients). The order is now AES256,
4795 AES128, 3DES, ARCFOUR128.
4797 31. Small patch to Sieve code - explicitly set From: when generating an
4800 32. Exim crashed if a remote delivery caused a very long error message to be
4801 recorded - for instance if somebody sent an entire SpamAssassin report back
4802 as a large number of 550 error lines. This bug was coincidentally fixed by
4803 increasing the size of one of Exim's internal buffers (big_buffer) that
4804 happened as part of the Exiscan merge. However, to be on the safe side, I
4805 have made the code more robust (and fixed the comments that describe what
4808 33. Some experimental protocols are using DNS PTR records for new purposes. The
4809 keys for these records are domain names, not reversed IP addresses. The
4810 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4811 leaves it alone. Component reversal etc. now happens only for IP addresses.
4814 34. The host_aton() function is supposed to be passed a string that is known
4815 to be a valid IP address. However, in the case of IPv6 addresses, it was
4816 not checking this. This is a hostage to fortune. Exim now panics and dies
4817 if the condition is not met. A case was found where this could be provoked
4818 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4819 components; fortuitously, this particular loophole had already been fixed
4820 by change 4.50/55 or 4.44/33 above.
4822 If there are any other similar loopholes, the new check in host_aton()
4823 itself should stop them being exploited. The report I received stated that
4824 data on the command line could provoke the exploit when Exim was running as
4825 exim, but did not say which command line option was involved. All I could
4826 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4827 running as the user.
4830 35. There was a buffer overflow vulnerability in the SPA authentication code
4831 (which came originally from the Samba project). I have added a test to the
4832 spa_base64_to_bits() function which I hope fixes it.
4835 36. The daemon start-up calls getloadavg() while still root for those OS that
4836 need the first call to be done as root, but it missed one case: when
4837 deliver_queue_load_max is set with deliver_drop_privilege. This is
4838 necessary for the benefit of the queue runner, because there is no re-exec
4839 when deliver_drop_privilege is set.
4841 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4842 was in use, was not putting the data itself into the right store pool;
4843 consequently, it could be overwritten for a subsequent message in the same
4844 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4847 38. Sometimes the final signoff response after QUIT could fail to get
4848 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4849 < 0 before doing a fflush(). This bug looks as though it goes back to the
4850 introduction of TLS in release 3.20, but "sometimes" must have been rare
4851 because the tests only now provoked it.
4857 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4858 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4859 Now it returns an integer. A number of calls were still expecting a T/F
4860 return. Fortuitously, in all cases, the tests worked in OK situations,
4861 which is the norm. However, things would have gone wrong on any write
4862 failures on the smtp file descriptor. This function is used when sending
4863 messages over SMTP and also when doing verify callouts.
4865 2. When Exim is called to do synchronous delivery of a locally submitted
4866 message (the -odf or -odi options), it no longer closes stderr before doing
4869 3. Implemented the mua_wrapper option.
4871 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4873 5. Implemented the functions header_remove(), header_testname(),
4874 header_add_at_position(), and receive_remove_recipient(), and exported them
4877 6. If an ACL "warn" statement specified the addition of headers, Exim already
4878 inserted X-ACL-Warn: at the start if there was no header name. However, it
4879 was not making this test for the second and subsequent header lines if
4880 there were newlines in the string. This meant that an invalid header could
4881 be inserted if Exim was badly configured.
4883 7. Allow an ACL "warn" statement to add header lines at the start or after all
4884 the Received: headers, as well as at the end.
4886 8. Added the rcpt_4xx retry error code.
4888 9. Added postmaster_mailfrom=xxx to callout verification option.
4890 10. Added mailfrom=xxxx to the callout verification option, for verify=
4893 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4894 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4896 12. Inserted some casts to stop certain compilers warning when using pointer
4897 differences as field lengths or precisions in printf-type calls (mostly
4898 affecting debugging statements).
4900 13. Added optional readline() support for -be (dynamically loaded).
4902 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4903 same clock tick as a message's arrival, so that its received time was the
4904 same as the "first fail" time on the retry record, and that message
4905 remained on the queue past the ultimate address timeout, every queue runner
4906 would try a delivery (because it was past the ultimate address timeout) but
4907 after another failure, the ultimate address timeout, which should have then
4908 bounced the address, did not kick in. This was a "< instead of <=" error;
4909 in most cases the first failure would have been in the next clock tick
4910 after the received time, and all would be well.
4912 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4913 being recognized when the domain list was tested by the match_domain
4914 condition in an expansion string.
4916 16. Added the ${str2b64: operator.
4918 17. Exim was always calling setrlimit() to set a large limit for the number of
4919 processes, without checking whether the existing limit was already
4920 adequate. (It did check for the limit on file descriptors.) Furthermore,
4921 errors from getrlimit() and setrlimit() were being ignored. Now they are
4922 logged to the main and panic logs, but Exim does carry on, to try to do its
4923 job under whatever limits there are.
4925 18. Imported PCRE 5.0.
4927 19. Trivial typo in log message " temporarily refused connection" (the leading
4930 20. If the log selector return_path_on_delivery was set and an address was
4931 redirected to /dev/null, the delivery process crashed because it assumed
4932 that a return path would always be set for a "successful" delivery. In this
4933 case, the whole delivery is bypassed as an optimization, and therefore no
4936 21. Internal re-arrangement: the function for sending a challenge and reading
4937 a response while authentication was assuming a zero-terminated challenge
4938 string. It's now changed to take a pointer and a length, to allow for
4939 binary data in such strings.
4941 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4943 23. Exim was not respecting finduser_retries when seeking the login of the
4944 uid under which it was called; it was always trying 10 times. (The default
4945 setting of finduser_retries is zero.) Also, it was sleeping after the final
4946 failure, which is pointless.
4948 24. Implemented tls_on_connect_ports.
4950 25. Implemented acl_smtp_predata.
4952 26. If the domain in control=submission is set empty, Exim assumes that the
4953 authenticated id is a complete email address when it generates From: or
4954 Sender: header lines.
4956 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4957 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4958 chown and chgrp in /bin and hostname in /usr/bin.
4960 28. Exim was keeping the "process log" file open after each use, just as it
4961 does for the main log. This opens the possibility of it remaining open for
4962 long periods when the USR1 signal hits a daemon. Occasional processlog
4963 errors were reported, that could have been caused by this. Anyway, it seems
4964 much more sensible not to leave this file open at all, so that is what now
4967 29. The long-running daemon process does not normally write to the log once it
4968 has entered its main loop, and it closes the log before doing so. This is
4969 so that log files can straightforwardly be renamed and moved. However,
4970 there are a couple of unusual error situations where the daemon does write
4971 log entries, and I had neglected to close the log afterwards.
4973 30. The text of an SMTP error response that was received during a remote
4974 delivery was being truncated at 512 bytes. This is too short for some of
4975 the long messages that one sometimes sees. I've increased the limit to
4978 31. It is now possible to make retry rules that apply only when a message has a
4979 specific sender, in particular, an empty sender.
4981 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4982 it possible to be selective about when SMTP synchronization is enforced.
4984 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4986 32. Implemented hosts_connection_nolog.
4988 33. Added an ACL for QUIT.
4990 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4993 35. Added mailbox_size and mailbox_filecount to appendfile.
4995 36. Added control = no_multiline_responses to ACLs.
4997 37. There was a bug in the logic of the code that waits for the clock to tick
4998 in the case where the clock went backwards by a substantial amount such
4999 that the microsecond fraction of "now" was more than the microsecond
5000 fraction of "then" (but the whole seconds number was less).
5002 38. Added support for the libradius Radius client library this is found on
5003 FreeBSD (previously only the radiusclient library was supported).
5009 1. When certain lookups returned multiple values in the form name=value, the
5010 quoting of the values was not always being done properly. Specifically:
5011 (a) If the value started with a double quote, but contained no whitespace,
5013 (b) If the value contained whitespace other than a space character (i.e.
5014 tabs or newlines or carriage returns) it was not quoted.
5015 This fix has been applied to the mysql and pgsql lookups by writing a
5016 separate quoting function and calling it from the lookup code. The fix
5017 should probably also be applied to nisplus, ibase and oracle lookups, but
5018 since I cannot test any of those, I have not disturbed their existing code.
5020 2. A hit in the callout cache for a specific address caused a log line with no
5021 reason for rejecting RCPT. Now it says "Previous (cached) callout
5022 verification failure".
5024 3. There was an off-by-one bug in the queryprogram router. An over-long
5025 return line was truncated at 256 instead of 255 characters, thereby
5026 overflowing its buffer with the terminating zero. As well as fixing this, I
5027 have increased the buffer size to 1024 (and made a note to document this).
5029 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5030 when Exim is waiting for an SMTP response from a remote server, Exim
5031 restarts its select() call on the socket, thereby resetting its timeout.
5032 This is not a problem when such interrupts are rare. Somebody set up a cron
5033 job to run exiwhat every 2 minutes, which is less than the normal select()
5034 timeout (5 or 10 minutes). This meant that the select() timeout never
5035 kicked in because it was always reset. I have fixed this by comparing the
5036 time when an interrupt arrives with the time at the start of the first call
5037 to select(). If more time than the timeout has elapsed, the interrupt is
5038 treated as a timeout.
5040 5. Some internal re-factoring in preparation for the addition of Sieve
5041 extensions (by MH). In particular, the "personal" test is moved to a
5042 separate function, and given an option for scanning Cc: and Bcc: (which is
5043 not set for Exim filters).
5045 6. When Exim created an email address using the login of the caller as the
5046 local part (e.g. when creating a From: or Sender: header line), it was not
5047 quoting the local part when it contained special characters such as @.
5049 7. Installed new OpenBSD configuration files.
5051 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5052 try to make them clearer.
5054 9. Callout options, other than the timeout value, were being ignored when
5055 verifying sender addresses in header lines. For example, when using
5057 verify = header_sender/callout=no_cache
5059 the cache was (incorrectly) being used.
5061 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5062 only the Cygwin environment.
5064 11. When return_path_on_delivery was set as a log selector, if different remote
5065 addresses in the same message used different return paths and parallel
5066 remote delivery occurred, the wrong values would sometimes be logged.
5067 (Whenever a remote delivery process finished, the return path value from
5068 the most recently started remote delivery process was logged.)
5070 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5071 lines when AUTH and/or TLS are in use. This is the "received protocol"
5072 field. Exim used to use "asmtp" for authenticated SMTP, without any
5073 indication (in the protocol name) for TLS use. Now it follows the RFC and
5074 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5075 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5076 names appear in log lines as well as in Received: header lines.
5078 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5079 extensions, and comparison tests, and to fix some bugs.
5081 14. Changes to the "personal" filter test:
5083 (1) The test was buggy in that it was just doing the equivalent of
5084 "contains" tests on header lines. For example, if a user's address was
5085 anne@some.where, the "personal" test would incorrectly be true for
5087 To: susanne@some.where
5089 This test is now done by extracting each address from the header in turn,
5090 and checking the entire address. Other tests that are part of "personal"
5091 are now done using regular expressions (for example, to check local parts
5092 of addresses in From: header lines).
5094 (2) The list of non-personal local parts in From: addresses has been
5095 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5096 taken from the Sieve specification recommendations.
5098 (3) If the message contains any header line starting with "List-" it is
5099 treated as non-personal.
5101 (4) The test for "circular" in the Subject: header line has been removed
5102 because it now seems ill-conceived.
5104 15. Minor typos in src/EDITME comments corrected.
5106 16. Installed latest exipick from John Jetmore.
5108 17. If headers_add on a router specified a text string that was too long for
5109 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5110 of string_sprintf() is now avoided.
5112 18. $message_body_size was not set (it was always zero) when running the DATA
5113 ACL and the local_scan() function.
5115 19. For the "mail" command in an Exim filter, no default was being set for
5116 the once_repeat time, causing a random time value to be used if "once" was
5117 specified. (If the value happened to be <= 0, no repeat happened.) The
5118 default is now 0s, meaning "never repeat". The "vacation" command was OK
5119 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5120 (I found it when inspecting the code).
5122 20. There is now an overall timeout for performing a callout verification. It
5123 defaults to 4 times the callout timeout, which applies to individual SMTP
5124 commands during the callout. The overall timeout applies when there is more
5125 than one host that can be tried. The timeout is checked before trying the
5126 next host. This prevents very long delays if there are a large number of
5127 hosts and all are timing out (e.g. when the network connections are timing
5128 out). The value of the overall timeout can be changed by specifying an
5129 additional sub-option for "callout", called "maxwait". For example:
5131 verify = sender/callout=5s,maxwait=20s
5133 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5134 to the end before writing, but this should make it even safer).
5136 22. Exim was forgetting that it had advertised PIPELINING for the second and
5137 subsequent messages on an SMTP connection. It was also not resetting its
5138 memory on STARTTLS and an internal HELO.
5140 23. When Exim logs an SMTP synchronization error within a session, it now
5141 records whether PIPELINING has been advertised or not.
5143 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5144 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5145 rather than long int.
5147 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5149 26. Added the never_mail option to autoreply.
5155 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5156 crash if the getsockname() call failed; this can happen if a connection is
5157 closed very soon after it is established. The problem was simply in the
5158 order in which certain operations were done, causing Exim to try to write
5159 to the SMTP stream before it had set up the file descriptor. The bug has
5160 been fixed by making things happen in the correct order.
5166 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5167 before the connection was closed, thus losing the rejection response.
5169 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5170 some early Solaris releases, but causes trouble in current releases where
5171 socklen_t is defined.
5173 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5176 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5179 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5180 that releasing the top of it at the end releases what was used for sub-
5181 expansions (unless the block got too big). However, discard this block if
5182 the first thing is a variable or header, so that we can use its block when
5183 it is dynamic (useful for very large $message_headers, for example).
5185 6. Lookups now cache *every* query, not just the most recent. A new, separate
5186 store pool is used for this. It can be recovered when all lookup caches are
5187 flushed. Lookups now release memory at the end of their result strings.
5188 This has involved some general refactoring of the lookup sources.
5190 7. Some code has been added to the store_xxx() functions to reduce the amount
5191 of flapping under certain conditions.
5193 8. log_incoming_interface used to affect only the <= reception log lines. Now
5194 it causes the local interface and port to be added to several more SMTP log
5195 lines, for example "SMTP connection from", and rejection lines.
5197 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5199 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5201 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5202 could be overwritten at the end of the current message (or the start of a
5203 new message if it was set in a HELO ACL). The value is now preserved for
5204 the duration of the SMTP connection.
5206 12. If a transport had a headers_rewrite setting, and a matching header line
5207 contained an unqualified address, that address was qualified, even if it
5208 did not match any rewriting rules. The underlying bug was that the values
5209 of the flags that permit the existence of unqualified sender and recipient
5210 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5211 non-local messages, and by -bnq for local messages) were not being
5212 preserved with the message after it was received.
5214 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5215 "next input=" as part of the text comprising the host identity instead of
5216 the correct text. The code was using the same buffer for two different
5217 strings. However, depending on which order the printing function evaluated
5218 its arguments, the bug did not always show up. Under Linux, for example, my
5219 test suite worked just fine.
5221 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5222 doesn't work with some older versions of Perl. It has been changed to "my",
5223 which in any case is probably the better facility to use.
5225 15. A really picky compiler found some instances of statements for creating
5226 error messages that either had too many or two few arguments for the format
5229 16. The size of the buffer for calls to the DNS resolver has been increased
5230 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5231 for addresses that have a lot of PTR records. This alleviates a problem; it
5232 does not fully solve it.
5234 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5235 buffer now truncates the list and logs the incident, which is the same
5236 action as happens when Exim is looking up a host name and its aliases.
5237 Previously in this situation something unpredictable would happen;
5238 sometimes it was "internal error: store_reset failed".
5240 18. If a server dropped the connection unexpectedly when an Exim client was
5241 using GnuTLS and trying to read a response, the client delivery process
5242 crashed while trying to generate an error log message.
5244 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5245 string, for example:
5247 warn message = H1: something\nH2: something
5249 the text was added as a single header line from Exim's point of view
5250 though it ended up OK in the delivered message. However, searching for the
5251 second and subsequent header lines using $h_h2: did not work. This has been
5252 fixed. Similarly, if a system filter added multiple headers in this way,
5253 the routers could not see them.
5255 20. Expanded the error message when iplsearch is called with an invalid key to
5256 suggest using net-iplsearch in a host list.
5258 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5259 ACLs are no longer actually imposed (and a message to that effect is
5262 22. If a "gecos" field in a passwd entry contained escaped characters, in
5263 particular, if it contained a \" sequence, Exim got it wrong when building
5264 a From: or a Sender: header from that name. A second bug also caused
5265 incorrect handling when an unquoted " was present following a character
5266 that needed quoting.
5268 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5269 was not being matched caselessly.
5271 24. Arranged for all hyphens in the exim.8 source to be escaped with
5274 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5275 properly. Recipient callout cache records were still being keyed to include
5276 the sender, even when use_sender was set false. This led to far more
5277 callouts that were necessary. The sender is no longer included in the key
5278 when use_sender is false.
5280 26. Added "control = submission" modifier to ACLs.
5282 27. Added the ${base62d: operator to decode base 62 numbers.
5284 28. dnsdb lookups can now access SRV records.
5286 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5287 the configuration file.
5289 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5290 (-v) mode. This makes the output for a verbose queue run more intelligible.
5292 31. Added a use_postmaster feature to recipient callouts.
5294 32. Added the $body_zerocount variable, containing the number of binary zero
5295 bytes in the message body.
5297 33. The time of last modification of the "new" subdirectory is now used as the
5298 "mailbox time last read" when there is a quota error for a maildir
5301 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5303 35. Added +ignore_unknown as a special item in host lists.
5305 36. Code for decoding IPv6 addresses in host lists is now included, even if
5306 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5307 address was recognized as an IP address, but was then not correctly decoded
5308 into binary, causing unexpected and incorrect effects when compared with
5315 1. Very minor rewording of debugging text in manualroute to say "list of
5316 hosts" instead of "hostlist".
5318 2. If verify=header_syntax was set, and a header line with an unqualified
5319 address (no domain) and a large number of spaces between the end of the
5320 name and the colon was received, the reception process suffered a buffer
5321 overflow, and (when I tested it) crashed. This was caused by some obsolete
5322 code that should have been removed. The fix is to remove it!
5324 3. When running in the test harness, delay a bit after writing a bounce
5325 message to get a bit more predictability in the log output.
5327 4. Added a call to search_tidyup() just before forking a reception process. In
5328 theory, someone could use a lookup in the expansion of smtp_accept_max_
5329 per_host which, without the tidyup, could leave open a database connection.
5331 5. Added the variables $recipient_data and $sender_data which get set from a
5332 lookup success in an ACL "recipients" or "senders" condition, or a router
5333 "senders" option, similar to $domain_data and $local_part_data.
5335 6. Moved the writing of debug_print from before to after the "senders" test
5338 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5339 problems for message scanning, either using a data ACL, or using
5340 local_scan() because the Received: header was not generated till after they
5341 were called (in order to set the time as the time of reception completion).
5342 I have revised the way this works. The header is now generated after the
5343 body is received, but before the ACL or local_scan() are called. After they
5344 are run, the timestamp in the header is updated.
5350 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5351 before starting a queue runner without re-exec. This happened only when
5352 deliver_drop_privilege was set or when the Exim user was set to root. The
5353 effect of the bug was that timeouts during subsequent deliveries caused
5354 crashes instead of being properly handled. The handler is now left at its
5355 default (and expected) setting.
5357 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5358 message, again when deliver_drop_privilege is set or Exim is run as root.
5359 The bug described in (1) was not present in this case, but the tidying up
5360 of the other signals was missing. I have made the two cases consistent.
5362 3. The ignore_target_hosts setting on a manualroute router was being ignored
5363 for hosts that were looked up using the /MX notation.
5365 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5368 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5369 operated on the sender address. After changing the $sender_address to <>
5370 for the sender address verify, Exim was re-instated it as the original
5371 (before rewriting) address, but remembering that it had rewritten it, so it
5372 wasn't rewriting it again. This bug also had the effect of breaking the
5373 sender address verification caching when the sender address was rewritten.
5375 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5376 This has been changed so that if the ip literal address matches
5377 ignore_target_hosts, the router declines.
5379 7. Added expansion conditions match_domain, match_address, and match_local_
5380 part (NOT match_host).
5382 8. The placeholder for the Received: header didn't have a length field set.
5384 9. Added code to Exim itself and to exim_lock to test for a specific race
5385 condition that could lead to file corruption when using MBX delivery. The
5386 issue is with the lockfile that is created in /tmp. If this file is removed
5387 after a process has opened it but before that process has acquired a lock,
5388 there is the potential for a second process to recreate the file and also
5389 acquire a lock. This could lead to two Exim processes writing to the file
5390 at the same time. The added code performs the same test as UW imapd; it
5391 checks after acquiring the lock that its file descriptor still refers to
5392 the same named file.
5394 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5395 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5398 11. Added the smtp_active_hostname option. If used, this will typically be made
5399 to depend on the incoming interface address. Because $interface_address is
5400 not set up until the daemon has forked a reception process, error responses
5401 that can happen earlier (such as "too many connections") no longer contain
5404 12. If an expansion in a condition on a "warn" statement fails because a lookup
5405 defers, the "warn" statement is abandoned, and the next ACL statement is
5406 processed. Previously this caused the whole ACL to be aborted.
5408 13. Added the iplsearch lookup type.
5410 14. Added ident_timeout as a log selector.
5412 15. Added tls_certificate_verified as a log selector.
5414 16. Added a global option tls_require_ciphers (compare the smtp transport
5415 option of the same name). This controls incoming TLS connections.
5417 17. I finally figured out how to make tls_require_ciphers do a similar thing
5418 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5419 before starting the TLS session.
5421 18. Tabs are now shown as \t in -bP output.
5423 19. If the log selector return_path_on_delivery was set, Exim crashed when
5424 bouncing a message because it had too many Received: header lines.
5426 20. If two routers both had headers_remove settings, and the first one included
5427 a superfluous trailing colon, the final name in the first list and the
5428 first name in the second list were incorrectly joined into one item (with a
5429 colon in the middle).
5435 1. Added -C and -D options to the exinext utility, mainly to make it easier
5436 to include in the automated testing, but these could be helpful when
5437 multiple configurations are in use.
5439 2. The exinext utility was not formatting the output nicely when there was
5440 an alternate port involved in the retry record key, nor when there was a
5441 message id as well (for retries that were specific to a specific message
5442 and a specific host). It was also confused by IPv6 addresses, because of
5443 the additional colons they contain. I have fixed the IPv4 problem, and
5444 patched it up to do a reasonable job for IPv6.
5446 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5447 delivery, the log line now contains "pipelined" if PIPELINING was used.
5449 4. An SMTP transport process used to panic and die if the bind() call to set
5450 an explicit outgoing interface failed. This has been changed; it is now
5451 treated in the same way as a connect() failure.
5453 5. A reference to $sender_host_name in the part of a conditional expansion
5454 that was being skipped was still causing a DNS lookup. This no longer
5457 6. The def: expansion condition was not recognizing references to header lines
5458 that used bh_ and bheader_.
5460 7. Added the _cache feature to named lists.
5462 8. The code for checking quota_filecount in the appendfile transport was
5463 allowing one more file than it should have been.
5465 9. For compatibility with Sendmail, the command line option
5473 and sets the incoming protocol and host name (for trusted callers). The
5474 host name and its colon can be omitted when only the protocol is to be set.
5475 Note the Exim already has two private options, -pd and -ps, that refer to
5476 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5477 "s", but I don't think that's a major issue.
5479 10. A number of refactoring changes to the code, none of which should affect
5482 (a) The number of logging options was getting close to filling up the
5483 32-bit word that was used as a bit map. I have split them into two classes:
5484 those that are passed in the argument to log_write(), and those that are
5485 only ever tested independently outside of that function. These are now in
5486 separate 32-bit words, so there is plenty of room for expansion again.
5487 There is no change in the user interface or the logging behaviour.
5489 (b) When building, for example, log lines, the code previously used a
5490 macro that called string_cat() twice, in order to add two strings. This is
5491 not really sufficiently general. Furthermore, there was one instance where
5492 it was actually wrong because one of the argument was used twice, and in
5493 one call a function was used. (As it happened, calling the function twice
5494 did not affect the overall behaviour.) The macro has been replaced by a
5495 function that can join an arbitrary number of extra strings onto a growing
5498 (c) The code for expansion conditions now uses a table and a binary chop
5499 instead of a serial search (which was left over from when there were very
5500 few conditions). Also, it now recognizes conditions like "pam" even when
5501 the relevant support is not compiled in: a suitably worded error message is
5502 given if an attempt is made to use such a condition.
5504 11. Added ${time_interval:xxxxx}.
5506 12. A bug was causing one of the ddress fields not to be passed back correctly
5507 from remote delivery subprocesses. The field in question was not being
5508 subsequently used, so this caused to problems in practice.
5510 13. Added new log selectors queue_time and deliver_time.
5512 14. Might have fixed a bug in maildirsizefile handling that threw up
5513 "unexpected character" debug warnings, and recalculated the data
5514 unnecessarily. In any case, I expanded the warning message to give more
5517 15. Added the message "Restricted characters in address" to the statements in
5518 the default ACL that block characters like @ and % in local parts.
5520 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5521 Three changes have been made:
5523 (a) There was a serious bug; a negative response to MAIL caused the whole
5524 recipient domain to be cached as invalid, thereby blocking all messages
5525 to all local parts at the same domain, from all senders. This bug has
5526 been fixed. The domain is no longer cached after a negative response to
5527 MAIL if the sender used is not empty.
5529 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5532 (c) A new callout option, "use_sender" has been added for people who want
5533 the modified behaviour.
5539 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5542 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5543 indeed breaks things for older releases.
5545 3. Added additional logging to the case where there is a problem reading data
5546 from a filter that is running in a subprocess using a pipe, in order to
5547 try to track down a specific problem.
5549 4. Testing facility fudge: when running in the test harness and attempting
5550 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5551 getting "No route to host". Convert this to a timeout.
5553 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5556 6. Some OS don't have socklen_t but use size_t instead. This affects the
5557 fifth argument of getsockopt() amongst other things. This is now
5558 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5559 can be set for individual OS. I have set it for SunOS5, OSF1, and
5560 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5561 some earlier ones do not.
5563 7. Change 4.30/15 was not doing the test caselessly.
5565 8. The standard form for an IPv6 address literal was being rejected by address
5566 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5567 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5568 this, as well as the form without the "IPv6" on the front (but only when
5569 address literals are enabled, of course).
5571 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5573 10. Exim crashed if a message with an empty sender address specified by -f
5574 encountered a router with an errors_to setting. This could be provoked only
5575 by a command such as
5579 where an empty string was supplied; "<>" did not hit this bug.
5581 11. Installed PCRE release 4.5.
5583 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5584 remained set. It is now erased.
5586 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5587 times from message ids (which are base 36 rather than the normal 62).
5589 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5590 were being counted as actual protocol errors, and logged if the log
5591 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5592 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5593 and DATA following a set of rejected RCPTs do not count as protocol errors.
5594 In other words, Exim assumes they were pipelined, though this may not
5595 actually be the case. Of course, in all cases the client gets an
5596 appropriate error code.
5598 15. If a lookup fails in an ACL condition, a message about the failure may
5599 be available; it is used if testing the ACL cannot continue, because most
5600 such messages specify what the cause of the deferral is. However, some
5601 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5602 that caused an old message to be retained and used if a later statement
5603 caused a defer, replacing the real cause of the deferral.
5605 16. If an IP address had so many PTR records that the DNS lookup buffer
5606 was not large enough to hold them, Exim could crash while trying to process
5607 the truncated data. It now detects and logs this case.
5609 17. Further to 4.21/58, another change has been made: if (and only if) the
5610 first line of a message (the first header line) ends with CRLF, a bare LF
5611 in a subsequent header line has a space inserted after it, so as not to
5612 terminate the header.
5614 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5615 unnecessarily, used atoi() instead of strtol(), and didn't check the
5616 termination when getting file sizes from file names by regex.
5618 19. Completely re-implemented the support for maildirsize files, in the light
5619 of a number of problems with the previous contributed implementation
5620 (4.30/29). In particular:
5622 . If the quota is zero, the maildirsize file is maintained, but no quota is
5625 . If the maildir directory does not exist, it is created before any attempt
5626 to write a maildirsize file.
5628 . The quota value in the file is just a cache; if the quota is changed in
5629 the transport, the new value overrides.
5631 . A regular expression is available for excluding directories from the
5634 20. The autoreply transport checks the characters in options that define the
5635 message's headers; it allows continued headers, but it was checking with
5636 isspace() after an embedded newline instead of explicitly looking for a
5639 21. If all the "regular" hosts to which an address was routed had passed their
5640 expiry times, and had not reached their retry times, the address was
5641 bounced, even if fallback hosts were defined. Now Exim should go on to try
5644 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5645 equivalent code in the SMTP transport. Some hosts send humungous responses
5646 to HELO/EHLO, more than 1024 it seems.
5648 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5649 gives compiler warnings in some environments. I've now done it "properly",
5652 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5653 (because of problems with the built-in one) was declared to return uschar *
5654 instead of char *, causing compiler failure.
5656 25. Fixed a file descriptor leak when processing alias/forward files.
5658 26. Fixed a minor format string issue in dbfn.c.
5660 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5662 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5663 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5666 29. When a sender address is verified, it is cached, to save repeating the test
5667 when there is more than one recipient in a message. However, when the
5668 verification involves a callout, it is possible for different callout
5669 options to be set for different recipients. It is too complicated to keep
5670 track of this in the cache, so now Exim always runs a verification when a
5671 callout is required, relying on the callout cache for the optimization.
5672 The overhead is duplication of the address routing, but this should not be
5675 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5676 to be verified with callout=postmaster, and the main callout worked but the
5677 postmaster check failed, the verification correctly failed. However, if a
5678 subsequent RCPT command asked for sender verification *without* the
5679 postmaster check, incorrect caching caused this verification also to fail,
5682 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5683 it was not caching the DNS options (qualify_single, search_parents) that
5684 were used when the lookup failed. A subsequent lookup with different
5685 options therefore always gave the same answer, though there were cases
5686 where it should not have. (Example: a "domains = !$mx_any" option on a
5687 dnslookup router: the "domains" option is always processed without any
5688 widening, but the router might have qualify_single set.) Now Exim uses the
5689 cached value only when the same options are set.
5691 32. Added John Jetmore's "exipick" utility to the distribution.
5693 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5694 than a timeout (e.g. a certificate is required, and is not provided), an
5695 Exim server now closes the connection immediately. Previously it waited for
5696 the client to close - but if the client is SSL, it seems that they each
5697 wait for each other, leading to a delay before one of them times out.
5699 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5700 maintained 0.8.x compatibility because I don't think many are using it, and
5701 it is clearly obsolete.
5703 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5706 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5707 client certificate was expired. A simple patch fixes this, though I don't
5708 understand the full logic of why the verify callback is called multiple
5711 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5712 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5713 which causes problems with some clients (such as the Certicom SSL Plus
5714 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5715 disables the coutermeasure allowing Eudora to connect."
5717 38. Exim was not checking that a write() to a log file succeeded. This could
5718 lead to Bad Things if a log got too big, in particular if it hit a file
5719 size limit. Exim now panics and dies if it cannot write to a log file, just
5720 as it does if it cannot open a log file.
5722 39. Modified OS/Makefile-Linux so that it now contains
5724 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5726 The two -D definitions ensure that Exim is compiled with large file
5727 support, which makes it possible to handle log files that are bigger than
5730 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5731 instance) a domain was checked against a named list that involved a lookup,
5732 causing $domain_data to be set, then another domain was checked against the
5733 same list, then the first domain was re-checked, the value of $domain_data
5734 after the final check could be wrong. In particular, if the second check
5735 failed, it could be set empty. This bug probably also applied to
5738 41. The strip_trailing_dot option was not being applied to the address given
5739 with the -f command-line option.
5741 42. The code for reading a message's header from the spool was incrementing
5742 $received_count, but never initializing it. This meant that the value was
5743 incorrect (doubled) while delivering a message in the same process in which
5744 it was received. In the most common configuration of Exim, this never
5745 happens - a fresh exec is done - but it can happen when
5746 deliver_drop_privilege is set.
5748 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5749 it now includes up to 150 characters of the unexpected data in the log
5752 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5753 and building data strings. The size of both of these buffers was 10 000
5754 bytes - far larger than anybody would *ever* want, thought I. Needless to
5755 say, somebody hit the limit. I have increased the maximum line length to
5756 20 000 and the maximum data length of concatenated lines to 100 000. I have
5757 also fixed two bugs, because there was no checking on these buffers. Tsk,
5758 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5759 buffer is too small.
5761 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5762 lsearch lookups. Now it does.
5764 46. When parsing a route_list item in a manualroute router, a fixed-length
5765 buffer was used for the list of hosts. I made this 1024 bytes long,
5766 thinking that nobody would ever have a list of hosts that long. Wrong.
5767 Somebody had a whole pile of complicated expansion conditions, and the
5768 string was silently truncated, leading to an expansion error. It turns out
5769 that it is easier to change to an unlimited length (owing to other changes
5770 that have happened since this code was originally written) than to build
5771 structure for giving a limitation error. The length of the item that
5772 expands into the list of hosts is now unlimited.
5774 47. The lsearch lookup could not handle data where the length of text line was
5775 more than 4095 characters. Such lines were truncated, leading to shortened
5776 data being returned. It should now handle lines of any length.
5778 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5779 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5782 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5783 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5784 no longer happens in comment lines. A list of replacements is now placed
5785 at the head of all of the source files, except those whose only change is
5786 to replace PERL_COMMAND in the very first #! line.
5788 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5789 messages on the queue, with a bottom-up merge sort, using code contributed
5790 by Michael Haardt. This should make operations like -bp somewhat faster on
5791 large queues. It won't affect queue runners, except when queue_run_in_order
5794 51. Installed eximstats 1.31 in the distribution.
5796 52. Added support for SRV lookups to the dnslookup router.
5798 53. If an ACL referred to $message_body or $message_body_end, the value was not
5799 reset for any messages that followed in the same SMTP session.
5801 54. The store-handling optimization for building very long strings was not
5802 differentiating between the different store pools. I don't think this
5803 actually made any difference in practice, but I've tidied it.
5805 55. While running the routers to verify a sender address, $sender_address
5806 was still set to the sender address. This is wrong, because when routing to
5807 send a bounce to the sender, it would be empty. Therefore, I have changed
5808 it so that, while verifying a sender address, $sender_address is set to <>.
5809 (There is no change to what happens when verifying a recipient address.)
5811 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5812 target A or AAAA records (if not already returned) without resetting the
5813 qualify_single or search_parents options of the DNS resolver. These are
5814 inappropriate in this case because the targets of MX and SRV records must
5815 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5816 target that, when qualified, matched something in the local domain. These
5817 two options are now turned off when doing these lookups.
5819 57. It seems that at least some releases of Reiserfs (which does not have the
5820 concept of a fixed number of inodes) returns zero and not -1 for the
5821 number of available inodes. This interacted badly with check_spool_inodes,
5822 which assumed that -1 was the "no such thing" setting. What I have done is
5823 to check that the total number of inodes is greater than zero before doing
5824 the test of how many are available.
5826 58. When a "warn" ACL statement has a log_message modifier, the message is
5827 remembered, and not repeated. This is to avoid a lot of repetition when a
5828 message has many recipients that cause the same warning to be written.
5829 However, Exim was preserving the list of already written lines for an
5830 entire SMTP session, which doesn't seem right. The memory is now reset if a
5831 new message is started.
5833 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5834 debugging output unless log_rewrite was also set.
5836 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5837 of (int)(handle) when we know that handle contains (void *)(-1).
5839 61. The Exim daemon panic-logs an error return when it closes the incoming
5840 connection. However "connection reset by peer" seems to be common, and
5841 isn't really an error worthy of noting specially, so that particular error
5844 62. When Exim is trying to find all the local interfaces, it used to panic and
5845 die if the ioctl to get the interface flags failed. However, it seems that
5846 on at least one OS (Solaris 9) it is possible to have an interface that is
5847 included in the list of interfaces, but for which you get a failure error
5848 for this call. This happens when the interface is not "plumbed" into a
5849 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5850 failure of the "get flags" call assumes that the interface is down.
5852 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5853 makes life easier for people who are doing arithmetic on fields extracted
5854 from dates, where you often get leading zeros that should not be
5855 interpreted as octal.
5857 64. Added qualify_domain to the redirect router, to override the global
5860 65. If a pathologically long header line contained very many addresses (the
5861 report of this problem mentioned 10 000) and each of them was rewritten,
5862 Exim could use up a very large amount of memory. (It kept on making new
5863 copies of the header line as it rewrote, and never released the old ones.)
5864 At the expense of a bit more processing, the header rewriting function has
5865 been changed so that it no longer eats memory in this way.
5867 66. The generation of the Received: header has been moved from the time that a
5868 message starts to be received, to the time that it finishes. The timestamp
5869 in the Received: header should now be very close to that of the <= log
5870 line. There are two side-effects of this change:
5872 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5873 logged header lines no longer include the local Received: line, because
5874 it has not yet been created. The same applies to a copy of the message
5875 that is returned to a non-SMTP sender when a message is rejected.
5877 (b) When a filter file is tested using -bf, no additional Received: header
5878 is added to the test message. After some thought, I decided that this
5881 This change does not affect the value of $received_for. It is still set
5882 after address rewriting, but before local_scan() is called.
5884 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5886 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5887 gave an unhelpful panic error message, and a defer error. I have managed to
5888 change this behaviour so that it now rejects any supplied certificate,
5889 which seems right, as the list of acceptable certificates is empty.
5891 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5892 gave an unhelpful defer error. I have not managed to make this reject any
5893 supplied certificates, but the error message it gives is "no certificate
5894 supplied", which is not helpful.
5896 70. exigrep's output now also includes lines that are not associated with any
5897 message, but which match the given pattern. Implemented by a patch from
5898 Martin Sluka, which also tidied up the Perl a bit.
5900 71. Recipient callout verification, like sender verification, was using <> in
5901 the MAIL FROM command. This isn't really the right thing, since the actual
5902 sender may affect whether the remote host accepts the recipient or not. I
5903 have changed it to use the actual sender in the callout; this means that
5904 the cache record is now keyed on a recipient/sender pair, not just the
5905 recipient address. There doesn't seem to be a real danger of callout loops,
5906 since a callout by the remote host to check the sender would use <>.
5907 [SEE ABOVE: changed after hitting problems.]
5909 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5910 temporary errors. However, in the case of such a code being given after
5911 the end of a data transmission (i.e. after ".") Exim was failing to write
5912 a retry record for the message. (Yes, there was some broken host that was
5913 actually sending 8xx at this point.)
5915 73. An unknown lookup type in a host list could cause Exim to panic-die when
5916 the list was checked. (An example that provoked this was putting <; in the
5917 middle of a list instead of at the start.) If this happened during a DATA
5918 ACL check, a -D file could be left lying around. This kind of configuration
5919 error no longer causes Exim to die; instead it causes a defer error. The
5920 incident is still logged to the main and panic logs.
5922 74. Buglet left over from Exim 3 conversion. The message "too many messages
5923 in one connection" was written to the rejectlog but not the mainlog, except
5924 when address rewriting (yes!) was being logged.
5926 75. Added write_rejectlog option.
5928 76. When a system filter was run not as root (that is, when system_filter_user
5929 was set), the values of the $n variables were not being returned to the
5930 main process; thus, they were not subsequently available in the $sn
5933 77. Added +return_path_on_delivery log selector.
5935 78. A connection timeout was being treated differently from recipients deferred
5936 when testing hosts_max_try with a message that was older than the host's
5937 retry timeout. (The host should not be counted, thus allowing all hosts to
5938 be tried at least once before bouncing.) This may have been the cause of an
5939 occasionally reported bug whereby a message would remain on the queue
5940 longer than the retry timeout, but would be bounced if a delivery was
5941 forced. I say "may" because I never totally pinned down the problem;
5942 setting up timeout/retry tests is difficult. See also the next item.
5944 79. The ultimate address timeout was not being applied to errors that involved
5945 a combination of host plus message (for example, a timeout on a MAIL
5946 command). When an address resolved to a number of possible hosts, and they
5947 were not all tried for each delivery (e.g. because of hosts_max_try), a
5948 message could remain on the queue longer than the retry timeout.
5950 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5953 81. Fixed an obscure SMTP outgoing bug which required at least the following
5954 conditions: (a) there was another message waiting for the same server;
5955 (b) the server returned 5xx to all RCPT commands in the first message so
5956 that the message was not completed; (c) the server dropped the connection
5957 or gave a negative response to the RSET that Exim sends to abort the
5958 transaction. The observed case was a dropped connection after DATA that had
5959 been sent in pipelining mode. That is, the server had advertised PIPELINING
5960 but was not implementing it correctly. The effect of the bug was incorrect
5961 behaviour, such as trying another host, and this could lead to a crash.
5967 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5968 and daemon.c were passed as pointers to ints; they should have been
5969 pointers to socklen_t variables (which are typically unsigned ints).
5971 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5974 3. Fixed a really odd bug that affected only the testing scheme; patching a
5975 certain fixed string in the binary changed the value of another string that
5976 happened to be identical to the end of the original first string.
5978 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5979 name", it returns that address as the IP address. On some operating
5980 systems (e.g. Solaris), it also passes back the IP address string as the
5981 "host name". However, on others (e.g. Linux), it passes back an empty
5982 string. Exim wasn't checking for this, and was changing the host name to an
5983 empty string, assuming it had been canonicalized.
5985 5. Although rare, it is permitted to have more than one PTR record for a given
5986 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5987 all the names associated with an address, because they do in Solaris.
5988 However, it seems that they do not in Linux for data that comes from the
5989 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5990 I found this out when I moved to a new Linux workstation and tried to run
5991 the Exim test suite.
5993 To get round this problem I have changed the code so that it now does its
5994 own call to the DNS to look up PTR records when searching for a host name.
5995 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5996 addresses that are only in /etc/hosts are still found.
5998 This behaviour is, however, controlled by an option called host_lookup_
5999 order, which defaults to "bydns:byaddr". If people want to use the other
6000 order, or indeed, just use one or the other means of lookup, they can
6001 specify it in this variable.
6003 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6004 some operating systems, this comes back from gethostbyaddr() as an empty
6005 string, and this is what Exim used to test for. However, it seems that in
6006 other systems, "." is yielded. Exim now tests for this case too.
6008 7. The values of check_spool_space and check_log_space are now held internally
6009 as a number of kilobytes instead of an absolute number of bytes. If a
6010 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6011 kilobyte. This means that much larger values can be stored.
6013 8. Exim monitor: an attempt to get the action menu when not actually pointing
6014 at a message produces an empty menu entitled "No message selected". This
6015 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6016 no entries in it ("Shell widget menu has zero width and/or height"). So I
6017 have added a single, blank menu entry in this case.
6019 9. Added ${quote_local_part.
6021 10. MIME decoding is now applied to the contents of Subject: header lines when
6024 11. Now that a reference to $sender_host_address automatically causes a reverse
6025 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6026 host lookup before query-style lookups in lists that might use this
6027 variable. This has therefore been abolished, and the "net-" prefix is no
6028 longer necessary for query-style lookups.
6030 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6031 have been a typo for LFLAGS, so it has been changed.
6033 13. The install script calls Exim with "-C /dev/null" in order to find the
6034 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6035 to be output. However, since Exim outputs its version number before the
6036 error, it didn't break the script. It just looked ugly. I fixed this by
6037 always allowing "-C /dev/null" if the caller is root.
6039 14. Ignore overlarge ACL variable number when reading spool file - insurance
6040 against a later release with more variables having written the file.
6042 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6043 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6044 this, as well as the form without the "IPv6" on the front.
6046 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6047 OS/Makefile-Darwin file.
6049 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6050 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6052 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6053 variable. However, some OS use ints for inodes. I've added cast to long int
6054 to get rid of the compiler warning.
6056 19. I had forgotten to lock out "/../" in configuration file names when
6057 ALT_CONFIG_PREFIX was set.
6059 20. Routers used for verification do not need to specify transports. However,
6060 if such a router generated a host list, and callout was configured, Exim
6061 crashed, because it could not find a port number from the (non-existent)
6062 transport. It now assumes port 25 in this circumstance.
6064 21. Added the -t option to exigrep.
6066 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6067 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6068 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6069 them is set, LOOKUP_LSEARCH is forced.
6071 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6073 24. Added sender and host information to the "rejected by local_scan()" log
6074 line; previously there was no indication of these.
6076 25. Added .include_if_exists.
6078 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6079 Linux. It turns out that not all file systems support this. Apparently some
6080 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6081 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6082 not supported on the file descriptor, is now ignored when Exim is trying to
6083 sync a directory. This applies only to Linux.
6085 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6087 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6088 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6091 29. Added support for maildirsize files from supplied patch (modified a bit).
6093 30. The use of :fail: followed by an empty string could lead Exim to respond to
6094 sender verification failures with (e.g.):
6096 550 Verification failed for <xxx>
6097 550 Sender verify failed
6099 where the first response line was missing the '-' that indicates it is not
6100 the final line of the response.
6102 31. The loop for finding the name of the user that called Exim had a hardwired
6103 limit of 10; it now uses the value of finduser_retries, which is used for
6104 all other user lookups.
6106 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6109 33. Exim was neglecting to zero errno before one call of strtol() when
6110 expanding a string and expecting an integer value. On some systems this
6111 resulted in spurious "integer overflow" errors. Also, it was casting the
6112 result into an int without checking.
6114 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6115 did not work. The code looks as if it has *never* worked, though it appears
6116 to have been documented since at least release 1.62. I have made it work.
6118 35. The "timeout_DNS" error in retry rules, also documented since at least
6119 1.62, also never worked. As it isn't clear exactly what this means, and
6120 clearly it isn't a major issue, I have abolished the feature by treating it
6121 as "timeout", and writing a warning to the main and panic logs.
6123 36. The display of retry rules for -brt wasn't always showing the error code
6126 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6127 timeout_connect_A, timeout_connect_MX.
6129 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6130 to the empty sender.
6132 39. The daemon was not analysing the content of -oX till after it had closed
6133 stderr and disconnected from the controlling terminal. This meant that any
6134 syntax errors were only noted on the panic log, and the return code from
6135 the command was 0. By re-arranging the code a little, I've made the
6136 decoding happen first, so such errors now appear on stderr, and the return
6137 code is 1. However, the actual setting up of the sockets still happens in
6138 the disconnected process, so errors there are still only recorded on the
6141 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6142 connections (as happens on some IP stacks) was logged at start up time as
6143 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6144 it from "IPv6 and IPv4", which means that two separate sockets are being
6147 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6148 says whether AF_INET or AF_INET6 was passed as an argument.
6150 42. Exiwhat output was messed up when time zones were included in log
6153 43. Exiwhat now gives more information about the daemon's listening ports,
6154 and whether -tls-on-connect was used.
6156 44. The "port" option of the smtp transport is now expanded.
6158 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6159 silently ignored. Now an error message is written to the main and panic
6162 46. There's a new ACL modifier called "logwrite" which writes to a log file
6163 as soon as it is encountered.
6165 47. Added $local_user_uid and $local_user_gid at routing time.
6167 48. Exim crashed when trying to verify a sender address that was being
6170 49. Exim was recognizing only a space character after ".include". It now also
6171 recognizes a tab character.
6173 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6174 extracting the relevant information from the specification. The man page no
6175 longer contains scrambled data for the -d option, and I've added a section
6176 at the front about calling Exim under different names.
6178 51. Added "extra_headers" argument to the "mail" command in filter files.
6180 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6183 53. Installed eximstats 1.29.
6185 54. Added transport_filter_timeout as a generic transport option.
6187 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6188 Cc: header lines. This was required by RFC 822, but it not required by RFC
6191 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6192 incoming messages that did not have them. Now it does so only if the
6193 message originates locally, that is, if there is no associated remote host
6194 address. When Resent- header lines are present, this applies to the Resent-
6195 lines rather than the non-Resent- lines.
6197 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6198 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6200 58. Messages for configuration errors now include the name of the main
6201 configuration file - useful now that there may be more than one file in a
6202 list (.included file names were always shown).
6204 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6205 for those rare installations that do not start the daemon as root or run it
6206 setuid root. I've cut out the call to initgroups() if the daemon is not
6209 60. The Exim user and group can now be bound into the binary as text strings
6210 that are looked up at the start of Exim's processing.
6212 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6214 62. Added $mailstore_basename variable.
6216 63. Installed patch to sieve.c from Michael Haardt.
6218 64. When Exim failed to open the panic log after failing to open the main log,
6219 the original message it was trying to log was written to stderr and debug
6220 output, but if they were not available (the usual case in production), it
6221 was lost. Now it is written to syslog before the two lines that record the
6222 failures to open the logs.
6224 65. Users' Exim filters run in subprocesses under the user's uid. It is
6225 possible for a "deliver" command or an alias in a "personal" command to
6226 provoke an address rewrite. If logging of address rewriting is configured,
6227 this fails because the process is not running as root or exim. There may be
6228 a better way of dealing with this, but for the moment (because 4.30 needs
6229 to be released), I have disabled address rewrite logging when running a
6230 filter in a non-root, non-exim process.
6236 1. The buildconfig auxiliary program wasn't quoting the value set for
6237 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6238 not defined. This bug was masked in 4.22 by the effect that was fixed in
6241 2. Some messages that were rejected after a message id was allocated were
6242 shown as "incomplete" by exigrep. It no longer does this for messages that
6243 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6245 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6246 have allow_domain_literals set, the ID did not get logged in the <= line.
6247 Domain literals are now always recognized in Message-ID: header lines.
6249 4. The first argument for a ${extract expansion item is the key name or field
6250 number. Leading and trailing spaces in this item were not being ignored,
6251 causing some misleading effects.
6253 5. When deliver_drop_privilege was set, single queue runner processes started
6254 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6255 same command in the process it spins off) were not dropping privilege.
6257 6. When the daemon running as "exim" started a queue runner, it always
6258 re-executed Exim in the spun-off process. This is a waste of effort when
6259 deliver_drop_privilege is set. The new process now just calls the
6260 queue-runner function directly.
6266 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6269 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6270 ignored. Though the use of -oP was forcing the writing of a pid file, it
6271 was always written to the default place.
6273 3. If the message "no IP address found for host xxxx" is generated during
6274 incoming verification, it is now followed by identification of the incoming
6275 connection (so you can more easily find what provoked it).
6277 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6279 5. Added some features to "harden" Exim a bit more against certain attacks:
6281 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6282 be put in Local/Makefile. This is like the never_users runtime option,
6283 but it cannot be overridden. The default setting is "root".
6285 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6286 prefix string with which any file named in a -C command line option
6289 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6290 is retained for -C and -D only if the caller of Exim is root. Without
6291 it, the exim user may also use -C and -D and retain privilege.
6293 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6294 command line option is disabled.
6296 6. Macro names set by the -D option must start with an upper case letter, just
6297 like macro names defined in the configuration file.
6299 7. Added "dereference=" facility to LDAP.
6301 8. Two instances of the typo "uknown" in the source files are fixed.
6303 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6304 the Configure-Makefile script screwed up while processing it.
6306 10. Incorporated PCRE 4.4.
6308 11. The SMTP synchronization check was not operating right at the start of an
6309 SMTP session. For example, it could not catch a HELO sent before the client
6310 waited for the greeting. There is now a check for outstanding input at the
6311 point when the greeting is written. Because of the duplex, asynchronous
6312 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6313 way, but not yet received, when the check is performed.
6315 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6316 on TCP/IP sockets, because this apparently causes some broken clients to
6319 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6320 unchanged) from the Cygwin maintainer.
6322 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6323 was supported instead of testing for mbx. Effectively a typo.
6325 15. The spa authenticator server code was not checking that the input it
6326 received was valid base64.
6328 16. The debug output line for the "set" modifier in ACLs was not showing the
6329 name of the variable that was being set.
6331 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6333 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6334 lookup on its own. Something else was needed to trigger the lookup. For
6335 example, a match in host_lookup or the need for a host name in a host list.
6336 Now, if $sender_host_name is referenced and the host name has not yet been
6337 looked up, a lookup is performed. If the lookup fails, the variable remains
6338 empty, and $host_lookup_failed is set to "1".
6340 19. Added "eqi" as a case-independent comparison operator.
6342 20. The saslauthd authentication condition could segfault if neither service
6343 nor realm was specified.
6345 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6346 error message that was logged was misleading, and incoming SMTP
6347 connections were dropped. The message is now more accurate, and temporary
6348 errors are given to SMTP connections.
6350 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6351 (except RSET) with a 421 error, until QUIT is received. However, it was
6352 failing to send a response to QUIT.
6354 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6355 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6356 ACL is now run after the helo_try_verify_hosts code.
6358 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6359 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6360 case-independent, so other case variants are also recognized). Apparently
6361 some systems use these upper case variants.
6363 25. If more than two messages were waiting for the same host, and a transport
6364 filter was specified for the transport, Exim sent two messages over the
6365 same TCP/IP connection, and then failed with "socket operation on non-
6366 socket" when it tried to send the third.
6368 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6370 27. The extern definition of crypt16() in expand.c was not being excluded when
6371 the OS had its own crypt16() function.
6373 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6374 as a preferred synonym for return_size_limit, both as an option and as an
6377 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6379 30. Changed the default configuration ACL to relax the local part checking rule
6380 for addresses that are not in any local domains. For these addresses,
6381 slashes and pipe symbols are allowed within local parts, but the sequence
6382 /../ is explicitly forbidden.
6384 31. SPA server authentication was not clearing the challenge buffer before
6387 32. log_message in a "warn" ACL statement was writing to the reject log as
6388 well as to the main log, which contradicts the documentation and doesn't
6389 seem right (because no rejection is happening). So I have stopped it.
6391 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6392 However, I am unable to do any testing of this.
6394 34. Fixed an infelicity in the appendfile transport. When checking directories
6395 for a mailbox, to see if any needed to be created, it was accidentally
6396 using path names with one or more superfluous leading slashes; tracing
6397 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6399 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6400 contents are added to the log line that is written for every discarded
6401 recipient. (Previously a log_message setting was ignored.)
6403 36. The ${quote: operator now quotes the string if it is empty.
6405 37. The install script runs exim in order to find its version number. If for
6406 some reason other than non-existence or emptiness, which it checks, it
6407 could not run './exim', it was installing it with an empty version number,
6408 i.e. as "exim-". This error state is now caught, and the installation is
6411 38. An argument was missing from the function that creates an error message
6412 when Exim fails to connect to the socket for saslauthd authentication.
6413 This could cause Exim to crash, or give a corrupted message.
6415 39. Added isip, isip4, and isip6 to ${if conditions.
6417 40. The ACL variables $acl_xx are now saved with the message, and can be
6418 accessed later in routers, transports, and filters.
6420 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6421 strings in the file are not string-expanded.
6423 42. If a MAIL command specified a SIZE value that was too large to fit into an
6424 int variable, the check against message_size_limit failed. Such values are
6425 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6426 day this will have to be increased, but I don't think I want to be around
6427 when emails are that large.
6434 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6435 iconv() is not standard in FreeBSD.
6437 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6438 IPv6 enabled. The observed symptom was a segmentation fault on return from
6439 the function os_common_find_running_interfaces() in src/os.c.
6441 3. In the check_special_case() function in daemon.c I had used "errno" as an
6442 argument name, which causes warnings on some systems. This was basically a
6443 typo, since it was named "eno" in the comments!
6445 4. The code that waits for the clock to tick (at a resolution of some fraction
6446 of a second) so as to ensure message-id uniqueness was always waiting for
6447 at least one whole tick, when it could have waited for less. [This is
6448 almost certainly not relevant at current processor speeds, where it is
6449 unlikely to ever wait at all. But we try to future-proof.]
6451 5. The function that sleeps for a time interval that includes fractions of a
6452 second contained a race. It did not block SIGALRM between setting the
6453 timer, and suspending (a couple of lines later). If the interval was short
6454 and the sigsuspend() was delayed until after it had expired, the suspension
6455 never ended. On busy systems this could lead to processes getting stuck for
6458 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6459 process, before it forks any delivery processes. The open lookup caching
6460 mechanism meant that the open file or database connection was passed into
6461 the delivery process. The problem was that delivery processes always tidy
6462 up cached lookup data. This could cause a problem for the next delivery
6463 process started by the queue runner, because the external queue runner
6464 process does not know about the closure. So the next delivery process
6465 still has data in the lookup cache. In the case of a file lookup, there was
6466 no problem because closing a file descriptor in a subprocess doesn't affect
6467 the parent. However, if the lookup was caching a connection to a database,
6468 the connection was closed, and the second delivery process was likely to
6469 see errors such as "PGSQL: query failed: server closed the connection
6470 unexpectedly". The problem has been fixed by closing all cached lookups
6471 in a queue runner before running a delivery process.
6473 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6474 seem to have the "const" qualifier which it has on other OS. I've
6477 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6478 *of the same type* (client or server) with the same public name that an
6479 error should be diagnosed.
6481 9. When Exim looked up a host name for an IP address, but failed to find the
6482 original IP address when looking up the host name (a safety check), it
6483 output the message "<ip address> does not match any IP for NULL", which was
6484 confusing, to say the least. The bug was that the host name should have
6485 appeared instead of "NULL".
6487 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6488 user that is built into the binary, and the -C or -D options is used, root
6489 privilege is dropped before the configuration file is read. In addition,
6490 logging is switched to stderr instead of the normal log files. If the
6491 configuration then re-defines the Exim user, the unprivileged environment
6492 is probably not what is expected, so Exim logs a panic warning message (but
6495 However, if deliver_drop_privilege is set, the unprivileged state may well
6496 be exactly what is intended, so the warning has been cut out in that case,
6497 and Exim is allowed to try to write to its normal log files.
6503 1. smtp_return_error_details was not giving details for temporary sender
6504 or receiver verification errors.
6506 2. Diagnose a configuration error if two authenticators have the same public
6509 3. Exim used not to create the message log file for a message until the first
6510 delivery attempt. This could be confusing when incoming messages were held
6511 for policy or load reasons. The message log file is now created at the time
6512 the message is received, and an initial "Received" line is written to it.
6514 4. The automatically generated man page for command line options had a minor
6515 bug that caused no ill effects; however, a more serious problem was that
6516 the procedure for building the man page automatically didn't always
6517 operate. Consequently, release 4.20 contains an out-of-date version. This
6518 shouldn't happen again.
6520 5. When building Exim with embedded Perl support, the script that builds the
6521 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6522 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6524 6. The freeze_tell option was not being used for messages that were frozen on
6525 arrival, either by an ACL or by local_scan().
6527 7. Added the smtp_incomplete_transaction log selector.
6529 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6530 was accepting AUTH without a new EHLO.
6532 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6533 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6536 10. Logging of TCP/IP connections (when configured) now happens in the main
6537 daemon process instead of the child process, so that the TCP/IP connection
6538 count is more accurate (but it can never be perfect).
6540 11. The use of "drop" in a nested ACL was not being handled correctly in the
6541 outer ACL. Now, if condition failure induced by the nested "drop" causes
6542 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6543 or "require"), the connection is dropped.
6545 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6546 that yield "discard" can now be used with an "accept" or a "discard" verb,
6547 but an error is generated for any others (because I can't see a useful way
6548 to define what should happen).
6550 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6551 are now processed in the same way as lines in the Exim configuration file.
6552 In particular, continuation lines are supported.
6554 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6556 15. Added -ti meaning -t -i.
6558 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6559 domains, and warn by logging if others are found.
6561 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6562 returned from GIFCONF when Exim is trying to find the list of interfaces on
6563 a host. The code in os.c has been modified to copy each ifreq to an aligned
6564 structure in all cases.
6566 Also, in some cases, the returned ifreq's were being copied to a 'struct
6567 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6568 means the last couple of bytes of an IPv6 address could be chopped if the
6569 ifreq contained only a normal sockaddr (14 bytes storage).
6571 18. Named domain lists were not supported in the hosts_treat_as_local option.
6572 An entry such as +xxxx was not recognized, and was treated as a literal
6575 19. Ensure that header lines added by a DATA ACL are included in the reject log
6576 if the ACL subsequently rejects the message.
6578 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6579 MD5 (which is deprecated).
6581 21. When testing a filter file using -bf, Exim was writing a message when it
6582 took the sender from a "From " line in the message, but it was not doing so
6583 when it took $return_path from a Return-Path: header line. It now does.
6585 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6586 with a valid header line field name (a series of printing characters
6587 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6589 23. Changed "disc" in the source to "disk" to conform to the documentation and
6590 the book and for uniformity.
6592 24. Ignore Sendmail's -Ooption=value command line item.
6594 25. When execve() failed while trying to run a command in a pipe transport,
6595 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6596 could be confused with a return value of 69 from the command itself. This
6597 has been changed to 127, the value the shell returns if it is asked to run
6598 a non-existent command. The wording for the related log line suggests a
6599 non-existent command as the problem.
6601 26. If received_header_text expands to an empty string, do not add a Received:
6602 header line to the message. (Well, it adds a token one on the spool, but
6603 marks it "old" so that it doesn't get used or transmitted.)
6605 27. Installed eximstats 1.28 (addition of -nt option).
6607 28. There was no check for failure on the call to getsockname() in the daemon
6608 code. This can fail if there is a shortage of resources on the system, with
6609 ENOMEM, for example. A temporary error is now given on failure.
6611 29. Contrary to the C standard, it seems that in some environments, the
6612 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6613 program. Exim now does this explicitly; it affects the formatting of
6614 timestamps using strftime().
6616 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6617 complaints. I've now initialized all the variables, to avoid this.
6619 32. Header lines added by a system filter were not being "seen" during
6620 transport-time rewrites.
6622 33. The info_callback() function passed to OpenSSL is set up with type void
6623 (*)(SSL *, int, int), as described somewhere. However, when calling the
6624 function (actually a macro) that sets it up, the type void(*)() is
6625 expected. I've put in a cast to prevent warnings from picky compilers.
6627 34. If a DNS black list lookup found a CNAME record, but there were no A
6628 records associated with the domain it pointed at, Exim crashed.
6630 35. If a DNS black list lookup returned more than one A record, Exim ignored
6631 all but the first. It now scans all returned addresses if a particular IP
6632 value is being sought. In this situation, the contents of the
6633 $dnslist_value variable are a list of all the addresses, separated by a
6636 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6637 to accept a host name and all its aliases if the forward lookup for any of
6638 them yielded the IP address of the incoming connection. Now it accepts only
6639 those names whose forward lookup yields the correct IP address. Any other
6640 names are discarded. This closes a loophole whereby a rogue DNS
6641 administrator could create reverse DNS records to break through a
6642 wildcarded host restriction in an ACL.
6644 37. If a user filter or a system filter that ran in a subprocess used any of
6645 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6646 the wrong values were passed to the pipe command ($thisaddress had the
6647 value of $0, $0 had the value of $1, etc). This bug was introduced by
6648 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6650 38. Improved the line breaking for long SMTP error messages from ACLs.
6651 Previously, if there was no break point between 40 and 75 characters, Exim
6652 left the rest of the message alone. Two changes have been made: (a) I've
6653 reduced the minimum length to 35 characters; (b) if it can't find a break
6654 point between 35 and 75 characters, it looks ahead and uses the first one
6655 that it finds. This may give the occasional overlong line, but at least the
6656 remaining text gets split now.
6658 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6659 file descriptors might be low, and that setting 1000 would always raise it.
6660 It turns out that in some environments, the limit is already over 1000 and
6661 that lowering it causes trouble. So now Exim takes care not to decrease it.
6663 40. When delivering a message, the value of $return_path is set to $sender_
6664 address at the start of routing (routers may change the value). By an
6665 oversight, this default was not being set up when an address was tested by
6666 -bt or -bv, which affected the outcome if any router or filter referred to
6669 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6670 log only when "message" or "log_message" are set. However, if one of the
6671 conditions was an address verification, or a call to a nested ACL, the
6672 messages generated by the underlying test were being passed through. This
6673 no longer happens. The underlying message is available in $acl_verify_
6674 message for both "message" and "log_message" expansions, so it can be
6675 passed through if needed.
6677 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6678 new expansion $bh_ to give the encoded byte string without charset
6679 translation. Translation happens only if iconv() is available; HAVE_ICONV
6680 indicates this at build time. HEADERS_CHARSET gives the charset to
6681 translate to; headers_charset can change it in the configuration, and
6682 "headers charset" can change it in an individual filter file.
6684 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6685 that creates RFC 2047 encoded "words" labels them as that charset instead
6686 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6687 expansion operator; (ii) when Exim creates a From: line for a local
6688 message; (iii) when a header line is rewritten to include a "phrase" part.
6690 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6691 buggy, causing it to skip the first lines of messages whose message ID
6692 ended in 'D'. This would not have bitten before Exim release 4.14, because
6693 message IDs were unlikely to end in 'D' before then. The effect was to have
6694 incorrect size information for certain domains.
6696 45. #include "config.h" was missing at the start of the crypt16.c module. This
6697 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6700 46. If there was a timeout during a "random" callout check, Exim treated it as
6701 a failure of the random address, and carried on sending RSET and the real
6702 address. If the delay was just some slowness somewhere, the response to the
6703 original RCPT would be taken as a response to RSET and so on, causing
6704 mayhem of various kinds.
6706 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6707 when I implemented it. It didn't allow for the fact that some option values
6708 may legitimately be negative (e.g. size_addition), and it didn't even do
6709 the right test for positive values.
6711 48. Domain names in DNS records are case-independent. Exim always looks them up
6712 in lower case. Some resolvers return domain names in exactly the case they
6713 appear in the zone file, that is, they may contain uppercase letters. Not
6714 all resolvers do this - some return always lower case. Exim was treating a
6715 change of case by a resolver as a change of domain, similar to a widening
6716 of a domain abbreviation. This triggered its re-routing code and so it was
6717 trying to route what was effectively the same domain again. This normally
6718 caused routing to fail (because the router wouldn't handle the domain
6719 twice). Now Exim checks for this case specially, and just changes the
6720 casing of the domain that it ultimately uses when it transmits the message
6723 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6726 50. If a filter generated a file delivery with a non-absolute name (possible if
6727 no home directory exists for the router), the forbid_file option was not
6730 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6731 the existing equality matching.
6733 52. Exim was using ints instead of ino_t variables in some places where it was
6734 dealing with inode numbers.
6736 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6737 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6738 called TMPDIR, and if it finds it is different, it changes its value.
6740 54. The smtp_printf() function is now made available to local_scan() so
6741 additional output lines can be written before returning. There is also an
6742 smtp_fflush() function to enable the detection of a dropped connection.
6743 The variables smtp_input and smtp_batched_input are exported to
6746 55. Changed the default runtime configuration: the message "Unknown user"
6747 has been removed from the ACL, and instead placed on the localuser router,
6748 using the cannot_route_message feature. This means that any verification
6749 failures that generate their own messages won't get overridden. Similarly,
6750 the "Unrouteable address" message that was in the ACL for unverifiable
6751 relay addresses has also been removed.
6753 56. Added hosts_avoid_esmtp to the smtp transport.
6755 57. The exicyclog script was not checking for the esoteric option
6756 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6757 will work only if exicyclog is run under the appropriate euid.
6759 58. Following a discussion on the list, the rules by which Exim recognises line
6760 endings on incoming messages have been changed. The -dropcr and drop_cr
6761 options are now no-ops, retained only for backwards compatibility. The
6762 following line terminators are recognized: LF CRLF CR. However, special
6763 processing applies to CR:
6765 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6766 nor a local message in the state where . is a terminator.
6768 (ii) If a bare CR is encountered in a header line, an extra space is added
6769 after the line terminator so as not to end the header. The reasoning
6770 behind this is that bare CRs in header lines are most likely either
6771 to be mistakes, or people trying to play silly games.
6773 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6774 was being incorrectly given as 18 bytes larger than it should have been.
6775 This is a VOB (very old bug).
6777 60. This may never have affected anything current, but just in case it has:
6778 When the local host is found other than at the start of a list of hosts,
6779 the local host, those with the same MX, and any that follow, are discarded.
6780 When the list in question was part of a longer list of hosts, the following
6781 hosts (not currently being processed) were also being discarded. This no
6782 longer happens. I'm not sure if this situation could ever has previously
6785 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6788 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6789 header. This is something that is recommended in a new Internet Draft, and
6790 is something that is documented as being done by Sendmail. There are two
6791 possible values. For messages generated by the autoreply transport, Exim
6794 Auto-Submitted: auto-replied
6796 whereas for all other generated messages (e.g. bounces) it adds
6798 Auto-Submitted: auto-generated
6800 63. The "personal" condition in filters now includes a test for the
6801 Auto-Submitted: header. If it contains the string "auto-" the message it
6802 not considered personal.
6804 64. Added rcpt_include_affixes as a generic transport option.
6806 65. Added queue_only_override (default true).
6808 66. Added the syslog_duplication option.
6810 67. If what should have been the first header line of a message consisted of
6811 a space followed by a colon, Exim was mis-interpreting it as a header line.
6812 It isn't of course - it is syntactically invalid and should therefore be
6813 treated as the start of the message body. The misbehaviour could have
6814 caused a number of strange effects, including loss of data in subsequent
6815 header lines, and spool format errors.
6817 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6818 client host had authenticated. This control can now be exercised by an ACL
6819 for more flexibility.
6821 69. By default, callouts do not happen when testing with -bh. There is now a
6822 variant, -bhc, which does actually run the callout code, including
6823 consulting and updating the callout cache.
6825 70. Added support for saslauthd authentication, courtesy of Alexander
6828 71. If statvfs() failed on the spool or log directories while checking their
6829 size for availability, Exim confusingly gave the error "space shortage".
6830 Furthermore, in debugging mode it crashed with a floating point exception.
6831 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6832 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6833 serious problem, Exim now writes to the main and panic logs when this
6834 happens, with details of the failure. It then refuses to accept the
6835 incoming message, giving the message "spool directory problem" or "log
6836 directory problem" with a 421 code for SMTP messages.
6838 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6839 0, 1, and 2 exist, because some OS complain for execs without them (see
6840 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6841 descriptors. However, the code omitted to check that the open succeeded,
6842 causing mysterious errors if for some reason the permissions on /dev/null
6843 got screwed. Now Exim writes a message to the main and panic logs, and
6844 bombs out if it can't open /dev/null.
6846 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6847 interact so that it is all more flexible. It is supposed to remain
6848 backwards compatible. Also added extra_local_interfaces.
6850 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6851 to bomb out with an assertion failure - to the client this appears as a
6852 connection drop. This problem occurs in the part of the code that was taken
6853 from the Samba project. Fortunately, the assertion is in a very simple
6854 function, so I have fixed this by reproducing the function inline in the
6855 one place where it is called, and arranging for authentication to fail
6856 instead of killing the process with assert().
6858 75. The SPA client code was not working when the server requested OEM rather
6859 than Unicode encoding.
6861 76. Added code to make require_files with a specific uid setting more usable in
6862 the case where statting the file as root fails - usually a non-root-mounted
6863 NFS file system. When this happens and the failure is EACCES, Exim now
6864 forks a subprocess and does the per-uid checking as the relevant uid.
6866 77. Added process_log_path.
6868 78. If log_file_path was not explicitly set, a setting of check_log_space or
6869 check_log_inodes was ignored.
6871 79. If a space check for the spool or log partitions fails, the incident is now
6872 logged. Of course, in the latter case the data may get lost...
6874 80. Added the %p formatting code to string_format() so that it can be used to
6875 print addresses in debug_print(). Adjusted all the address printing in the
6876 debugging in store.c to use %p rather than %d.
6878 81. There was a concern that a line of code in smtp_in.c could overflow a
6879 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6880 initially expressed, the concern was not well-founded, because trailing
6881 spaces are removed early. However, if the trailing spaces were followed by
6882 a NULL, they did not get removed, so the overflow was possible. Two fixes
6885 (a) I re-wrote the offending code in a cleaner fashion.
6886 (b) If an incoming SMTP command contains a NULL character, it is rejected
6889 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6890 runs initgroups(), so that if the Exim user is in any additional groups,
6891 they will be used during message reception.
6897 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob