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
130 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
131 than -2003 did; needs libidn2 in addition to libidn.
133 JH/02 The path option on a pipe transport is now expanded before use.
135 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
136 Patch provided by "Björn", documentation fix added too.
138 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
139 missing a wire-to-host endian conversion.
141 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
142 close after a BDAT command line could be taken as a following command,
143 giving a synch failure. Fix by only checking for synch immediately
144 before acknowledging the chunk.
146 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
147 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
148 macro. Patches provided by Josh Soref.
150 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
151 Previously we did not; the RFC seems ambiguous and VRFY is not listed
152 by IANA as a service extension. However, John Klensin suggests that we
155 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
156 the dkim code may be unix-mode line endings rather than smtp wire-format
157 CRLF, so prepend a CR to any bare LF.
159 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
160 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
162 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
163 the first were themselves being wrongly included in the feed into dkim
164 processing; with most chunk sizes in use this resulted in an incorrect
165 body hash calculated value.
167 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
168 DKIM signature block, for verification. Although advised against by
169 standards it is specifically not ruled illegal.
171 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
173 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
174 missing a body hash (the bh= tag).
176 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
177 It seems that HAProxy sends the Proxy Protocol information in clear and
178 only then does a TLS startup, so do the same.
180 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
181 TCP connections (such as for Spamd) unless the daemon successfully set
182 Fast Open mode on its listening sockets. This fixes breakage seen on
183 too-old kernels or those not configured for Fast Open, at the cost of
184 requiring both directions being enabled for TFO, and TFO never being used
185 by non-daemon-related Exim processes.
187 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
188 endings, at least on the first header line. Try to canonify any that get
189 past that check, despite the cost.
191 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
192 now limited to an arbitrary five deep, while parsing addresses with the
193 strip_excess_angle_brackets option enabled.
195 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
196 instead leave the unprompted TLS handshake in socket buffer for the
197 TLS library to consume.
199 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
201 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
203 JH/16 Drop variables when they go out of scope. Memory management drops a whole
204 region in one operation, for speed, and this leaves assigned pointers
205 dangling. Add checks run only under the testsuite which checks all
206 variables at a store-reset and panics on a dangling pointer; add code
207 explicitly nulling out all the variables discovered. Fixes one known
208 bug: a transport crash, where a dangling pointer for $sending_ip_address
209 originally assigned in a verify callout, is re-used.
211 PP/06 Drop '.' from @INC in various Perl scripts.
213 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
215 PP/08 Reduce a number of compilation warnings under clang; building with
216 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
217 should be warning-free.
219 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
221 HS/01 Fix portability problems introduced by PP/08 for platforms where
222 realloc(NULL) is not equivalent to malloc() [SunOS et al].
224 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
225 chunk. This allows us to accept broken chunked messages. We need a more
226 general solution here.
228 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
229 already-broken messages in the queue.
231 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
233 JH/19 Fix reference counting bug in routing-generated-address tracking.
239 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
240 supports it and a size is available (ie. the sending peer gave us one).
242 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
243 years of being deprecated). The replacements are the ACLs
244 acl_smtp_mime and acl_not_smtp_mime.
246 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
247 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
248 or even an in-clear connection were permitted. Now, if the host lookup
249 was dnssec and dane was requested then the host is only used if the
250 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
251 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
252 if one fails this test.
253 This means that a poorly-configured remote DNS will make it incommunicado;
254 but it protects against a DNS-interception attack on it.
256 JH/04 Bug 1810: make continued-use of an open smtp transport connection
257 non-noisy when a race steals the message being considered.
259 JH/05 If main configuration option tls_certificate is unset, generate a
260 self-signed certificate for inbound TLS connections.
262 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
263 in rewrites and routers.
265 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
266 and logged a warning sing 4.83; now they are a configuration file error.
268 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
269 (lacking @domain). Apply the same qualification processing as RCPT.
271 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
273 JH/10 Support ${sha256:} applied to a string (as well as the previous
276 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
277 a cutthrough deliver is pending, as we always want to make a connection.
278 This also avoids re-routing the message when later placing the cutthrough
279 connection after a verify cache hit.
280 Do not update it with the verify result either.
282 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
283 when routing results in more than one destination address.
285 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
286 signing (which inhibits the cutthrough capability). Previously only
287 the presence of an option was tested; now an expansion evaluating as
288 empty is permissible (obviously it should depend only on data available
289 when the cutthrough connection is made).
291 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
292 the relevant preceding SMTP command did not note the pipelining mode.
294 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
295 Previously they were not counted.
297 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
298 as one having no matching records. Previously we deferred the message
299 that needed the lookup.
301 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
302 distinguished as "(=".
304 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
305 for missing MX records. Previously it only worked for missing A records.
307 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
309 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
310 after the data-go-ahead and data-ack. Patch from Jason Betts.
312 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
313 even for a "none" policy. Patch from Tony Meyer.
315 JH/22 Fix continued use of a connection for further deliveries. If a port was
316 specified by a router, it must also match for the delivery to be
319 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
320 When one of the recipients of a message was unsuitable for the connection
321 (has no matching addresses), we lost track of needing to mark it
322 deferred. As a result mail would be lost.
324 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
326 JH/25 Decoding ACL controls is now done using a binary search; the source code
327 takes up less space and should be simpler to maintain. Merge the ACL
328 condition decode tables also, with similar effect.
330 JH/26 Fix problem with one_time used on a redirect router which returned the
331 parent address unchanged. A retry would see the parent address marked as
332 delivered, so not attempt the (identical) child. As a result mail would
335 JH/27 Fix a possible security hole, wherein a process operating with the Exim
336 UID can gain a root shell. Credit to http://www.halfdog.net/ for
337 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
340 JH/28 Enable {spool,log} filesystem space and inode checks as default.
341 Main config options check_{log,spool}_{inodes,space} are now
342 100 inodes, 10MB unless set otherwise in the configuration.
344 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
345 Previously it only applied to the main-section connection policy
348 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
350 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
351 by me. Added RFC7919 DH primes as an alternative.
353 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
354 are not in the system include path.
356 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
357 GnuTLS, when a session startup failed (eg because the client disconnected)
358 Exim did stdio operations after fclose. This was exposed by a recent
359 change which nulled out the file handle after the fclose.
361 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
362 signed directly by the cert-signing cert, rather than an intermediate
363 OCSP-signing cert. This is the model used by LetsEncrypt.
365 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
367 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
368 an incoming connection.
370 HS/02 Bug 1802: Do not half-close the connection after sending a request
373 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
374 fallback to "prime256v1".
376 JH/34 SECURITY: Use proper copy of DATA command in error message.
377 Could leak key material. Remotely exploitable. CVE-2016-9963.
383 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
384 and 3.4.4 - once the server is enabled to respond to an OCSP request
385 it does even when not requested, resulting in a stapling non-aware
386 client dropping the TLS connection.
388 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
389 support variable-length bit vectors. No functional change.
391 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
392 The I= interface field on outgoing lines is now after the H= remote
393 host field, same as incoming lines. There is a separate
394 outgoing_interface log selector which allows you to disable the
397 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
398 If not running log_selector +smtp_connection the mainlog would be held
399 open indefinitely after a "too many connections" event, including to a
400 deleted file after a log rotate. Leave the per net connection logging
401 leaving it open for efficiency as that will be quickly detected by the
402 check on the next write.
404 HS/01 Bug 1671: Fix post transport crash.
405 Processing the wait-<transport> messages could crash the delivery
406 process if the message IDs didn't exist for some reason. When
407 using 'split_spool_directory=yes' the construction of the spool
408 file name failed already, exposing the same netto behaviour.
410 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
411 mime_regex ACL conditions.
413 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
414 to DSN fail messages (bounces): remote IP, remote greeting, remote response
415 to HELO, local diagnostic string.
417 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
418 log line to debug. Even when configured with a tls authenticator many
419 client connections are expected to not authenticate in this way, so
420 an authenticate fail is not an error.
422 HS/02 Add the Exim version string to the process info. This way exiwhat
423 gives some more detail about the running daemon.
425 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
426 matter for fast-change records such as DNSBLs.
428 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
429 retry record, even if constant. There may be multiple transports with
430 different interface settings and the retry behaviour needs to be kept
433 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
435 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
437 JH/10 Bug 840: fix log_defer_output option of pipe transport
439 JH/11 Bug 830: use same host for all RCPTS of a message, even under
440 hosts_randomize. This matters a lot when combined with mua_wrapper.
442 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
443 ${quote_pgsql:<string>} operator.
445 JH/13 Bug 1708: avoid misaligned access in cached lookup.
447 JH/14 Change header file name for freeradius-client. Relevant if compiling
448 with Radius support; from the Gentoo tree and checked under Fedora.
450 JH/15 Bug 1712: Introduce $prdr_requested flag variable
452 JH/16 Bug 1714: Permit an empty string as expansion result for transport
453 option transport_filter, meaning no filtering.
455 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
457 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
458 defaults to "*" (all hosts). The variable is now available when not built
459 with TLS, default unset, mainly to enable keeping the testsuite sane.
460 If a server certificate is not supplied (via tls_certificate) an error is
461 logged, and clients will find TLS connections fail on startup. Presumably
462 they will retry in-clear.
463 Packagers of Exim are strongly encouraged to create a server certificate
464 at installation time.
466 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
467 with the $config_file variable.
469 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
470 in transport context, after the attempt, and per-recipient. The latter type
471 is per host attempted. The event data is the error message, and the errno
472 information encodes the lookup type (A vs. MX) used for the (first) host,
473 and the trailing two digits of the smtp 4xx response.
475 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
476 to write to mainlog (or rejectlog, paniclog) in the window between file
477 creation and permissions/ownership being changed. Particularly affects
478 installations where exicyclog is run as root, rather than exim user;
479 result is that the running daemon panics and dies.
481 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
483 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
484 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
485 "pri" and "weight". Note that the previous implicit priority given by the
486 list order is no longer honoured.
488 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
491 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
492 by defining SUPPORT_SOCKS.
494 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
495 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
496 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
497 variables are renamed to proxy_{local,external}_{address,port}.
499 JH/27 Move Internationalisation support from Experimental to mainline, enabled
500 for a build by defining SUPPORT_I18N
502 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
503 of the query string, and make ${quote_redis:} do that quoting.
505 JH/29 Move Events support from Experimental to mainline, enabled by default
506 and removable for a build by defining DISABLE_EVENT.
508 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
510 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
511 cached by the daemon.
513 JH/32 Move Redis support from Experimental to mainline, enabled for a build
514 by defining LOOKUP_REDIS. The libhiredis library is required.
516 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
517 keys are given for lookup.
519 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
520 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
521 only supported when built with TLS support. The PolarSSL SHA routines
522 are still used when the TLS library is too old for convenient support.
524 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
525 openssl_options), for security. OpenSSL forces this from version 1.1.0
526 server-side so match that on older versions.
528 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
529 allocation for $value could be released as the expansion processing
530 concluded, but leaving the global pointer active for it.
532 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
533 and to use the domains and local_parts ACL conditions.
535 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
536 incorrectly not doubled on cutthrough transmission, hence seen as a
537 body-termination at the receiving system - resulting in truncated mails.
538 Commonly the sender saw a TCP-level error, and retransmitted the message
539 via the normal store-and-forward channel. This could result in duplicates
540 received - but deduplicating mailstores were liable to retain only the
541 initial truncated version.
543 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
545 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
547 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
548 we're in there, support oversigning also; bug 1309.
550 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
552 HS/04 Add support for keep_environment and add_environment options.
554 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
555 either intentional arithmetic overflow during PRNG, or testing config-
558 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
559 delivery resulted in actual delivery. Cancel cutthrough before DATA
562 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
563 rejects a recipient: pass the reject to the originator.
565 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
566 Many were false-positives and ignorable, but it's worth fixing the
569 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
570 for the new environment-manipulation done at startup. Move the routines
571 from being local to tls.c to being global via the os.c file.
573 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
574 an extract embedded as result-arg for a map, the first arg for extract
575 is unavailable so we cannot tell if this is a numbered or keyed
576 extraction. Accept either.
582 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
585 JH/02 The smtp transport option "multi_domain" is now expanded.
587 JH/03 The smtp transport now requests PRDR by default, if the server offers
590 JH/04 Certificate name checking on server certificates, when exim is a client,
591 is now done by default. The transport option tls_verify_cert_hostnames
592 can be used to disable this per-host. The build option
593 EXPERIMENTAL_CERTNAMES is withdrawn.
595 JH/05 The value of the tls_verify_certificates smtp transport and main options
596 default to the word "system" to access the system default CA bundle.
597 For GnuTLS, only version 3.0.20 or later.
599 JH/06 Verification of the server certificate for a TLS connection is now tried
600 (but not required) by default. The verification status is now logged by
601 default, for both outbound TLS and client-certificate supplying inbound
604 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
607 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
608 Status Notification (bounce) messages are now MIME format per RFC 3464.
609 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
610 under the control of the dsn_advertise_hosts option, and routers may
611 have a dsn_lasthop option.
613 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
614 default, modifiable by a malware= option. The list separator for
615 the options can now be changed in the usual way. Bug 68.
617 JH/10 The smtp_receive_timeout main option is now expanded before use.
619 JH/11 The incoming_interface log option now also enables logging of the
620 local interface on delivery outgoing connections.
622 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
623 if the interface and destination host and port all match.
625 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
628 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
629 Patch from Andrew Lewis.
631 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
632 now supports optional time-restrictions, weighting, and priority
633 modifiers per server. Patch originally by <rommer@active.by>.
635 JH/16 The spamd_address main option now supports a mixed list of local
636 and remote servers. Remote servers can be IPv6 addresses, and
637 specify a port-range.
639 JH/17 Bug 68: The spamd_address main option now supports an optional
640 timeout value per server.
642 JH/18 Bug 1581: Router and transport options headers_add/remove can
643 now have the list separator specified.
645 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
648 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
651 JH/21 Support for the A6 type of dns record is withdrawn.
653 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
654 rather than the verbs used.
656 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
657 from 255 to 1024 chars.
659 JH/24 Verification callouts now attempt to use TLS by default.
661 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
662 are generic router options now. The defaults didn't change.
664 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
665 Original patch from Alexander Shikoff, worked over by JH.
667 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
668 files if ZCAT_COMMAND is not executable.
670 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
672 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
674 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
675 Normally benign, it bites when the pair was led to by a CNAME;
676 modern usage is to not canonicalize the domain to a CNAME target
677 (and we were inconsistent anyway for A-only vs AAAA+A).
679 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
681 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
682 when evaluating $sender_host_dnssec.
684 JH/31 Check the HELO verification lookup for DNSSEC, adding new
685 $sender_helo_dnssec variable.
687 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
689 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
691 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
693 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
694 documented as working, but never had. Support all but $spam_report.
696 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
697 added for tls authenticator.
699 HS/03 Add perl_taintmode main config option
705 TL/01 When running the test suite, the README says that variables such as
706 no_msglog_check are global and can be placed anywhere in a specific
707 test's script, however it was observed that placement needed to be near
708 the beginning for it to behave that way. Changed the runtest perl
709 script to read through the entire script once to detect and set these
710 variables, reset to the beginning of the script, and then run through
711 the script parsing/test process like normal.
713 TL/02 The BSD's have an arc4random API. One of the functions to induce
714 adding randomness was arc4random_stir(), but it has been removed in
715 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
716 function when detected.
718 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
719 cause callback expansion.
721 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
722 syntax errors in an expansion can be treated as a string instead of
723 logging or causing an error, due to the internal use of bool_lax
724 instead of bool when processing it.
726 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
727 server certificates when making smtp deliveries.
729 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
731 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
733 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
735 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
736 Merged patch from Sebastian Wiedenroth.
738 JH/05 Fix results-pipe from transport process. Several recipients, combined
739 with certificate use, exposed issues where response data items split
740 over buffer boundaries were not parsed properly. This eventually
741 resulted in duplicates being sent. This issue only became common enough
742 to notice due to the introduction of connection certificate information,
743 the item size being so much larger. Found and fixed by Wolfgang Breyha.
745 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
746 size buffer was used, resulting in syntax errors when an expansion
749 JH/07 Add support for directories of certificates when compiled with a GnuTLS
750 version 3.3.6 or later.
752 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
753 is EXPERIMENTAL_EVENT, the main-configuration and transport options
754 both become "event_action", the variables become $event_name, $event_data
755 and $event_defer_errno. There is a new variable $verify_mode, usable in
756 routers, transports and related events. The tls:cert event is now also
757 raised for inbound connections, if the main configuration event_action
760 TL/06 In test suite, disable OCSP for old versions of openssl which contained
761 early OCSP support, but no stapling (appears to be less than 1.0.0).
763 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
764 server certificate names available under the smtp transport option
765 "tls_verify_cert_hostname" now do not permit multi-component wildcard
768 JH/10 Time-related extraction expansions from certificates now use the main
769 option "timezone" setting for output formatting, and are consistent
770 between OpenSSL and GnuTLS compilations. Bug 1541.
772 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
773 encoded parameter in the incoming message. Bug 1558.
775 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
776 include certificate info, eximon was claiming there were spoolfile
779 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
781 JH/14 Log delivery-related information more consistently, using the sequence
782 "H=<name> [<ip>]" wherever possible.
784 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
785 are problematic for Debian distribution, omit them from the release
788 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
790 JH/16 Fix string representation of time values on 64bit time_t architectures.
793 JH/17 Fix a null-indirection in certextract expansions when a nondefault
794 output list separator was used.
799 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
800 checkers that were complaining about end of non-void function with no
803 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
804 This was a regression introduced in 4.83 by another bugfix.
806 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
808 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
809 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
815 TF/01 Correctly close the server side of TLS when forking for delivery.
817 When a message was received over SMTP with TLS, Exim failed to clear up
818 the incoming connection properly after forking off the child process to
819 deliver the message. In some situations the subsequent outgoing
820 delivery connection happened to have the same fd number as the incoming
821 connection previously had. Exim would try to use TLS and fail, logging
822 a "Bad file descriptor" error.
824 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
825 utilities have not been installed.
827 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
828 temporary space as the ACL may create new global variables.
830 TL/01 LDAP support uses per connection or global context settings, depending
831 upon the detected version of the libraries at build time.
833 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
834 to extract and use the src ip:port in logging and expansions as if it
835 were a direct connection from the outside internet. PPv2 support was
836 updated based on HAProxy spec change in May 2014.
838 JH/02 Add ${listextract {number}{list}{success}{fail}}.
840 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
841 Properly escape header and check for NULL return.
843 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
846 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
848 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
849 characters in header names, implemented as a verify condition.
850 Contributed by Michael Fischer v. Mollard.
852 TL/05 Rename SPF condition results err_perm and err_temp to standardized
853 results permerror and temperror. Previous values are deprecated but
854 still accepted. In a future release, err_perm and err_temp will be
855 completely removed, which will be a backward incompatibility if the
856 ACL tests for either of these two old results. Patch contributed by
857 user bes-internal on the mailing list.
859 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
861 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
862 selectors, in both main and reject logs.
864 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
867 JH/07 Add malware type "sock" for talking to simple daemon.
869 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
871 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
872 routers/transports under cutthrough routing.
874 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
875 numbers. Touch up "bool" conditional to keep the same definition.
877 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
879 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
881 JH/12 Expand items in router/transport headers_add or headers_remove lists
882 individually rather than the list as a whole. Bug 1452.
884 Required for reasonable handling of multiple headers_ options when
885 they may be empty; requires that headers_remove items with embedded
886 colons must have them doubled (or the list-separator changed).
888 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
889 view the policy declared in the DMARC record. Currently, $dmarc_status
890 is a combined value of both the record presence and the result of the
893 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
895 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
896 dnslookup router and the smtp transport (applying to the forward
899 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
900 of ldap servers used for a specific lookup. Patch provided by Heiko
903 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
904 New variable $lookup_dnssec_authenticated for observability.
906 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
907 Patch submitted by Lars Timman.
909 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
911 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
912 Requires trusted mode and valid format message id, aborts otherwise.
913 Patch contributed by Heiko Schlichting.
915 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
916 certextract with support for various fields. Bug 1358.
918 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
919 is requested by default, modifiable by smtp transport option
922 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
923 operate on certificate variables to give certificate fingerprints
924 Also new ${sha256:cert_variable}.
926 JH/23 The PRDR feature is moved from being Experimental into the mainline.
928 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
929 Christian Aistleitner.
931 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
933 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
934 file. Patch from Wolfgang Breyha.
936 JH/25 Expand the coverage of the delivery $host and $host_address to
937 client authenticators run in verify callout. Bug 1476.
939 JH/26 Port service names are now accepted for tls_on_connect_ports, to
940 align with daemon_smtp_ports. Bug 72.
942 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
943 support and error reporting did not work properly.
945 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
946 and is readable. Patch from Andrew Colin Kissa.
948 TL/14 Enhance documentation of ${run expansion and how it parses the
949 commandline after expansion, particularly in the case when an
950 unquoted variable expansion results in an empty value.
952 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
954 PP/02 Fix internal collision of T_APL on systems which support RFC3123
955 by renaming away from it. Addresses GH issue 15, reported by
958 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
960 TL/15 SECURITY: prevent double expansion in math comparison functions
961 (can expand unsanitized data). Not remotely exploitable.
968 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
970 PP/02 Make -n do something, by making it not do something.
971 When combined with -bP, the name of an option is not output.
973 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
976 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
977 $sender_host_name and config options to manage this, and basic check
980 PP/05 DSCP support for outbound connections and control modifier for inbound.
982 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
983 (Only plugin which currently uses this is kerberos4, which nobody should
984 be using, but we should make it available and other future plugins might
985 conceivably use it, even though it would break NAT; stuff *should* be
986 using channel bindings instead).
988 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
989 name; added for Sendmail compatibility; requires admin caller.
990 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
991 just ignore it); requires trusted caller.
992 Also parse but ignore: -Ac -Am -X<logfile>
995 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
997 TL/02 Add +smtp_confirmation as a default logging option.
999 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1000 Patch by Magnus Holmgren from 2007-02-20.
1002 TL/04 Bugzilla 1281 - Spec typo.
1003 Bugzilla 1283 - Spec typo.
1004 Bugzilla 1290 - Spec grammar fixes.
1006 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1008 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1010 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1011 mailing list by by Dmitry Isaikin.
1013 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1015 JH/02 Support "G" suffix to numbers in ${if comparisons.
1017 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1019 NM/01 Bugzilla 1197 - Spec typo
1020 Bugzilla 1196 - Spec examples corrections
1022 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1024 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1025 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1028 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1029 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1031 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1032 "acl {{name}{arg}...}", and optional args on acl condition
1035 JH/05 Permit multiple router/transport headers_add/remove lines.
1037 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1039 JH/07 Avoid using a waiting database for a single-message-only transport.
1040 Performance patch from Paul Fisher. Bugzilla 1262.
1042 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1045 JH/09 Add $headers_added variable, with content from use of ACL modifier
1046 add_header (but not yet added to the message). Bugzilla 199.
1048 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1049 Pulled from Bugzilla 817 by Wolfgang Breyha.
1051 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1053 (nb: this is the same fix as in Exim 4.80.1)
1055 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1058 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1061 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1063 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1064 advertise SMTP AUTH mechanism to us, instead of a generic
1065 protocol violation error. Also, make Exim more robust to bad
1066 data from the Dovecot auth socket.
1068 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1070 When a queue runner is handling a message, Exim first routes the
1071 recipient addresses, during which it prunes them based on the retry
1072 hints database. After that it attempts to deliver the message to
1073 any remaining recipients. It then updates the hints database using
1076 So if a recipient address works intermittently, it can get repeatedly
1077 deferred at routing time. The retry hints record remains fresh so the
1078 address never reaches the final cutoff time.
1080 This is a fairly common occurrence when a user is bumping up against
1081 their storage quota. Exim had some logic in its local delivery code
1082 to deal with this. However it did not apply to per-recipient defers
1083 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1085 This change adds a proper retry rule check during routing so that the
1086 final cutoff time is checked against the message's age. We only do
1087 this check if there is an address retry record and there is not a
1088 domain retry record; this implies that previous attempts to handle
1089 the address had the retry_use_local_parts option turned on. We use
1090 this as an approximation for the destination being like a local
1091 delivery, as in LMTP.
1093 I suspect this new check makes the old local delivery cutoff check
1094 redundant, but I have not verified this so I left the code in place.
1096 TF/02 Correct gecos expansion when From: is a prefix of the username.
1098 Test 0254 submits a message to Exim with the header
1102 When I ran the test suite under the user fanf2, Exim expanded
1103 the header to contain my full name, whereas it should have added
1104 a Resent-Sender: header. It erroneously treats any prefix of the
1105 username as equal to the username.
1107 This change corrects that bug.
1109 GF/01 DCC debug and logging tidyup
1110 Error conditions log to paniclog rather than rejectlog.
1111 Debug lines prefixed by "DCC: " to remove any ambiguity.
1113 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1115 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1116 Bug spotted by Jeremy Harris; was flawed since initial commit.
1117 Would have resulted in OCSP responses post-SNI triggering an Exim
1118 NULL dereference and crash.
1120 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1122 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1123 Bug detection, analysis and fix by Samuel Thibault.
1124 Bugzilla 1331, Debian bug #698092.
1126 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1128 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1129 Server implementation by Todd Lyons, client by JH.
1130 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1131 config variable "prdr_enable" controls whether the server
1132 advertises the facility. If the client requests PRDR a new
1133 acl_data_smtp_prdr ACL is called once for each recipient, after
1134 the body content is received and before the acl_smtp_data ACL.
1135 The client is controlled by both of: a hosts_try_prdr option
1136 on the smtp transport, and the server advertisement.
1137 Default client logging of deliveries and rejections involving
1138 PRDR are flagged with the string "PRDR".
1140 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1141 fclose(). Diagnosis by Todd Lyons.
1143 PP/17 Update configure.default to handle IPv6 localhost better.
1144 Patch by Alain Williams (plus minor tweaks).
1147 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1148 This is now consistent with GnuTLS, and is now documented: the
1149 previous undocumented portable approach to treating the option as
1150 unset was to force an expansion failure. That still works, and
1151 an empty string is now equivalent.
1153 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1154 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1155 not performing validation itself.
1157 PP/20 Added force_command boolean option to pipe transport.
1158 Patch from Nick Koston, of cPanel Inc.
1160 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1163 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1165 PP/21 Fix eximon continuous updating with timestamped log-files.
1166 Broken in a format-string cleanup in 4.80, missed when I repaired the
1167 other false fix of the same issue.
1168 Report and fix from Heiko Schlichting.
1171 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1172 Report from Prashanth Katuri.
1174 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1175 It's SecureTransport, so affects any MacOS clients which use the
1176 system-integrated TLS libraries, including email clients.
1178 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1179 using a MIME ACL for non-SMTP local injection.
1180 Report and assistance in diagnosis by Warren Baker.
1182 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1184 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1186 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1187 last id that failed so it may be referenced in subsequent ACL's.
1189 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1192 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1193 ldap library initialization, allowing self-signed CA's to be
1194 used. Also properly sets require_cert option later in code by
1195 using NULL (global ldap config) instead of ldap handle (per
1196 session). Bug diagnosis and testing by alxgomz.
1198 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1199 the src/util/ subdirectory.
1201 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1202 renamed to Transport Post Delivery Action by Jeremy Harris, as
1205 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1206 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1207 redis_servers = needs to be configured which will be used by the redis
1208 lookup. Patch from Warren Baker, of The Packet Hub.
1210 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1212 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1213 hostname or reverse DNS when processing a host list. Used suggestions
1214 from multiple comments on this bug.
1216 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1218 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1219 Missed a few lines, added it to make the runtest require no keyboard
1222 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1223 contains upper case chars. Make router use caseful_local_part.
1225 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1226 support when GnuTLS has been built with p11-kit.
1232 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1234 This, or similar/improved, will also be change PP/11 of 4.82.
1240 PP/01 Handle short writes when writing local log-files.
1241 In practice, only affects FreeBSD (8 onwards).
1242 Bugzilla 1053, with thanks to Dmitry Isaikin.
1244 NM/01 Bugzilla 949 - Documentation tweak
1246 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1249 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1251 PP/02 Implemented gsasl authenticator.
1253 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1255 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1256 `pkg-config foo` for cflags/libs.
1258 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1259 with rest of GSASL and with heimdal_gssapi.
1261 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1262 `pkg-config foo` for cflags/libs for the TLS implementation.
1264 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1265 properties get this fed in as external SSF. A number of robustness
1266 and debugging improvements to the cyrus_sasl authenticator.
1268 PP/08 cyrus_sasl server now expands the server_realm option.
1270 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1271 Patch by Jeremy Harris.
1273 PP/10 Added dbmjz lookup type.
1275 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1277 PP/12 MAIL args handles TAB as well as SP, for better interop with
1278 non-compliant senders.
1279 Analysis and variant patch by Todd Lyons.
1281 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1282 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1283 Patch from Dirk Mueller <dmueller@suse.com>
1285 PP/13 tls_peerdn now print-escaped for spool files.
1286 Observed some $tls_peerdn in wild which contained \n, which resulted
1287 in spool file corruption.
1289 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1290 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1291 or write after TLS renegotiation, which otherwise led to messages
1294 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1295 as a tracking header (ie: a signed header comes before the signature).
1296 Patch from Wolfgang Breyha.
1298 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1299 comma-sep list; embedded commas doubled.
1301 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1303 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1305 Report and patch from Dmitry Banschikov.
1307 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1308 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1309 failures. We appear to now support TLS1.1+ with Exim.
1311 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1312 lets Exim select keys and certificates based upon TLS SNI from client.
1313 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1314 before an outbound SMTP session. New log_selector, +tls_sni.
1316 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1317 NULL dereference. Report and patch from Alun Jones.
1319 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1320 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1321 Not seeing resolver debug output on NetBSD, but suspect this is a
1322 resolver implementation change.
1324 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1325 Left warnings. Added "eximon gdb" invocation mode.
1327 PP/21 Defaulting "accept_8bitmime" to true, not false.
1329 PP/22 Added -bw for inetd wait mode support.
1331 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1332 locate the relevant includes and libraries. Made this the default.
1334 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1335 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1337 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1338 This may cause build issues on older platforms.
1340 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1341 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1342 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1343 Added SNI support via GnuTLS too.
1344 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1346 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1348 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1349 Applied second patch from Janne, implementing suggestion to default
1350 multiple-strings-in-record handling to match SPF spec.
1352 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1354 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1355 read-only, out of scope).
1356 Patch from Wolfgang Breyha, report from Stuart Northfield.
1358 PP/29 Fix three issues highlighted by clang analyser static analysis.
1359 Only crash-plausible issue would require the Cambridge-specific
1360 iplookup router and a misconfiguration.
1361 Report from Marcin Mirosław.
1363 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1365 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1366 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1367 As part of this, removing so much warning spew let me fix some minor
1368 real issues in debug logging.
1370 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1371 assignment on my part. Fixed.
1373 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1374 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1375 Janne Snabb (who went above and beyond: thank you).
1377 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1378 string otherwise requires a connection and a bunch more work and it's
1379 relatively easy to get wrong. Should also expose TLS library linkage
1382 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1383 64-bit ${eval} (JH/03).
1385 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1386 GNU libc to support some of the 64-bit stuff, should not lead to
1387 conflicts. Defined before os.h is pulled in, so if a given platform
1388 needs to override this, it can.
1390 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1391 protection layer was required, which is not implemented.
1392 Bugzilla 1254, patch from Wolfgang Breyha.
1394 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1395 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1396 tls_dhparam take prime identifiers. Also unbreak combination of
1397 OpenSSL+DH_params+TLSSNI.
1399 PP/39 Disable SSLv2 by default in OpenSSL support.
1405 PP/01 Solaris build fix for Oracle's LDAP libraries.
1406 Bugzilla 1109, patch from Stephen Usher.
1408 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1410 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1413 TF/02 Fix a couple more cases where we did not log the error message
1414 when unlink() failed. See also change 4.74-TF/03.
1416 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1417 lock up or crash if it happened to be inside a call to libc when it
1418 got a SIGUSR1 from exiwhat.
1420 The SIGUSR1 handler appends the current process status to the process
1421 log which is later printed by exiwhat. It used to use the general
1422 purpose logging code to do this, but several functions it calls are
1423 not safe for signals.
1425 The new output code in the SIGUSR1 handler is specific to the process
1426 log, and simple enough that it's easy to inspect for signal safety.
1427 Removing some special cases also simplifies the general logging code.
1428 Removing the spurious timestamps from the process log simplifies
1431 TF/04 Improved ratelimit ACL condition.
1433 The /noupdate option has been deprecated in favour of /readonly which
1434 has clearer semantics. The /leaky, /strict, and /readonly update modes
1435 are mutually exclusive. The update mode is no longer included in the
1436 database key; it just determines when the database is updated. (This
1437 means that when you upgrade Exim will forget old rate measurements.)
1439 Exim now checks that the per_* options are used with an update mode that
1440 makes sense for the current ACL. For example, when Exim is processing a
1441 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1442 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1443 must specify per_mail/readonly. If you omit the update mode it defaults to
1444 /leaky where that makes sense (as before) or /readonly where required.
1446 The /noupdate option is now undocumented but still supported for
1447 backwards compatibility. It is equivalent to /readonly except that in
1448 ACLs where /readonly is required you may specify /leaky/noupdate or
1449 /strict/noupdate which are treated the same as /readonly.
1451 A useful new feature is the /count= option. This is a generalization
1452 of the per_byte option, so that you can measure the throughput of other
1453 aggregate values. For example, the per_byte option is now equivalent
1454 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1456 The per_rcpt option has been generalized using the /count= mechanism
1457 (though it's more complicated than the per_byte equivalence). When it is
1458 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1459 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1460 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1461 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1462 Note that using per_rcpt with a non-readonly update mode in more than
1463 one ACL will cause the recipients to be double-counted. (The per_mail
1464 and per_byte options don't have this problem.)
1466 The handling of very low rates has changed slightly. If the computed rate
1467 is less than the event's count (usually one) then this event is the first
1468 after a long gap. In this case the rate is set to the same as this event's
1469 count, so that the first message of a spam run is counted properly.
1471 The major new feature is a mechanism for counting the rate of unique
1472 events. The new per_addr option counts the number of different
1473 recipients that someone has sent messages to in the last time period. It
1474 behaves like per_rcpt if all the recipient addresses are different, but
1475 duplicate recipient addresses do not increase the measured rate. Like
1476 the /count= option this is a general mechanism, so the per_addr option
1477 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1478 example, measure the rate that a client uses different sender addresses
1479 with the options per_mail/unique=$sender_address. There are further
1480 details in the main documentation.
1482 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1484 TF/06 Removed a few PCRE remnants.
1486 TF/07 Automatically extract Exim's version number from tags in the git
1487 repository when doing development or release builds.
1489 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1490 Bugzilla 879. Patch from Paul Fisher.
1492 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1493 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1496 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1498 PP/05 Variable $av_failed, true if the AV scanner deferred.
1499 Bugzilla 1078. Patch from John Horne.
1501 PP/06 Stop make process more reliably on build failure.
1502 Bugzilla 1087. Patch from Heiko Schlittermann.
1504 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1505 Bugzilla 1089. Patch from Heiko Schlittermann.
1507 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1508 Bugzilla 1131. Patch from Holger Weiß.
1510 PP/09 Handle IPv6 addresses with SPF.
1511 Bugzilla 860. Patch from Wolfgang Breyha.
1513 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1515 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1518 PP/11 match_* no longer expand right-hand-side by default.
1519 New compile-time build option, EXPAND_LISTMATCH_RHS.
1520 New expansion conditions, "inlist", "inlisti".
1522 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1524 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1526 PP/14 fix log_write() format string regression from TF/03.
1527 Bugzilla 1152. Patch from Dmitry Isaikin.
1533 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1535 PP/02 Harmonised TLS library version reporting; only show if debugging.
1536 Layout now matches that introduced for other libraries in 4.74 PP/03.
1538 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1540 PP/04 New "dns_use_edns0" global option.
1542 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1545 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1546 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1548 TK/01 Updated PolarSSL code to 0.14.2.
1549 Bugzilla 1097. Patch from Andreas Metzler.
1551 PP/07 Catch divide-by-zero in ${eval:...}.
1552 Fixes bugzilla 1102.
1554 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1557 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1558 format-string attack -- SECURITY: remote arbitrary code execution.
1560 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1561 time unintentionally subject to list matching rules, letting the header
1562 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1563 arbitrary string expansion). This allowed for information disclosure.
1565 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1566 INT_MIN/-1 -- value coerced to INT_MAX.
1572 NM/01 Workaround for PCRE version dependency in version reporting
1575 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1576 This fixes portability to compilers other than gcc, notably
1577 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1579 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1580 makefiles for portability to HP-UX and POSIX correctness.
1582 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1583 Also via indented variable definition in the Makefile.
1584 (Debugging by Oliver Heesakkers).
1586 PP/02 Restore caching of spamd results with expanded spamd_address.
1587 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1589 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1590 Improves build reliability. Fix from: Frank Elsner
1592 NM/02 Fix wide character breakage in the rfc2047 coding
1593 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1595 NM/03 Allow underscore in dnslist lookups
1596 Fixes bug 1026. Patch from Graeme Fowler
1598 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1599 Code patches from Adam Ciarcinski of NetBSD.
1601 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1604 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1605 is logged, to avoid truncation. Patch from John Horne.
1607 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1608 Patch from Jakob Hirsch.
1610 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1611 SQL string expansion failure details.
1612 Patch from Andrey Oktyabrski.
1614 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1615 Patch from Simon Arlott.
1617 PP/09 New lookups functionality failed to compile on old gcc which rejects
1618 extern declarations in function scope.
1619 Patch from Oliver Fleischmann
1621 PP/10 Use sig_atomic_t for flags set from signal handlers.
1622 Check getgroups() return and improve debugging.
1623 Fixed developed for diagnosis in bug 927 (which turned out to be
1626 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1627 Patch from Mark Zealey.
1629 PP/12 Bugzilla 1056: Improved spamd server selection.
1630 Patch from Mark Zealey.
1632 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1633 Based on patch from Heiko Schlittermann.
1635 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1636 Patch from Uwe Doering, sign-off by Michael Haardt.
1638 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1639 variable declaration deep within a block. Bug and patch from
1642 PP/15 lookups-Makefile IRIX compatibility coercion.
1644 PP/16 Make DISABLE_DKIM build knob functional.
1646 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1647 Patch by Simon Arlott
1649 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1650 variable argument macros. Our copy now differs from upstream.
1656 TF/01 Failure to get a lock on a hints database can have serious
1657 consequences so log it to the panic log.
1659 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1660 controlled using the smtp_confirmation log selector.
1662 TF/03 Include the error message when we fail to unlink a spool file.
1664 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1665 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1666 for maintaining out-of-tree patches for some time.
1668 PP/01 Bugzilla 139: Documentation and portability issues.
1669 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1670 Handle per-OS dynamic-module compilation flags.
1672 PP/02 Let /dev/null have normal permissions.
1673 The 4.73 fixes were a little too stringent and complained about the
1674 permissions on /dev/null. Exempt it from some checks.
1675 Reported by Andreas M. Kirchwitz.
1677 PP/03 Report version information for many libraries, including
1678 Exim version information for dynamically loaded libraries. Created
1679 version.h, now support a version extension string for distributors
1680 who patch heavily. Dynamic module ABI change.
1682 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1683 privilege escalation vulnerability whereby the Exim run-time user
1684 can cause root to append content of the attacker's choosing to
1687 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1690 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1691 If dropping privileges for untrusted macros, we disabled normal logging
1692 on the basis that it would fail; for the Exim run-time user, this is not
1693 the case, and it resulted in successful deliveries going unlogged.
1694 Fixed. Reported by Andreas Metzler.
1700 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1701 only prepend for the Resent-* case. Fixes regression introduced in
1702 Exim 4.70 by NM/22 for Bugzilla 607.
1704 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1705 increasing numbers of administrators be bitten by this.
1707 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1709 PP/03 Bugzilla 994: added openssl_options main configuration option.
1711 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1713 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1715 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1717 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1718 without a peer certificate, leading to a segfault because of an
1719 assumption that peers always have certificates. Be a little more
1720 paranoid. Problem reported by Martin Tscholak.
1722 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1723 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1724 NB: ClamAV planning to remove STREAM in "middle of 2010".
1725 CL also introduces -bmalware, various -d+acl logging additions and
1726 more caution in buffer sizes.
1728 PP/09 Implemented reverse_ip expansion operator.
1730 PP/10 Bugzilla 937: provide a "debug" ACL control.
1732 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1734 PP/12 Bugzilla 973: Implement --version.
1736 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1738 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1740 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1742 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1743 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1744 ignore trailing whitespace.
1746 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1748 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1751 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1752 Notification from Dr Andrew Aitchison.
1754 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1755 ExtendedDetectionInfo response format.
1756 Notification from John Horne.
1758 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1761 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1762 XSL and documented dependency on system catalogs, with examples of how
1765 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1768 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1769 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1770 configuration file which is writeable by the Exim user or group.
1772 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1773 of configuration files to cover files specified with the -C option if
1774 they are going to be used with root privileges, not just the default
1777 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1778 option (effectively making it always true).
1780 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1781 files to be used while preserving root privileges.
1783 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1784 that rogue child processes cannot use them.
1786 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1787 run-time user, instead of root.
1789 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1790 Exim run-time user without dropping privileges.
1792 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1793 result string, instead of calling string_vformat() twice with the same
1796 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1797 for other users. Others should always drop root privileges if they use
1798 -C on the command line, even for a whitelisted configure file.
1800 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1802 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1808 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1809 $data_path, and $header_path variables; fixed documentation bugs and
1812 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1813 exipick to access non-standard spools, including the "frozen" queue
1816 NM/01 Bugzilla 965: Support mysql stored procedures.
1817 Patch from Alain Williams
1819 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1821 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1822 Patch from Andreas Metzler
1824 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1825 Patch from Kirill Miazine
1827 NM/05 Bugzilla 671: Added umask to procmail example.
1829 JJ/03 installed exipick 20100323.0, fixing doc bug
1831 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1832 directory. Notification and patch from Dan Rosenberg.
1834 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1836 TK/02 Improve log output when DKIM signing operation fails.
1838 MH/01 Treat the transport option dkim_domain as a colon separated
1839 list, not as a single string, and sign the message with each element,
1840 omitting multiple occurences of the same signer.
1842 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1843 Bugzilla 985, 986. Patch by Simon Arlott
1845 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1846 Patch by Simon Arlott
1848 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1849 MBX locking. Notification from Dan Rosenberg.
1855 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1857 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1859 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1861 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1863 NM/04 Bugzilla 915: Flag broken perl installation during build.
1869 TK/01 Added patch by Johannes Berg that expands the main option
1870 "spamd_address" if it starts with a dollar sign.
1872 TK/02 Write list of recipients to X-Envelope-Sender header when building
1873 the mbox-format spool file for content scanning (suggested by Jakob
1876 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1877 (http://www.dcc-servers.net/) support via dccifd. Activated by
1878 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1880 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1881 by Mark Daniel Reidel <mr@df.eu>.
1883 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1884 When building exim an external PCRE library is now needed -
1885 PCRE is a system library on the majority of modern systems.
1886 See entry on PCRE_LIBS in EDITME file.
1888 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1889 conversation. Added nologin parameter to request.
1890 Patch contributed by Kirill Miazine.
1892 TF/01 Do not log submission mode rewrites if they do not change the address.
1894 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1896 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1897 log files in place. Contributed by Roberto Lima.
1899 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1901 TF/03 Bugzilla 615: When checking the local_parts router precondition
1902 after a local_part_suffix or local_part_prefix option, Exim now
1903 does not use the address's named list lookup cache, since this
1904 contains cached lookups for the whole local part.
1906 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1907 Robert Millan. Documentation is in experimental-spec.txt.
1909 TF/04 Bugzilla 668: Fix parallel build (make -j).
1911 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1913 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1914 Patch provided by Jan Srzednicki.
1916 TF/05 Leading white space used to be stripped from $spam_report which
1917 wrecked the formatting. Now it is preserved.
1919 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1920 that they are available at delivery time.
1922 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1924 TF/08 TLS error reporting now respects the incoming_interface and
1925 incoming_port log selectors.
1927 TF/09 Produce a more useful error message if an SMTP transport's hosts
1928 setting expands to an empty string.
1930 NM/06 Bugzilla 744: EXPN did not work under TLS.
1931 Patch provided by Phil Pennock.
1933 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1934 Patch provided by Richard Godbee.
1936 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1937 acl_smtp_notquit, added index entry.
1939 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1940 Patch provided by Eugene Bujak.
1942 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1943 accept(). Patch provided by Maxim Dounin.
1945 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1946 Patch provided by Phil Pennock.
1948 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1950 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1951 Patch provided by Brad "anomie" Jorsch.
1953 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1954 Patch provided by Dean Brooks.
1956 TK/05 Add native DKIM support (does not depend on external libraries).
1958 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1959 Patch provided by Graeme Fowler.
1961 NM/16 Bugzilla 851: Documentation example syntax fix.
1963 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1965 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1968 NM/19 Bugzilla 745: TLS version reporting.
1969 Patch provided by Phil Pennock.
1971 NM/20 Bugzilla 167: bool: condition support.
1972 Patch provided by Phil Pennock.
1974 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1975 clients. Patch provided by Phil Pennock.
1977 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1978 Patch provided by Brad "anomie" Jorsch.
1980 NM/23 Bugzilla 687: Fix misparses in eximstats.
1981 Patch provided by Heiko Schlittermann.
1983 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1984 Patch provided by Heiko Schlittermann.
1986 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1987 plus update to original patch.
1989 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1991 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1992 Patch provided by David Brownlee.
1994 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1996 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1998 NM/30 Bugzilla 888: TLS documentation bugfixes.
2000 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2002 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2003 Unlike the original bugzilla I have changed all shell scripts in src tree.
2005 NM/33 Bugzilla 898: Transport filter timeout fix.
2006 Patch by Todd Rinaldo.
2008 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2009 Patch by Serge Demonchaux.
2011 NM/35 Bugzilla 39: Base64 decode bug fixes.
2012 Patch by Jakob Hirsch.
2014 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2016 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2018 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2020 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2026 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2027 ALT-N's libdkim that I have put here:
2028 http://duncanthrax.net/exim-experimental/
2030 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2031 were called 'true' and it seems that C99 defines that as a reserved
2032 keyword to be used with 'bool' variable types. That means you could
2033 not include C99-style headers which use bools without triggering
2034 build errors in sieve.c.
2036 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2037 as mailq or other aliases. Changed the --help handling significantly
2038 to do whats expected. exim_usage() emits usage/help information.
2040 SC/01 Added the -bylocaldomain option to eximstats.
2042 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2044 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2046 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2052 PH/01 Another patch from the Sieve maintainer.
2054 PH/02 When an IPv6 address is converted to a string for single-key lookup
2055 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2056 dots are used instead of colons so that keys in lsearch files need not
2057 contain colons. This was done some time before quoting was made available
2058 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2059 (notated using the quote facility) so as to distinguish them from IPv4
2060 keys. This meant that lookups for IP addresses in host lists did not work
2061 for iplsearch lookups.
2063 This has been fixed by arranging for IPv6 addresses to be expressed with
2064 colons if the lookup type is iplsearch. This is not incompatible, because
2065 previously such lookups could never work.
2067 The situation is now rather anomalous, since one *can* have colons in
2068 ordinary lsearch keys. However, making the change in all cases is
2069 incompatible and would probably break a number of configurations.
2071 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2074 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2075 conversion specification without a maximum field width, thereby enabling
2076 a rogue spamd server to cause a buffer overflow. While nobody in their
2077 right mind would setup Exim to query an untrusted spamd server, an
2078 attacker that gains access to a server running spamd could potentially
2079 exploit this vulnerability to run arbitrary code as the Exim user.
2081 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2082 $primary_hostname instead of what libspf2 thinks the hosts name is.
2084 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2085 a directory entry by the name of the lookup key. Previously, if a
2086 symlink pointed to a non-existing file or a file in a directory that
2087 Exim lacked permissions to read, a lookup for a key matching that
2088 symlink would fail. Now it is enough that a matching directory entry
2089 exists, symlink or not. (Bugzilla 503.)
2091 PH/03 The body_linecount and body_zerocount variables are now exported in the
2094 PH/04 Added the $dnslist_matched variable.
2096 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2097 This means they are set thereafter only if the connection becomes
2100 PH/06 Added the client_condition to authenticators so that some can be skipped
2101 by clients under certain conditions.
2103 PH/07 The error message for a badly-placed control=no_multiline_responses left
2104 "_responses" off the end of the name.
2106 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2108 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2109 (without spaces) instead of just copying the configuration text.
2111 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2113 PH/11 Added $max_received_linelength.
2115 PH/12 Added +ignore_defer and +include_defer to host lists.
2117 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2118 way in which PCRE > 7.0 is built.
2120 PH/14 Implemented queue_only_load_latch.
2122 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2123 MAIL command. The effect was to mangle the value on 64-bit systems.
2125 PH/16 Another patch from the Sieve maintainer.
2127 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2129 PH/18 If a system quota error occurred while trying to create the file for
2130 a maildir delivery, the message "Mailbox is full" was not appended to the
2131 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2132 only to a quota excession during the actual writing of the file.
2134 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2135 characters?) which causes problems in log lines. The DN values are now
2136 passed through string_printing() before being added to log lines.
2138 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2139 and InterBase are left for another time.)
2141 PH/21 Added message_body_newlines option.
2143 PH/22 Guard against possible overflow in moan_check_errorcopy().
2145 PH/23 POSIX allows open() to be a macro; guard against that.
2147 PH/24 If the recipient of an error message contained an @ in the local part
2148 (suitably quoted, of course), incorrect values were put in $domain and
2149 $local_part during the evaluation of errors_copy.
2155 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2156 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2159 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2160 issue a MAIL command.
2162 PH/02 In an ACL statement such as
2164 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2166 if a client was not listed at all, or was listed with a value other than
2167 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2168 the condition was not true (as it should be), so access was not denied.
2169 The bug was that the ! inversion was incorrectly passed on to the second
2170 item. This has been fixed.
2172 PH/03 Added additional dnslists conditions == and =& which are different from
2173 = and & when the dns lookup returns more than one IP address.
2175 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2176 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2178 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2179 FSYNC, which compiles an option called disable_fsync that allows for
2180 bypassing fsync(). The documentation is heavily laced with warnings.
2182 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2184 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2185 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2186 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2187 including adding "make clean"; (3) Added -fPIC when compiling the test
2188 dynamically loaded module, to get rid of a warning.
2190 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2191 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2192 = 0s. The bug is otherwise harmless.
2194 PH/07 There was a bug in the dovecot authenticator such that the value of
2195 $auth1 could be overwritten, and so not correctly preserved, after a
2196 successful authentication. This usually meant that the value preserved by
2197 the server_setid option was incorrect.
2199 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2201 PH/09 Installed PCRE release 7.0.
2203 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2204 run for batched SMTP input. It is now run at the start of every message
2205 in the batch. While fixing this I discovered that the process information
2206 (output by running exiwhat) was not always getting set for -bs and -bS
2207 input. This is fixed, and it now also says "batched" for BSMTP.
2209 PH/11 Added control=no_pipelining.
2211 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2212 patch, slightly modified), and move the expansion of helo_data till after
2213 the connection is made in the smtp transport (so it can use these
2216 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2218 PH/14 Added log_selector = +pid.
2220 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2222 PH/16 Add ${if forany and ${if forall.
2224 PH/17 Added dsn_from option to vary the From: line in DSNs.
2226 PH/18 Flush SMTP output before performing a callout, unless control =
2227 no_callout_flush is set.
2229 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2230 was true (the default) a successful delivery failed to delete the retry
2231 item, thus causing premature timeout of the address. The bug is now
2234 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2236 PH/21 Long custom messages for fakedefer and fakereject are now split up
2237 into multiline responses in the same way that messages for "deny" and
2238 other ACL rejections are.
2240 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2241 with slight modification.
2243 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2244 draft, changing the syntax and factoring some duplicate code".
2246 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2247 for deliveries of the second and subsequent messages over the same SMTP
2250 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2251 ${reduce, with only minor "tidies".
2253 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2255 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2256 expansion side effects.
2258 PH/27 When a message times out after an over-quota error from an Exim-imposed
2259 quota, the bounce message says "mailbox is full". This message was not
2260 being given when it was a system quota that was exceeded. It now should
2263 MH/03 Made $recipients available in local_scan(). local_scan() already has
2264 better access to the recipient list through recipients_list[], but
2265 $recipients can be useful in postmaster-provided expansion strings.
2267 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2268 in the case of a MAIL command with additional options following the
2269 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2270 were accidentally chopped off.
2272 PH/29 SMTP synchronization checks are implemented when a command is read -
2273 there is a check that no more input is waiting when there shouldn't be
2274 any. However, for some commands, a delay in an ACL can mean that it is
2275 some time before the response is written. In this time, more input might
2276 arrive, invalidly. So now there are extra checks after an ACL has run for
2277 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2278 pipelining has not been advertised.
2280 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2282 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2283 returned in the SMTP response when smtp_return_error_details was false.
2284 This has been fixed.
2286 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2287 instead of the C I/O that was originally supplied, because problems were
2288 reported on Solaris.
2290 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2291 Exim which did not show up earlier: it was assuming that a call to
2292 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2293 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2294 was a macro that became an assignment, so it seemed to work. This has
2295 changed to a proper function call with a void return, hence the compile
2296 error. Exim's code has been fixed.
2298 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2301 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2303 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2305 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2306 because it turns out to be too big or there is a timeout) there is no
2307 "Completed" line in the log. When some messages of this type were
2308 selected by exigrep, they were listed as "not completed". Others were
2309 picked up by some special patterns. I have improved the selection
2310 criteria to be more general.
2312 PH/38 The host_find_failed option in the manualroute router can now be set
2313 to "ignore", to completely ignore a host whose IP address cannot be
2314 found. If all hosts are ignored, the behaviour is controlled by the new
2315 host_all_ignored option.
2317 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2318 homing or because of multiple MX records with /mx) generated more than
2319 one IP address, and the following item turned out to be the local host,
2320 all the secondary addresses of the first item were incorrectly removed
2321 from the list, along with the local host and any following hosts (which
2322 is what is supposed to happen).
2324 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2325 whoever called Exim into the -H file. In the case of the daemon it was
2326 behaving confusingly. When first started, it used values for whoever
2327 started the daemon, but after a SIGHUP it used the Exim user (because it
2328 calls itself on a restart). I have changed the code so that it now always
2331 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2332 message are rejected with the same error (e.g. no authentication or bad
2333 sender address), and a DATA command is nevertheless sent (as can happen
2334 with PIPELINING or a stupid MUA), the error message that was given to the
2335 RCPT commands is included in the rejection of the DATA command. This is
2336 intended to be helpful for MUAs that show only the final error to their
2339 PH/42 Another patch from the Sieve maintainer.
2341 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2342 Eximstats - Fixed some broken HTML links and added missing column headers
2344 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2345 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2347 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2349 PH/43 Yet another patch from the Sieve maintainer.
2351 PH/44 I found a way to check for a TCP/IP connection going away before sending
2352 the response to the final '.' that terminates a message, but only in the
2353 case where the client has not sent further data following the '.'
2354 (unfortunately, this is allowed). However, in many cases there won't be
2355 any further data because there won't be any more messages to send. A call
2356 to select() can be used: if it shows that the input is "ready", there is
2357 either input waiting, or the socket has been closed. An attempt to read
2358 the next input character can distinguish the two cases. Previously, Exim
2359 would have sent an OK response which the client would never have see.
2360 This could lead to message repetition. This fix should cure that, at
2361 least in a lot of common cases.
2363 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2364 advertised in response to EHLO.
2370 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2371 fixed by 4.65/MH/01 (is this a record?) are fixed:
2373 (i) An empty string was always treated as zero by the numeric comparison
2374 operators. This behaviour has been restored.
2376 (ii) It is documented that the numeric comparison operators always treat
2377 their arguments as decimal numbers. This was broken in that numbers
2378 starting with 0 were being interpreted as octal.
2380 While fixing these problems I realized that there was another issue that
2381 hadn't been noticed. Values of message_size_limit (both the global option
2382 and the transport option) were treated as octal if they started with 0.
2383 The documentation was vague. These values are now always treated as
2384 decimal, and I will make that clear in the documentation.
2390 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2391 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2394 MH/01 Don't check that the operands of numeric comparison operators are
2395 integers when their expansion is in "skipping" mode (fixes bug
2396 introduced by 4.64-PH/07).
2398 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2399 child addresses, Exim now panics and dies. Previously, because the count
2400 is held in a short int, deliveries were likely to be lost. As such a
2401 large number of recipients for a single message is ridiculous
2402 (performance will be very, very poor), I have chosen to impose a limit
2403 rather than extend the field.
2409 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2410 leftover -K file (the existence of which was triggered by #402).
2411 While we were at it, introduced process PID as part of the -K
2412 filename. This should rule out race conditions when creating
2415 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2416 processing considerably. Previous code took too long for large mails,
2417 triggering a timeout which in turn triggers #401.
2419 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2420 in the DK code in transports.c. sendfile() is not really portable,
2421 hence the _LINUX specificness.
2423 TF/01 In the add_headers option to the mail command in an Exim filter,
2424 there was a bug that Exim would claim a syntax error in any
2425 header after the first one which had an odd number of characters
2428 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2429 callout verification, Exim cached a "reject" for the entire domain. This
2430 is correct for most verifications, but it is not correct for a recipient
2431 verification with use_sender or use_postmaster set, because in that case
2432 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2433 case of MAIL FROM:<> rejection from other early rejections (e.g.
2434 rejection of HELO). When verifying a recipient using a non-null MAIL
2435 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2436 Whatever the result of the callout, the value of the domain cache is
2437 left unchanged (for any other kind of callout, getting as far as trying
2438 RCPT means that the domain itself is ok).
2440 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2443 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2444 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2445 session, the daemon ignores ECONNECT errors and logs others; it now
2446 ignores EPIPE as well.
2448 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2449 (quoted-printable decoding).
2451 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2452 later the small subsequent patch to fix an introduced bug.
2454 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2456 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2458 PH/08 An error is now given if message_size_limit is specified negative.
2460 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2461 to be given (somewhat) arbitrary names.
2463 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2466 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2467 miscellaneous code fixes
2469 PH/10 Added the log_reject_target ACL modifier to specify where to log
2472 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2473 hostname. This is wrong, because it relates to the incoming message (and
2474 probably the interface on which it is arriving) and not to the outgoing
2475 callout (which could be using a different interface). This has been
2476 changed to use the value of the helo_data option from the smtp transport
2477 instead - this is what is used when a message is actually being sent. If
2478 there is no remote transport (possible with a router that sets up host
2479 addresses), $smtp_active_hostname is used.
2481 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2482 tweaks were necessary in order to get it to work (see also 21 below):
2483 (a) The code assumed that strncpy() returns a negative number on buffer
2484 overflow, which isn't the case. Replaced with Exim's string_format()
2486 (b) There were several signed/unsigned issues. I just did the minimum
2487 hacking in of casts. There is scope for a larger refactoring.
2488 (c) The code used strcasecmp() which is not a standard C function.
2489 Replaced with Exim's strcmpic() function.
2490 (d) The code set only $1; it now sets $auth1 as well.
2491 (e) A simple test gave the error "authentication client didn't specify
2492 service in request". It would seem that Dovecot has changed its
2493 interface. Fortunately there's a specification; I followed it and
2494 changed what the client sends and it appears to be working now.
2496 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2499 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2500 address A is aliased to B and C, where B exists and C does not. Without
2501 -v the output is "A verified" because verification stops after a
2502 successful redirection if more than one address is generated. However,
2503 with -v the child addresses are also verified. Exim was outputting "A
2504 failed to verify" and then showing the successful verification for C,
2505 with its parentage. It now outputs "B failed to verify", showing B's
2506 parentage before showing the successful verification of C.
2508 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2509 look up a TXT record in a specific list after matching in a combined
2512 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2513 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2514 they consult the DNS. I had assumed they would set it the way they
2515 wanted; and indeed my experiments on Linux seem to show that in some
2516 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2517 To be on the safe side, however, I have now made the interface to
2518 host_find_byname() similar to host_find_bydns(), with an argument
2519 containing the DNS resolver options. The host_find_byname() function now
2520 sets these options at its start, just as host_find_bydns() does. The smtp
2521 transport options dns_qualify_single and dns_search_parents are passed to
2522 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2523 of host_find_byname() use the default settings of RES_DEFNAMES
2524 (qualify_single) but not RES_DNSRCH (search_parents).
2526 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2527 spool_read_header() do less string testing, by means of a preliminary
2528 switch on the second character of optional "-foo" lines. (This is
2529 overdue, caused by the large number of possibilities that now exist.
2530 Originally there were few.) While I was there, I also converted the
2531 str(n)cmp tests so they don't re-test the leading "-" and the first
2532 character, in the hope this might squeeze out yet more improvement.
2534 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2535 flag allowing group syntax was set by the header_syntax check but not
2536 turned off, possible causing trouble later; (2) The flag was not being
2537 set at all for the header_verify test, causing "group"-style headers to
2538 be rejected. I have now set it in this case, and also caused header_
2539 verify to ignore an empty address taken from a group. While doing this, I
2540 came across some other cases where the code for allowing group syntax
2541 while scanning a header line wasn't quite right (mostly, not resetting
2542 the flag correctly in the right place). These bugs could have caused
2543 trouble for malformed header lines. I hope it is now all correct.
2545 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2546 with the "reply" argument non-NULL. The code, however (which originally
2547 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2548 but it didn't always do it. This confused somebody who was copying the
2549 code for some other use. I have removed all the tests.
2551 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2552 feature that was used to support insecure browsers during the U.S. crypto
2553 embargo. It requires special client support, and Exim is probably the
2554 only MTA that supported it -- and would never use it because real RSA is
2555 always available. This code has been removed, because it had the bad
2556 effect of slowing Exim down by computing (never used) parameters for the
2557 RSA_EXPORT functionality.
2559 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2560 authenticator to fail if there's a tab character in the incoming data
2561 (there should never be unless someone is messing about, as it's supposed
2562 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2563 if the connection is using TLS or if the remote IP is the same as the
2564 local IP, and the "valid-client-cert option" if a client certificate has
2567 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2568 authenticators. This can be used for authorization after authentication
2569 succeeds. (In the case of plaintext, it servers for both authentication
2572 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2573 if any retry times were supplied.
2575 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2576 connection, where there is no client IP address to check. In this
2577 situation, the verify now always succeeds.
2579 PH/25 Applied John Jetmore's -Mset patch.
2581 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2583 PH/27 In a string expansion for a processed (not raw) header when multiple
2584 headers of the same name were present, leading whitespace was being
2585 removed from all of them, but trailing whitespace was being removed only
2586 from the last one. Now trailing whitespace is removed from each header
2587 before concatenation. Completely empty headers in a concatenation (as
2588 before) are ignored.
2590 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2591 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2593 PH/29 [Removed. This was a change that I later backed out, and forgot to
2594 correct the ChangeLog entry (that I had efficiently created) before
2595 committing the later change.]
2597 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2598 address errors (4xx response to RCPT) over the same connection as other
2599 messages routed to the same hosts. Such deliveries are always "forced",
2600 so retry times are not inspected. This resulted in far too many retries
2601 for the affected addresses. The effect occurred only when there were more
2602 hosts than the hosts_max_try setting in the smtp transport when it had
2603 the 4xx errors. Those hosts that it had tried were not added to the list
2604 of hosts for which the message was waiting, so if all were tried, there
2605 was no problem. Two fixes have been applied:
2607 (i) If there are any address or message errors in an SMTP delivery, none
2608 of the hosts (tried or untried) are now added to the list of hosts
2609 for which the message is waiting, so the message should not be a
2610 candidate for sending over the same connection that was used for a
2611 successful delivery of some other message. This seems entirely
2612 reasonable: after all the message is NOT "waiting for some host".
2613 This is so "obvious" that I'm not sure why it wasn't done
2614 previously. Hope I haven't missed anything, but it can't do any
2615 harm, as the worst effect is to miss an optimization.
2617 (ii) If, despite (i), such a delivery is accidentally attempted, the
2618 routing retry time is respected, so at least it doesn't keep
2619 hammering the server.
2621 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2622 in ${readsocket because some servers need this prod.
2624 PH/32 Added some extra debug output when updating a wait-xxx database.
2626 PH/33 The hint "could be header name not terminated by colon", which has been
2627 given for certain expansion errors for a long time, was not being given
2628 for the ${if def:h_colon_omitted{... case.
2630 PH/34 The spec says: "With one important exception, whenever a domain list is
2631 being scanned, $domain contains the subject domain." There was at least
2632 one case where this was not true.
2634 PH/35 The error "getsockname() failed: connection reset by peer" was being
2635 written to the panic log as well as the main log, but it isn't really
2636 panic-worthy as it just means the connection died rather early on. I have
2637 removed the panic log writing for the ECONNRESET error when getsockname()
2640 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2641 runs only) independently of the message's sender address. This meant
2642 that, if the 4xx error was in fact related to the sender, a different
2643 message to the same recipient with a different sender could confuse
2644 things. In particular, this can happen when sending to a greylisting
2645 server, but other circumstances could also provoke similar problems.
2646 I have changed the default so that the retry time for these errors is now
2647 based a combination of the sender and recipient addresses. This change
2648 can be overridden by setting address_retry_include_sender=false in the
2651 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2652 remote server are returned as part of bounce messages. This was not
2653 happening for LMTP over a pipe (the lmtp transport), but now it is the
2654 same for both kinds of LMTP.
2656 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2657 in header lines that were in fact CNAMEs. This is no longer the case.
2659 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2660 and queue runs started by the daemon processed all messages. This has
2661 been fixed so that -R and -S can now usefully be given with -q<time>.
2663 PH/40 Import PCRE release 6.7 (fixes some bugs).
2665 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2667 PH/42 Give an error if -q is specified more than once.
2669 PH/43 Renamed the variables $interface_address and $interface_port as
2670 $received_ip_address and $received_port, to make it clear that these
2671 values apply to message reception, and not to the outgoing interface when
2672 a message is delivered. (The old names remain recognized, of course.)
2674 PH/44 There was no timeout on the connect() call when using a Unix domain
2675 socket in the ${readsocket expansion. There now is.
2677 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2678 be meaningful with "accept".
2680 SC/01 Eximstats V1.43
2681 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2683 SC/02 Eximstats V1.44
2684 Use a glob alias rather than an array ref in the generated
2685 parser. This improves both readability and performance.
2687 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2688 Collect SpamAssassin and rejection statistics.
2689 Don't display local sender or destination tables unless
2690 there is data to show.
2691 Added average volumes into the top table text output.
2693 SC/04 Eximstats V1.46
2694 Collect data on the number of addresses (recipients)
2695 as well as the number of messages.
2697 SC/05 Eximstats V1.47
2698 Added 'Message too big' to the list of mail rejection
2699 reasons (thanks to Marco Gaiarin).
2701 SC/06 Eximstats V1.48
2702 Mainlog lines which have GMT offsets and are too short to
2703 have a flag are now skipped.
2705 SC/07 Eximstats V1.49 (Alain Williams)
2706 Added the -emptyok flag.
2708 SC/08 Eximstats V1.50
2709 Fixes for obtaining the IP address from reject messages.
2711 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2712 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2713 whitespace changes from 4.64-PH/27
2715 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2718 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2719 are found, allow negative numbers in numeric criteria)
2721 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2723 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2724 to match changes made in 4.64-PH/43
2726 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2728 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2729 "domain" is actually an IP address, and force a failure. This locks out
2730 those revolvers/nameservers that support "A-for-A" lookups, in
2731 contravention of the specifications.
2733 PH/48 When a host name was looked up from an IP address, and the subsequent
2734 forward lookup of the name timed out, the host name was left in
2735 $sender_host_name, contrary to the specification.
2737 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2738 restricted to single-key lookups, Exim was not diagnosing an error if
2739 * or *@ was used with a query-style lookup.
2741 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2743 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2744 long ago, but noone interested enough thought of it. Let's just say that
2745 the "1.1" means that there are some new functions that weren't there at
2746 some point in the past.
2748 PH/51 Error processing for expansion failure of helo_data from an smtp
2749 transport during callout processing was broken.
2751 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2752 tested/used via the -bh/-bhc/-bs options.
2754 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2755 bug, fixed in subsequent PCRE releases).
2757 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2758 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2760 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2766 SC/01 Use a glob alias rather than an array ref in eximstats generated
2767 parser. This improves both readability and performance.
2769 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2770 Don't display local sender or destination tables in eximstats unless
2771 there is data to show.
2772 Added average volumes into the eximstats top table text output.
2774 SC/03 Collect data on the number of addresses (recipients) as well
2775 as the number of messages in eximstats.
2777 TF/01 Correct an error in the documentation for the redirect router. Exim
2778 does (usually) call initgroups() when daemonizing.
2780 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2781 with consistent privilege compared to when running as a daemon.
2783 TF/03 Note in the spec that $authenticated_id is not set for local
2784 submissions from trusted users.
2786 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2787 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2789 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2790 by adding some example configuration directives to the default
2791 configuration file. A little bit of work is required to uncomment the
2792 directives and define how usernames and passwords are checked, but
2793 there is now a framework to start from.
2795 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2796 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2797 without this. I don't know how relevant this is to other LDAP libraries.
2799 PH/02 Add the verb name to the "unknown ACL verb" error.
2801 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2803 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2805 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2806 directory not expanded when it should be if an expanded home directory
2807 was set for the address (which is overridden by the transport).
2809 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2812 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2813 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2814 because it is too late at that time, and has no effect.
2816 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2817 security issue with \' (bugzilla #107). I could not use the
2818 PQescapeStringConn() function, because it needs a PGconn value as one of
2821 PH/08 When testing addresses using -bt, indicate those final addresses that
2822 are duplicates that would not cause an additional delivery. At least one
2823 person was confused, thinking that -bt output corresponded to deliveries.
2824 (Suppressing duplicates isn't a good idea as you lose the information
2825 about possibly different redirections that led to the duplicates.)
2827 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2828 systems where poll() doesn't work, in particular OS X.
2830 PH/10 Added more information to debugging output for retry time not reached.
2832 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2833 operations in malware.c.
2835 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2838 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2839 syslog_duplication set false, log lines that would normally be written
2840 both the the main log and to the reject log were not written to syslog at
2843 PH/14 In the default configuration, change the use of "message" in ACL warn
2844 statements to "add_header".
2846 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2847 not followed by a command (e.g. "seen endif").
2849 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2850 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2853 PH/17 Added extra conditions to the default value of delay_warning_condition
2857 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2858 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2859 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2862 The Auto-Submitted: and various List- headers are standardised, whereas I
2863 don't think Precedence: ever was.
2865 PH/18 Refactored debugging code in route_finduser() to show more information,
2866 in particular, the error code if getpwnam() issues one.
2868 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2869 This is apparently needed in addition to the PH/07 change above to avoid
2870 any possible encoding problems.
2872 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2873 but not after initializing Perl.
2875 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2876 output them only if debugging. By default they are written stderr,
2877 apparently, which is not desirable.
2879 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2882 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2885 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2887 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2888 authenticated or an ident call has been made. Suppress the default
2889 values for $authenticated_id and $authenticated_sender (but permit -oMai
2890 and -oMas) when testing with -bh.
2892 PH/24 Re-jigged the order of the tests in the default configuration so that the
2893 tests for valid domains and recipients precede the DNS black list and CSA
2894 tests, on the grounds that those ones are more expensive.
2896 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2897 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2898 This bug exists in every version of Exim that I still have, right back to
2901 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2902 However, an attempt to turn on case-sensitivity in a regex key by
2903 including (?-i) didn't work because the subject string was already
2904 lowercased, and the effects were non-intuitive. It turns out that a
2905 one-line patch can be used to allow (?-i) to work as expected.
2911 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2912 other effects) broke the use of negated acl sub-conditions.
2914 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2917 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2918 "Deny" causes Exim to reject the incoming connection with a 554 error.
2919 Unfortunately, if there is a major crisis, such as a disk failure,
2920 tcp-wrappers gives "deny", whereas what one would like would be some
2921 kind of temporary error. A kludge has been added to help with this.
2922 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2923 554 error is used if errno is still zero or contains ENOENT (which occurs
2924 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2927 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2929 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2930 errors. Otherwise a message that provokes a temporary error (when other
2931 messages do not) can cause a whole host to time out.
2933 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2934 addresses were routed directly to files or pipes from a redirect router.
2935 File deliveries just didn't batch; pipe deliveries might have suffered
2938 PH/06 A failure to get a lock for a hints database would erroneously always say
2939 "Failed to get write lock", even when it was really a read lock.
2941 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2942 of 0600. This has been changed to use the value of the lockfile_mode
2943 option (which defaults to 0600).
2945 PH/08 Applied small patch from the Sieve maintainer.
2947 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2948 folder from quota calculations, a direct delivery into this folder messed
2949 up the contents of the maildirsize file. This was because the regex was
2950 used only to exclude .Trash (or whatever) when the size of the mailbox
2951 was calculated. There was no check that a delivery was happening into an
2952 excluded directory. This bug has been fixed by ignoring all quota
2953 processing for deliveries into excluded directories.
2955 PH/10 Added the maildirfolder_create_regex option to appendfile.
2961 PH/01 The code for finding all the local interface addresses on a FreeBSD
2962 system running IPv6 was broken. This may well have applied to all BSD
2963 systems, as well as to others that have similar system calls. The broken
2964 code found IPv4 interfaces correctly, but gave incorrect values for the
2965 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2966 that it would not match correctly against @[] and not recognize the IPv6
2969 PH/02 The ipliteral router was not recognizing addresses of the form user@
2970 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2972 PH/03 Added disable_ipv6.
2974 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2975 decoded form, because it is most often used to construct To: headers
2976 lines in autoreplies, and the decoded form may well be syntactically
2977 invalid. However, $reply_address has leading white space removed, and all
2978 newlines turned into spaces so that the autoreply transport does not
2981 PH/05 If group was specified without a user on a router, and no group or user
2982 was specified on a transport, the group from the router was ignored.
2984 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2985 for visible compile-time settings that can be used to change these
2986 numbers, for those that want even more. Backwards compatibility with old
2987 spool files has been maintained. However, going back to a previous Exim
2988 release will lost any variables that are in spool files.
2990 PH/07 Two small changes when running in the test harness: increase delay when
2991 passing a TCP/IP connection to a new process, in case the original
2992 process has to generate a bounce, and remove special handling of
2993 127.0.0.2 (sic), which is no longer necessary.
2995 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2996 be the same on different OS.
2998 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3001 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3002 whether --show-vars was specified or not
3004 JJ/02 exipick: Added support for new ACL variable spool format introduced
3007 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3008 syntactically invalid From: or Reply-to: line, and a filter used this to
3009 generate an autoreply, and therefore failed to obtain an address for the
3010 autoreply, Exim could try to deliver to a non-existent relative file
3011 name, causing unrelated and misleading errors. What now happens is that
3012 it logs this as a hard delivery error, but does not attempt to create a
3015 PH/11 The exinext utility has a -C option for testing purposes, but although
3016 the given file was scanned by exinext itself; it wasn't being passed on
3017 when Exim was called.
3019 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3020 an end-of-file indication when reading a command response.
3022 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3023 compiled. In many other places in Exim, IPv6 addresses are always
3024 recognized, so I have changed this. It also means that IPv4 domain
3025 literals of the form [IPV4:n.n.n.n] are now always recognized.
3027 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3028 used if the router is not running as root, for example, when verifying at
3029 ACL time, or when using -bh. The debugging output from this situation was
3030 non-existent - all you got was a failure to exec. I have made two
3033 (a) Failures to set uid/gid, the current directory, or a process leader
3034 in a subprocess such as that created by queryprogram now generate
3035 suitable debugging output when -d is set.
3037 (b) The queryprogram router detects when it is not running as root,
3038 outputs suitable debugging information if -d is set, and then runs
3039 the subprocess without attempting to change uid/gid.
3041 PH/15 Minor change to Makefile for building test_host (undocumented testing
3044 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3045 additional section of a DNS packet that returns MX or SRV records.
3046 Instead, it always explicitly searches for A/AAAA records. This avoids
3047 major problems that occur when a DNS server includes only records of one
3048 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3049 fixed another bug: if SRV records were looked up and the corresponding
3050 address records were *not* found in the additional section, the port
3051 values from the SRV records were lost.
3053 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3054 using the correct key (the original address) when searching the retry
3055 rules in order to find which one to use for generating the retry hint.
3057 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3058 adding the default one. Similarly, if it contains a Reply-To: header, the
3059 errors_reply_to option, if set, is not used.
3061 PH/19 When calculating a retry time, Exim used to measure the "time since
3062 failure" by looking at the "first failed" field in the retry record. Now
3063 it does not use this if it is later than than the arrival time of the
3064 message. Instead it uses the arrival time. This makes for better
3065 behaviour in cases where some deliveries succeed, thus re-setting the
3066 "first failed" field. An example is a quota failure for a huge message
3067 when small messages continue to be delivered. Without this change, the
3068 "time since failure" will always be short, possible causing more frequent
3069 delivery attempts for the huge message than are intended.
3070 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3072 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3073 $1, $2, $3) because the numerical variables can be reset during some
3074 expansion items (e.g. "match"), thereby losing the authentication data.
3076 PH/21 Make -bV show the size of off_t variables so that the test suite can
3077 decide whether to run tests for quotas > 2G.
3079 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3080 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3081 filecount value is greater than 2G or if a quota value is greater than 2G
3082 on a system where the size of off_t is not greater than 4, a panic error
3085 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3086 never match. The debug and -bh output now contains an explicit error
3087 message indicating a malformed IPv4 address or mask.
3089 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3090 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3091 PH/23 above applies.
3093 PH/25 Do not write to syslog when running in the test harness. The only
3094 occasion when this arises is a failure to open the main or panic logs
3095 (for which there is an explicit test).
3097 PH/26 Added the /no_tell option to "control=freeze".
3099 PH/27 If a host name lookup failed very early in a connection, for example, if
3100 the IP address matched host_lookup and the reverse lookup yielded a name
3101 that did not have a forward lookup, an error message of the form "no IP
3102 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3103 could be logged. Now it outputs the IP address instead of "NULL".
3105 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3106 allows the sender and the authenticated sender to be set when
3107 submitting a message from within Exim. Since child_open_exim() is
3108 documented for local_scan(), the new function should be too.
3110 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3111 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3112 results in an empty string is now treated as unset.
3114 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3116 PH/31 Added sender_verify_fail logging option.
3118 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3119 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3120 tidied the source and removed it altogether.
3122 PH/33 When a queue run was abandoned because the load average was too high, a
3123 log line was always written; now it is written only if the queue_run log
3124 selector is set. In addition, the log line for abandonment now contains
3125 information about the queue run such as the pid. This is always present
3126 in "start" and "stop" lines but was omitted from the "abandon" line.
3128 PH/34 Omit spaces between a header name and the colon in the error message that
3129 is given when verify = headers_syntax fails (if there are lots of them,
3130 the message gets confusing).
3132 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3133 names, as there are now some PTR records that contain slashes. This check
3134 is only to protect against broken name servers that fall over on strange
3135 characters, so the fact that it applies to all lookups doesn't matter.
3137 PH/36 Now that the new test suite is complete, we can remove some of the
3138 special code in Exim that was needed for the old test suite. For example,
3139 sorting DNS records because real resolvers return them in an arbitrary
3140 order. The new test suite's fake resolver always returns records in the
3143 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3144 bounces) except when queue_only is set, to avoid logging races between
3145 the different processes.
3147 PH/38 Panic-die if .include specifies a non-absolute path.
3149 PH/39 A tweak to the "H" retry rule from its user.
3151 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3152 a label. They prevented compilation on older perls.
3154 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3155 a warning to be raised on newish perls.
3157 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3158 on queue. Changes to match documented behaviour of showing count of
3159 messages matching specified criteria.
3161 PH/40 Changed the default ident timeout from 30s to 5s.
3163 PH/41 Added support for the use of login_cap features, on those BSD systems
3164 that have them, for controlling the resources used by pipe deliveries.
3166 PH/42 The content-scanning code uses fopen() to create files in which to put
3167 message data. Previously it was not paying any attention to the mode of
3168 the files. Exim runs with umask(0) because the rest of the code creates
3169 files with open(), and sets the required mode explicitly. Thus, these
3170 files were ending up world-writeable. This was not a big issue, because,
3171 being within the spool directory, they were not world-accessible. I have
3172 created a function called modefopen, which takes an additional mode
3173 argument. It sets umask(777), creates the file, chmods it to the required
3174 mode, then resets the umask. All the relevant calls to fopen() in the
3175 content scanning code have been changed to use this function.
3177 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3178 to 24 hours. This avoids potential overflow problems when processing G
3179 and H retry rules. I suspect nobody ever tinkers with this value.
3181 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3183 PH/45 When the plaintext authenticator is running as a client, the server's
3184 challenges are checked to ensure they are valid base64 strings. By
3185 default, the authentication attempt is cancelled if an invalid string is
3186 received. Setting client_ignore_invalid_base64 true ignores these errors.
3187 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3188 they are received. Thus, the responses can be made to depend on the
3189 challenges. If an invalid string is ignored, an empty string is placed in
3192 PH/46 Messages that are created by the autoreply transport now contains a
3193 References: header, in accordance with RFCs 2822 and 3834.
3195 PH/47 Added authenticated_sender_force to the smtp transport.
3197 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3199 PH/49 Installed latest patch from the Sieve maintainer.
3201 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3202 was also set, the appendfile transport was unnecessarily scanning a
3203 directory of message files (e.g. for maildir delivery) to find the count
3204 of files (along with the size), even though it did not need this
3205 information. It now does the scan only if it needs to find either the
3206 size of the count of files.
3208 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3210 PH/52 Two bugs concerned with error handling when the smtp transport is
3213 (i) Exim was not creating retry information for temporary errors given
3214 for individual recipients after the DATA command when the smtp transport
3215 was used in LMTP mode. This meant that they could be retried too
3216 frequently, and not timed out correctly.
3218 (ii) Exim was setting the flag that allows error details to be returned
3219 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3220 recipients that were returned after the DATA command.
3222 PH/53 This is related to PH/52, but is more general: for any failing address,
3223 when detailed error information was permitted to be returned to the
3224 sender, but the error was temporary, then after the final timeout, only
3225 "retry timeout exceeded" was returned. Now it returns the full error as
3226 well as "retry timeout exceeded".
3228 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3229 do this, and (what is worse) MTAs that accept it.
3231 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3232 will now be deprecated.
3234 PH/56 New os.c-cygwin from the Cygwin maintainer.
3236 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3237 formats (previously only available in exim formats via -bpr, -bpru,
3238 and -bpra. Now also available in native and exiqgrep formats)
3240 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3241 with very large, slow to parse queues
3243 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3245 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3247 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3248 responses to authentication challenges, though it was showing the
3249 challenges; (ii) I've removed the CR characters from the debug output for
3252 PH/58 Allow for the insertion of a newline as well as a space when a string
3253 is turned into more than one encoded-word during RFC 2047 encoding. The
3254 Sieve code now uses this.
3256 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3257 data_4xx, lost_connection, tls_required.
3259 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3260 message was being sent as an SMTP response.
3262 PH/61 Add -l and -k options to exicyclog.
3264 PH/62 When verifying, if an address was redirected to one new address, so that
3265 verification continued, and the new address failed or deferred after
3266 having set something in $address_data, the value of $address_data was not
3267 passed back to the ACL. This was different to the case when no
3268 redirection occurred. The value is now passed back in both cases.
3270 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3271 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3272 use login_cap.h, so on its own it isn't the distinguishing feature. The
3273 new name refers directly to the setclassresources() function.
3275 PH/65 Added configuration files for NetBSD3.
3277 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3279 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3280 is preferred over IPv4.
3282 PH/68 The bounce_return_message and bounce_return_body options were not being
3283 honoured for bounces generated during the reception of non-SMTP messages.
3284 In particular, this applied to messages rejected by the ACL. This bug has
3285 been fixed. However, if bounce_return_message is true and bounce_return_
3286 body is false, the headers that are returned for a non-SMTP message
3287 include only those that have been read before the error was detected.
3288 (In the case of an ACL rejection, they have all been read.)
3290 PH/69 The HTML version of the specification is now built in a directory called
3291 spec_html instead of spec.html, because the latter looks like a path with
3292 a MIME-type, and this confuses some software.
3294 PH/70 Catch two compiler warnings in sieve.c.
3296 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3297 function verify_get_ident() calls ip_connect() to connect a socket, but
3298 if the "connect()" function timed out, ip_connect() used to close the
3299 socket. However, verify_get_ident() also closes the socket later, and in
3300 between Exim writes to the log, which may get opened at this point. When
3301 the socket was closed in ip_connect(), the log could get the same file
3302 descriptor number as the socket. This naturally causes chaos. The fix is
3303 not to close the socket in ip_connect(); the socket should be closed by
3304 the function that creates it. There was only one place in the code where
3305 this was missing, in the iplookup router, which I don't think anybody now
3306 uses, but I've fixed it anyway.
3308 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3309 well as to direct DNS lookups. Otherwise the handling of names in host
3310 lists is inconsistent and therefore confusing.
3316 PH/01 Two changes to the default runtime configuration:
3318 (1) Move the checks for relay_from_hosts and authenticated clients from
3319 after to before the (commented out) DNS black list checks.
3321 (2) Add control=submission to the relay_from_hosts and authenticated
3322 clients checks, on the grounds that messages accepted by these
3323 statements are most likely to be submissions.
3325 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3327 (1) Generate an error if the third argument for the ${prvs expansion is
3330 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3333 (3) Reset the variables that are obtained from the first argument of
3334 ${prvscheck and used in the second argument before leaving the code,
3335 because their memory is reclaimed, so using them afterwards may do
3338 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3339 one (it's much easier than Tom thought :-).
3341 (5) Because of (4), we can now allow for the use of $prvscheck_result
3342 inside the third argument.
3344 PH/03 For some reason, the default setting of PATH when running a command from
3345 a pipe transport was just "/usr/bin". I have changed it to
3348 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3349 anything to be listed in the output from -bV.
3351 PH/05 When a filter generated an autoreply, the entire To: header line was
3352 quoted in the delivery log line, like this:
3354 => >A.N.Other <ano@some.domain> <original@ddress> ...
3356 This has been changed so that it extracts the operative address. There
3357 may be more than one such address. If so, they are comma-separated, like
3360 => >ano@some.domain,ona@other.domain <original@ddress> ...
3362 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3363 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3364 not being looked up in the rDNS to get a host name, Exim was showing the
3365 IP address twice in Received: lines, even though the IP addresses were
3366 identical. For example:
3368 Received: from [1.2.3.4] (helo=[1.2.3.4])
3370 However, if the real host name was known, it was omitting the HELO data
3371 if it matched the actual IP address. This has been tidied up so that it
3372 doesn't show the same IP address twice.
3374 PH/07 When both +timestamp and +memory debugging was on, the value given by
3375 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3376 called by the debug printing, thereby overwriting the timestamp buffer.
3377 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3379 PH/08 When the original message was included in an autoreply transport, it
3380 always said "this is a copy of the message, including all the headers",
3381 even if body_only or headers_only was set. It now gives an appropriate
3384 PH/09 Applied a patch from the Sieve maintainer which:
3386 o fixes some comments
3387 o adds the (disabled) notify extension core
3388 o adds some debug output for the result of if/elsif tests
3389 o points to the current vacation draft in the documentation
3390 and documents the missing references header update
3394 o fixes a bug in processing the envelope test (when testing
3395 multiple envelope elements, the last element determined the
3398 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3399 Electronic Mail") by including:
3401 Auto-submitted: auto-generated
3403 in the messages that it generates (bounce messages and others, such as
3404 warnings). In the case of bounce messages for non-SMTP messages, there was
3405 also a typo: it was using "Auto_submitted" (underscore instead of
3406 hyphen). Since every message generated by Exim is necessarily in response
3407 to another message, thes have all been changed to:
3409 Auto-Submitted: auto-replied
3411 in accordance with these statements in the RFC:
3413 The auto-replied keyword:
3415 - SHOULD be used on messages sent in direct response to another
3416 message by an automatic process,
3418 - MUST NOT be used on manually-generated messages,
3420 - MAY be used on Delivery Status Notifications (DSNs) and Message
3421 Disposition Notifications (MDNs),
3423 - MUST NOT be used on messages generated by automatic or periodic
3424 processes, except for messages which are automatic responses to
3427 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3428 to the default Received: header definition.
3430 PH/12 Added log selector acl_warn_skipped (default on).
3432 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3433 variables because (a) they are in the wrong storage pool and (b) even if
3434 they were copied, it wouldn't work properly because of the caching.
3436 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3437 checking when decoding. Apparently there are clients that generate
3438 overlong encoded strings. Why am I not surprised?
3440 PH/15 If the first argument of "${if match_address" was not empty, but did not
3441 contain an "@" character, Exim crashed. Now it writes a panic log message
3442 and treats the condition as false.
3444 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3446 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3447 extension "envelope-auth". The code is finished and in agreement with
3448 other implementations, but there is no documentation so far and in fact,
3449 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3450 not changing the active code.
3452 Print executed "if" and "elsif" statements when debugging is used. This
3453 helps a great deal to understand what a filter does.
3455 Document more things not specified clearly in RFC3028. I had all this
3456 sorted out, when out of a sudden new issues came to my mind. Oops."
3458 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3461 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3462 canonical form (as documented). However, after a host name lookup from
3463 the IP address, check_host() was doing a simple string comparison with
3464 addresses acquired from the DNS when checking that the found name did
3465 have the original IP as one of its addresses. Since any found IPv6
3466 addresses are likely to be in abbreviated form, the comparison could
3467 fail. Luckily, there already exists a function for doing the comparison
3468 by converting both addresses to binary, so now that is used instead of
3469 the text comparison.
3471 PH/20 There was another similar case to PH/19, when a complete host name was
3472 given in a host list; looking up its IP address could give an abbreviated
3473 form, whereas the current host's name might or might not be abbreviated.
3474 The same fix has been applied.
3480 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3481 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3484 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3485 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3487 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3489 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3490 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3491 Also, for "Auto-Submitted", treat anything other than "no" as
3492 non-personal, in accordance with RFC 3834. (Previously it treated
3493 anything starting "auto-" as non-personal.)
3495 TF/01 The control=submission/name=... option had a problem with syntax
3496 errors if the name included a slash character. The /name= option
3497 now slurps the rest of the string, so it can include any characters
3498 but it must come last in the list of options (after /sender_retain
3501 PH/05 Some modifications to the interface to the fake nameserver for the new
3509 TK/01 Added the "success_on_redirect" address verification option. See
3510 NewStuff for rationale and an example.
3512 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3514 PH/02 Patch to exigrep to allow it to work on syslog lines.
3516 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3517 fread() to skip over the body file's header line, because in Cygwin the
3518 header line is locked and is inaccessible.
3520 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3521 co-exist for some time) to make it clear that it is the Exim ID that is
3522 referenced, not the Message-ID: header line.
3524 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3525 string_format() function, because snprintf() does not exist on all
3528 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3529 ${stat: expansion item.
3531 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3532 protocol synchronization error", to keep the pedants happy.
3534 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3535 well as for IRIX systems, when gcc is being used. See the host.c source
3538 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3540 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3543 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3544 transport and to the smtp transport in LMTP mode.
3546 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3548 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3549 run a filter in a subprocess. This could lead to confusion in subsequent
3550 lookups in the parent process. There should also be a search_tidyup() at
3551 the end of the subprocess.
3553 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3554 only if the host matched helo_try_verify_hosts, which caused the
3555 verification to occur when the EHLO/HELO command was issued. The ACL just
3556 tested the remembered result. Now, if a previous verification attempt has
3557 not happened, "verify = helo" does it there and then.
3559 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3561 TK/03 Fix log output including CR from clamd.
3563 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3564 exist provoked a memory error which could cause a segfault.
3566 PH/15 Installed PCRE 6.2
3568 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3570 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3571 of the problem. Specifically, suggested +O2 rather than +O1 for the
3574 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3576 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3577 record, the port was not correctly logged when the outgoing_port log
3578 selector was set (it logged the transort's default port).
3580 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3581 fallback_hosts, and "hosts" in the smtp transport.
3583 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3584 host errors such as "Connection refused".
3586 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3587 authentication with radiusclient 0.4.9:
3589 - Error returned from rc_read_config was caught wrongly
3590 - Username/password not passed on to radius server due to wrong length.
3592 The presumption is that some radiusclient API changes for 4.51/PH/17
3593 were not taken care of correctly. The code is still untested by me (my
3594 Linux distribution still has 0.3.2 of radiusclient), but it was
3595 contributed by a Radius user.
3597 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3598 expanding the "port" option of the smtp transport.
3600 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3601 while reading a MIME header. Thanks to Tom Hughes for a patch.
3603 PH/24 Include config.h inside local_scan.h so that configuration settings are
3606 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3607 that in an ACL for RCPT (for example), you can examine exactly what was
3610 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3611 commands, but it was not correctly comparing the address with the actual
3612 client host address. Thus, it would show the EHLO address in Received:
3613 header lines when this was not necessary.
3615 PH/27 Added the % operator to ${eval:}.
3617 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3618 it should be ignoring failures (because with -C, for example, it has lost
3619 privilege). It wasn't ignoring creation failures other than "already
3622 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3625 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3626 because an input line was too long, either on its own, or by virtue of
3627 too many continuations, the temporary file was not being removed, and the
3628 return code was incorrect.
3630 PH/31 Missing "BOOL" in function definition in filtertest.c.
3632 PH/32 Applied Sieve patches from the maintainer.
3634 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3636 PH/33 Added "verify = not_blind".
3638 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3639 Local/Makefile (with some defaults set). These are used in built scripts
3640 such as exicyclog, but they have never been used in the exim_install
3641 script (though there are many overriding facilities there). I have
3642 arranged that the exim_install script now takes note of these two
3645 PH/35 Installed configuration files for Dragonfly.
3647 PH/36 When a locally submitted message by a trusted user did not contain a
3648 From: header, and the sender address was obtained from -f or from an SMTP
3649 MAIL command, and the trusted user did not use -F to supply a sender
3650 name, $originator_name was incorrectly used when constructing a From:
3651 header. Furthermore, $originator_name was used for submission mode
3652 messages from external hosts without From: headers in a similar way,
3653 which is clearly wrong.
3655 PH/37 Added control=suppress_local_fixups.
3657 PH/38 When log_selector = +received_sender was set, and the addition of the
3658 sender made the log line's construction buffer exactly full, or one byte
3659 less than full, an overflow happened when the terminating "\n" was
3662 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3663 when the result of a list match is failure because a DNS lookup failed.
3665 PH/40 RM_COMMAND is now used in the building process.
3667 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3668 the "build-* directories that it finds.
3670 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3671 address in a domain literal was a prefix of an interface address.
3673 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3674 when verifying a sender address, unless rewrite_headers is false.
3676 PH/44 Wrote a long comment about why errors_to addresses are verified as
3677 recipients, not senders.
3679 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3680 the ratelimit ACL was added.
3682 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3684 PH/46 Added extra information about PostgreSQL errors to the error string.
3686 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3687 avoiding the need to install special zones in a real server. This is
3688 backwards compatible; if it can't find the fake resolver, it drops back.
3689 Thus, both old and new test suites can be run.
3691 TF/02 Added util/ratelimit.pl
3693 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3694 clock is set back in time.
3696 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3697 Candler <B.Candler@pobox.com>.
3699 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3700 OK for addresses that are the result of redirections.
3702 PH/48 A number of further additions for the benefit of the new test suite,
3703 including a fake gethostbyname() that interfaces to the fake DNS resolver
3706 TF/06 The fix for widen_domains has also been applied to qualify_single and
3707 search_parents which are the other dnslookup options that can cause
3710 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3713 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3715 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3716 service" if it encountered an MX record with an empty target hostname.
3717 The message is now "an MX or SRV record indicated no SMTP service".
3719 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3720 if the DNS of the sending site is misconfigured. This is quite a
3721 common situation. This change restores the behaviour of treating a
3722 helo verification defer as a failure.
3724 PH/51 If self=fail was set on a router, the bounce message did not include the
3725 actual error message.
3731 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3733 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3734 command itself ended in error, the underlying message about the transport
3735 filter timeout was being overwritten with the pipe command error. Now the
3736 underlying error message should be appended to the second error message.
3738 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3740 PH/02 Reduce the amount of output that "make" produces by default. Full output
3741 can still be requested.
3743 PH/03 The warning log line about a condition test deferring for a "warn" verb
3744 was being output only once per connection, rather than after each
3745 occurrence (because it was using the same function as for successful
3746 "warn" verbs). This seems wrong, so I have changed it.
3748 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3749 it should not have, which might have caused a crash in the right
3750 circumstances, but probably never did.
3752 PH/04 Installed a modified version of Tony Finch's patch to make submission
3753 mode fix the return path as well as the Sender: header line, and to
3754 add a /name= option so that you can make the user's friendly name appear
3757 TF/03 Added the control = fakedefer ACL modifier.
3759 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3760 Mark Lowes for thorough testing.
3762 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3764 TK/03 Merged latest SRS patch from Miles Wilton.
3766 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3767 with the definition in sysexits.h (which is #included earlier).
3768 Fortunately, Exim does not actually use EX_OK. The code used to try to
3769 preserve the sysexits.h value, by assuming that macro definitions were
3770 scanned for macro replacements. I have been disabused of this notion,
3771 so now the code just undefines EX_OK before #including unistd.h.
3773 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3774 in the smtp transport. When a block could not be written in a single
3775 write() function, the timeout was being re-applied to each part-write.
3776 This seems wrong - if the receiver was accepting one byte at a time it
3777 would take for ever. The timeout is now adjusted when this happens. It
3778 doesn't have to be particularly precise.
3780 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3781 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3783 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3784 without a domain if the check to <postmaster@domain> fails.
3786 SC/01 Eximstats: added -xls and the ability to specify output files
3787 (patch written by Frank Heydlauf).
3789 SC/02 Eximstats: use FileHandles for outputting results.
3791 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3793 SC/04 Eximstats: fixed display of large numbers with -nvr option
3795 SC/05 Eximstats: fixed merging of reports with empty tables.
3797 SC/06 Eximstats: added the -include_original_destination flag
3799 SC/07 Eximstats: removed tabs and trailing whitespace.
3801 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3803 TK/06 MBOX spool code: Add real "From " MBOX separator line
3804 so the .eml file is really in mbox format (even though
3805 most programs do not really care). Patch from Alex Miller.
3807 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3808 The latter is generated from $received_to and is only set if the
3809 message has one envelope recipient. SA can use these headers,
3810 obviously out-of-the-box. Patch from Alex Miller.
3812 PH/08 The ${def test on a variable was returning false if the variable's
3813 value was "0", contrary to what the specification has always said!
3814 The result should be true unless the variable is empty.
3816 PH/09 The syntax error of a character other than { following "${if
3817 def:variable_name" (after optional whitespace) was not being diagnosed.
3818 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3819 accidental colon was present, for example, could give incorrect results.
3821 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3822 result is used (not including appendfile, where other changes are about
3825 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3826 This involved changing a lot of size variables from int to off_t. It
3827 should work with maildirs and everything.
3829 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3830 spamd dying while we are connected to it.
3832 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3835 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3836 function for BDB 4.3.
3838 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3840 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3841 the sources for PCRE have been re-arranged and more files are now
3844 PH/15 The code I had for printing potentially long long variables in PH/11
3845 above was not the best (it lost precision). The length of off_t variables
3846 is now inspected at build time, and an appropriate printing format (%ld
3847 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3848 to be "long long int" or "long int". This is needed for the internal
3849 formatting function string_vformat().
3851 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3852 the configuration file to be ":syslog", then the script "guesses" where
3853 the logs files are, rather than using the compiled in default. In our
3854 case the guess is not the same as the compiled default, so the script
3855 suddenly stopped working when I started to use syslog. The patch checks
3856 to see if log_file_path is "". If so, it attempts to read it from exim
3857 with no configuration file to get the compiled in version, before it
3858 falls back to the previous guessing code."
3860 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3861 implementing BATV in an Exim configuration. See NewStuff for the gory
3864 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3865 Makefile that are specific to HP-UX.
3867 PH/18 If the "use_postmaster" option was set for a recipient callout together
3868 with the "random" option, the postmaster address was used as the MAIL
3869 FROM address for the random test, but not for the subsequent recipient
3870 test. It is now used for both.
3872 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3873 patch removes a few documentation additions to RFC 3028, because the
3874 latest draft now contains them. It adds the new en;ascii-case comparator
3875 and a new error check for 8bit text in MIME parts. Comparator and
3876 require names are now matched exactly. I enabled the subaddress
3877 extension, but it is not well tested yet (read: it works for me)."
3879 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3880 rework some of the code of TK/09 above to avoid the hardwired use of
3881 "%lld" and "long long". Replaced the call to snprintf() with a call to
3884 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3885 records point to non-existent hosts", "retry timeout exceeded", and
3886 "retry time not reached for any host after a long failure period".
3888 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3889 experimental DomainKeys support:
3891 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3892 (2) On an error such as an illegally used "control", the wrong name for
3893 the control was given.
3895 These problems did NOT occur unless DomainKeys support was compiled.
3897 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3899 PH/24 Added ${if match_ip condition.
3901 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3902 databases so that it will be absolutely obvious if a crash occurs in the
3903 DB library. This is a regular occurrence (often caused by mis-matched
3906 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3907 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3908 picked up on a user's system that detects such things. There doesn't seem
3909 to be a gcc warning option for this - only an attribute that has to be
3910 put on the function's prototype. It seems that in Fedora Core 4 they have
3911 set this on a number of new functions. No doubt there will be more in due
3914 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3915 specify a transport. However, if an address that was verified by such a
3916 router was the subject of a callout, Exim crashed because it tried to
3917 read the rcpt_include_affixes from the non-existent transport. Now it
3918 just assumes that the setting of that option is false. This bug was
3919 introduced by 4.51/PH/31.
3921 PH/28 Changed -d+all to exclude +memory, because that information is very
3922 rarely of interest, but it makes the output a lot bigger. People tend to
3923 do -d+all out of habit.
3925 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3926 code in os-type was giving problems when libc.so lives in lib64, like on
3929 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3930 aren't the modern standard, and it seems that some systems' include files
3931 don't always have them. Exim was already checking for some of the newer
3932 ones like T_AAAA, and defining it itself. I've added checks for all the
3933 record types that Exim uses.
3935 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3936 not automatically generating a new one, as it is supposed to. This
3937 prevented TLS from working. If the file did exist, but contained invalid
3938 data, a new version was generated, as expected. It was only the case of a
3939 non-existent file that was broken.
3941 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3942 with a change in libdomainkeys > 0.64.
3944 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3945 from DNS. If the selector record carries the flag, it now has
3946 precedence over the domain-wide flag.
3948 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3950 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3951 the use of a port name that isn't defined in /etc/services) occurred, the
3952 message was deferred as in a normal delivery, and thus remained on the
3953 spool, instead of being failed because of the mua_wrapper setting. This
3954 is now fixed, and I tidied up some of the mua_wrapper messages at the
3957 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3958 the messages in a hash of arrays rather than using individual hashes.
3959 This is a bit cleaner and results in dramatic memory savings, albeit
3960 at a slight CPU cost.
3962 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3963 as requested by Marc Sherman.
3965 SC/10 Eximstats: added histograms for user specified patterns as requested
3968 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3970 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3971 fopen() in the content-scanning modules that did not already have it.
3977 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3978 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3980 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3982 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3984 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3985 to test Sieve filters that use "vacation".
3987 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3988 that changes the way the GnuTLS parameters are stored in the cache file.
3989 The new format can be generated externally. For backward compatibility,
3990 if the data in the cache doesn't make sense, Exim assumes it has read an
3991 old-format file, and it generates new data and writes a new file. This
3992 means that you can't go back to an older release without removing the
3995 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3996 work if there are any delivery delays because "one_time" forces the
3997 parent to be marked "delivered", so its unseen clone is never tried
3998 again. For this reason, Exim now forbids the simultaneous setting of
4001 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4002 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4003 entry if you want to know the details. The fix, however, neglected to
4004 consider the case when local delivery batching is involved. The test for
4005 "previously delivered" was not happening when checking to see if an
4006 address could be batched with a previous (undelivered) one; under
4007 certain circumstances this could lead to multiple deliveries to the same
4010 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4011 in its include files, and this causes problems building Exim.
4013 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4014 header_syntax) but Exim was just ignoring anything given after a slash.
4015 In particular, this caused confusion with an attempt to use "verify =
4016 reverse_host_lookup/defer_ok". An error is now given when options are
4017 supplied for verify items that do not have them. (Maybe reverse_host_
4018 lookup should have a defer_ok option, but that's a different point.)
4020 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4021 defined by RFC 821) to 2048, because there were problems with some AUTH
4022 commands, and RFC 1869 says the size should be increased for extended
4023 SMTP commands that take arguments.
4025 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4028 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4029 "unknown" error; now it says that the functionality isn't in the binary.
4031 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4032 an address' error message when a string expansion fails (syntax or
4033 whatever). Otherwise the password may appear in the log. Following change
4034 PH/42 below, there is no longer a chance of it appearing in a bounce
4037 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4039 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4040 of its addresses was ever tried. (Bugzilla bug #2.)
4042 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4043 the result incorrectly in the debug output. (It correctly added a newline
4044 to what was transported.)
4046 TF/01 Added $received_time.
4048 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4049 commented out examples of how to interface to a virus scanner and to
4050 SpamAssassin. Also added commented examples of av_scanner and
4051 spamd_address settings.
4053 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4054 and controls are allowed in which ACLs. There were a couple of minor
4055 errors. Some of the entries in the conditions table (which is a table of
4056 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4057 negation of where the condition IS allowed.
4059 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4061 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4062 header file does not have a version number, so I've had to invent a new
4063 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4064 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4065 radiusclient), but it was contributed by a Radius user.
4067 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4068 files or directories, for OpenSSL.
4070 PH/19 When an Exim process that is running as root has to create an Exim log
4071 file, it does so in a subprocess that runs as exim:exim so as to get the
4072 ownership right at creation (otherwise, other Exim processes might see
4073 the file with the wrong ownership). There was no test for failure of this
4074 fork() call, which would lead to the process getting stuck as it waited
4075 for a non-existent subprocess. Forks do occasionally fail when resources
4076 run out. I reviewed all the other calls to fork(); they all seem to check
4079 PH/20 When checking for unexpected SMTP input at connect time (before writing
4080 the banner), Exim was not dealing correctly with a non-positive return
4081 from the read() function. If the client had disconnected by this time,
4082 the result was a log entry for a synchronization error with an empty
4083 string after "input=" when read() returned zero. If read() returned -1
4084 (an event I could not check), uninitialized data bytes were printed.
4085 There were reports of junk text (parts of files, etc) appearing after
4088 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4090 PH/22 Added support for macro redefinition, and (re)definition in between
4091 driver and ACL definitions.
4093 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4094 forgetting to use the resulting value; it was using the unexpanded value.
4096 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4097 hadn't been configured. The fix is from Juergen Kreileder, who
4098 understands it better than I do:
4100 "Here's what I see happening with three configured cyrus_sasl
4101 authenticators configured (plain, login, cram-md5):
4103 On startup auth_cyrus_sasl_init() gets called for each of these.
4104 This means three calls to sasl_listmech() without a specified mech_list.
4105 => SASL tests which mechs of all available mechs actually work
4106 => three warnings about OTP not working
4107 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4109 With the patch, sasl_listmech() also gets called three times. But now
4110 SASL's mech_list option is set to the server_mech specified in the the
4111 authenticator. Or in other words, the answer from sasl_listmech()
4112 gets limited to just the mech you're testing for (which is different
4114 => the return list contains just 'plain' or 'login', 'cram-md5' or
4115 nothing depending on the value of ob->server_mech.
4117 I've just tested the patch: Authentication still works fine,
4118 unavailable mechs specified in the exim configuration are still
4119 caught, and the auth.log warnings about OTP are gone."
4121 PH/25 When debugging is enabled, the contents of the command line are added
4122 to the debugging output, even when log_selector=+arguments is not
4125 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4126 answer is "GNU", and only if the return is "GNU/something" is the answer
4129 PH/27 $acl_verify_message is now set immediately after the failure of a
4130 verification in an ACL, and so is available in subsequent modifiers. In
4131 particular, the message can be preserved by coding like this:
4133 warn !verify = sender
4134 set acl_m0 = $acl_verify_message
4136 Previously, $acl_verify_message was set only while expanding "message"
4137 and "log_message" when a very denied access.
4139 PH/28 Modified OS/os.c-Linux with
4141 -#ifndef OS_LOAD_AVERAGE
4142 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4144 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4145 nomenclature these days.)
4147 PH/29 Installed patch from the Sieve maintainer that adds the options
4148 sieve_useraddress and sieve_subaddress to the redirect router.
4150 PH/30 In these circumstances:
4151 . Two addresses routed to the same list of hosts;
4152 . First host does not offer TLS;
4153 . First host accepts first address;
4154 . First host gives temporary error to second address;
4155 . Second host offers TLS and a TLS session is established;
4156 . Second host accepts second address.
4157 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4158 and peerdn, if requested) that were in fact used only for the second
4161 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4162 attention to any local part prefix or suffix that was matched by the
4163 router that accepted the address. It now behaves in the same way as it
4164 does for delivery: the affixes are removed from the local part unless
4165 rcpt_include_affixes is set on the transport.
4167 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4168 timeout during the DATA phase of an incoming message.
4170 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4171 applied a patch sanctioned by the Sieve maintainer.
4173 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4174 the uid or gid is negative. A case of a negative gid caused this to be
4175 noticed. The fix allows for either to be negative.
4177 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4178 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4181 PH/36 The change PH/12 above was broken. Fixed it.
4183 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4184 the grounds that routing the same address twice would always produce the
4185 same answer. This might have been true once, but it is certainly no
4186 longer true now. Routing a child address may depend on the previous
4187 routing that produced that child. Some complicated redirection strategies
4188 went wrong when messages had multiple recipients, and made Exim's
4189 behaviour dependent on the order in which the addresses were given.
4191 I have moved the duplicate checking until after the routing is complete.
4192 Exim scans the addresses that are assigned to local and remote
4193 transports, and removes any duplicates. This means that more work will be
4194 done, as duplicates will always all be routed, but duplicates are
4195 presumably rare, so I don't expect this is of any significance.
4197 For deliveries to pipes, files, and autoreplies, the duplicate checking
4198 still happens during the routing process, since they are not going to be
4201 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4202 It corrects a timeout issue with spamd. This is Ian's comment: "The
4203 background is that sometimes spamd either never reads data from a
4204 connection it has accepted, or it never writes response data. The exiscan
4205 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4206 blindly assumes that writes won't block so it may never time out."
4208 PH/39 Allow G after quota size as well as K and M.
4210 PH/40 The value set for $authenticated_id in an authenticator may not contain
4211 binary zeroes or newlines because the value is written to log lines and
4212 to spool files. There was no check on this. Now the value is run through
4213 the string_printing() function so that such characters are converted to
4214 printable escape sequences.
4216 PH/41 $message_linecount is a new variable that contains the total number of
4217 lines in the message. Compare $body_linecount, which is the count for the
4220 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4221 bounce and delay warning messages, except in certain special cases, which
4224 (a) An SMTP error message from a remote host;
4225 (b) A message specified in a :fail: redirection;
4226 (c) A message specified in a "fail" command in a system filter;
4227 (d) A message specified in a FAIL return from the queryprogram router;
4228 (e) A message specified by the cannot_route_message router option.
4230 In these cases only, Exim does include the error details in bounce and
4231 warning messages. There are also a few cases where bland messages such
4232 as "unrouteable address" or "local delivery error" are given.
4234 PH/43 $value is now also set for the "else" part of a ${run expansion.
4236 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4237 being worked on, but at least Exim now implements the latest version to
4240 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4241 process to complete was treated as a delivery failure, a timeout while
4242 writing the message to the pipe was logged, but erroneously treated as a
4243 successful delivery. Such timeouts include transport filter timeouts. For
4244 consistency with the overall process timeout, these timeouts are now
4245 treated as errors, giving rise to delivery failures by default. However,
4246 there is now a new Boolean option for the pipe transport called
4247 timeout_defer, which, if set TRUE, converts the failures into defers for
4248 both kinds of timeout. A transport filter timeout is now identified in
4251 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4252 systems where "make" and "gmake" are different, calling "gmake" at top
4253 level broke things. I've arranged for the value of $(MAKE) to be passed
4254 from the Makefile to this script so that it can call the same version of
4258 A note about Exim versions 4.44 and 4.50
4259 ----------------------------------------
4261 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4262 changes of various kinds. As a consequence, a big documentation update was
4263 needed. This delayed the release for rather longer than seemed good, especially
4264 in the light of a couple of (minor) security issues. Therefore, the changes
4265 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4266 release. So 4.44 and 4.50 are in effect two different branches that both start
4269 I have left the 4.50 change log unchanged; it contains all the changes since
4270 4.43. The change log for 4.44 is below; many of its items are identical to
4271 those for 4.50. This seems to be the most sensible way to preserve the
4272 historical information.
4278 1. Minor wording change to the doc/README.SIEVE file.
4280 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4281 computation of the current number of files was incorrect.
4283 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4284 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4287 4. Give more explanation in the error message when the command for a transport
4288 filter fails to execute.
4290 5. There are several places where Exim runs a non-Exim command in a
4291 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4292 was being done only for the command run by the queryprogram router. It is
4293 now done for all such subprocesses. The other cases are: ${run, transport
4294 filters, and the commands run by the lmtp and pipe transports.
4296 6. Added CONFIGURE_GROUP build-time option.
4298 7. Some older OS have a limit of 256 on the maximum number of file
4299 descriptors. Exim was using setrlimit() to set 1000 as a large value
4300 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4301 systems. I've change it so that if it can't get 1000, it tries for 256.
4303 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4304 was an oversight, and furthermore, ever since the addition of extra
4305 controls (e.g. 4.43/32), the checks on when to allow different forms of
4306 "control" were broken. There should now be diagnostics for all cases when a
4307 control that does not make sense is encountered.
4309 9. Added the /retain_sender option to "control=submission".
4311 10. $recipients is now available in the predata ACL (oversight).
4313 11. Tidy the search cache before the fork to do a delivery from a message
4314 received from the command line. Otherwise the child will trigger a lookup
4315 failure and thereby defer the delivery if it tries to use (for example) a
4316 cached ldap connection that the parent has called unbind on.
4318 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4319 of $address_data from the recipient verification was clobbered by the
4320 sender verification.
4322 13. The value of address_data from a sender verification is now available in
4323 $sender_address_data in subsequent conditions in the ACL statement.
4325 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4327 15. Added a new option "connect=<time>" to callout options, to set a different
4330 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4331 was its contents. (It was OK if the option was not defined at all.)
4333 17. A "Completed" log line is now written for messages that are removed from
4334 the spool by the -Mrm option.
4336 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4337 information about exactly what failed.
4339 19. Added -dd to debug only the daemon process.
4341 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4342 handles timeouts, both on the server side and network timeouts. Renamed the
4343 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4345 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4346 It is now set to "smtps".
4348 22. $host_address is now set to the target address during the checking of
4349 ignore_target_hosts.
4351 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4352 being passed; this would have caused $sender_host_name to have been used if
4353 matching the list had actually called for a host name (not very likely,
4354 since this list is usually IP addresses). A host name is now passed as
4357 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4358 code that specifies a non-restarting handler (typically sigaction() in
4359 modern systems) in an attempt to fix a rare and obscure crash bug.
4361 25. Narrowed the window for a race in the daemon that could cause it to ignore
4362 SIGCHLD signals. This is not a major problem, because they are used only to
4363 wake it up if nothing else does.
4365 26. A malformed maildirsize file could cause Exim to calculate negative values
4366 for the mailbox size or file count. Odd effects could occur as a result.
4367 The maildirsize information is now recalculated if the size or filecount
4370 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4371 support for a long time. Removed HAVE_SYS_VFS_H.
4373 28. Installed the latest version of exipick from John Jetmore.
4375 29. In an address list, if the pattern was not a regular expression, an empty
4376 subject address (from a bounce message) matched only if the pattern was an
4377 empty string. Non-empty patterns were not even tested. This was the wrong
4378 because it is perfectly reasonable to use an empty address as part of a
4379 database query. An empty address is now tested by patterns that are
4380 lookups. However, all the other forms of pattern expect the subject to
4381 contain a local part and a domain, and therefore, for them, an empty
4382 address still always fails if the pattern is not itself empty.
4384 30. Exim went into a mad DNS loop when attempting to do a callout where the
4385 host was specified on an smtp transport, and looking it up yielded more
4386 than one IP address.
4388 31. Re-factored the code for checking spool and log partition space into a
4389 function that finds that data and another that does the check. The former
4390 is then used to implement four new variables: $spool_space, $log_space,
4391 $spool_inodes, and $log_inodes.
4393 32. The RFC2047 encoding function was originally intended for short strings
4394 such as real names; it was not keeping to the 75-character limit for
4395 encoded words that the RFC imposes. It now respects the limit, and
4396 generates multiple encoded words if necessary. To be on the safe side, I
4397 have increased the buffer size for the ${rfc2047: expansion operator from
4400 33. It is now permitted to omit both strings after an "if" condition; if the
4401 condition is true, the result is "true". As before, when the second string
4402 is omitted, a false condition yields an empty string. This makes it less
4403 cumbersome to write custom ACL and router conditions.
4405 34. Failure to deliver a bounce message always caused it to be frozen, even if
4406 there was an errors_to setting on the router. The errors_to setting is now
4409 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4410 canonical form (fully expanded) before being placed in
4411 $sender_host_address.
4413 36. The table in the code that translates DNS record types into text (T_A to
4414 "A" for instance) was missing entries for NS and CNAME. It is just possible
4415 that this could have caused confusion if both these types were looked up
4416 for the same domain, because the text type is used as part of Exim's
4417 per-process caching. But the chance of anyone hitting this buglet seems
4420 37. The dnsdb lookup has been extended in a number of ways.
4422 (1) There is a new type, "zns", which walks up the domain tree until it
4423 finds some nameserver records. It should be used with care.
4425 (2) There is a new type, "mxh", which is like "mx" except that it returns
4426 just the host names, not the priorities.
4428 (3) It is now possible to give a list of domains (or IP addresses) to be
4429 looked up. The behaviour when one of the lookups defers can be
4430 controlled by a keyword.
4432 (4) It is now possible to specify the separator character for use when
4433 multiple records are returned.
4435 38. The dnslists ACL condition has been extended: it is now possible to supply
4436 a list of IP addresses and/or domains to be looked up in a particular DNS
4439 39. Added log_selector=+queue_time_overall.
4441 40. When running the queue in the test harness, wait just a tad after forking a
4442 delivery process, to get repeatability of debugging output.
4444 41. Include certificate and key file names in error message when GnuTLS fails
4445 to set them up, because the GnuTLS error message doesn't include the name
4446 of the failing file when there is a problem reading it.
4448 42. Allow both -bf and -bF in the same test run.
4450 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4452 44. The "Exiscan patch" is now merged into the mainline Exim source.
4454 45. Sometimes the final signoff response after QUIT could fail to get
4455 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4456 < 0 before doing a fflush(). This bug looks as though it goes back to the
4457 introduction of TLS in release 3.20, but "sometimes" must have been rare
4458 because the tests only now provoked it.
4460 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4461 (this can affect the format of dates).
4463 47. exim_tidydb, when checking for the continued existence of a message for
4464 which it has found a message-specific retry record, was not finding
4465 messages that were in split spool directories. Consequently, it was
4466 deleting retry records that should have stayed in existence.
4468 48. Steve fixed some bugs in eximstats.
4470 49. The SPA authentication driver was not abandoning authentication and moving
4471 on to the next authenticator when an expansion was forced to fail,
4472 contradicting the general specification for all authenticators. Instead it
4473 was generating a temporary error. It now behaves as specified.
4475 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4476 (the order specifies the preference for clients). The order is now AES256,
4477 AES128, 3DES, ARCFOUR128.
4479 51. Small patch to Sieve code - explicitly set From: when generating an
4482 52. Exim crashed if a remote delivery caused a very long error message to be
4483 recorded - for instance if somebody sent an entire SpamAssassin report back
4484 as a large number of 550 error lines. This bug was coincidentally fixed by
4485 increasing the size of one of Exim's internal buffers (big_buffer) that
4486 happened as part of the Exiscan merge. However, to be on the safe side, I
4487 have made the code more robust (and fixed the comments that describe what
4490 53. Now that there can be additional text after "Completed" in log lines (if
4491 the queue_time_overall log selector is set), a one-byte patch to exigrep
4492 was needed to allow it to recognize "Completed" as not the last thing in
4495 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4496 patch that reportedly fixes this has been added. I am not expert enough to
4497 create a test for it. This is what the patch creator wrote:
4499 "I found a little strange behaviour of ldap code when working with
4500 Windows 2003 AD Domain, where users was placed in more than one
4501 Organization Units. When I tried to give exim partial DN, the exit code
4502 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4503 But simultaneously result of request was absolutely normal ldap result,
4504 so I produce this patch..."
4506 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4507 REFERENCE, so I have modified the code to exclude the patch when that macro
4510 55. Some experimental protocols are using DNS PTR records for new purposes. The
4511 keys for these records are domain names, not reversed IP addresses. The
4512 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4513 leaves it alone. Component reversal etc. now happens only for IP addresses.
4516 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4518 57. Double the size of the debug message buffer (to 2048) so that more of very
4519 long debug lines gets shown.
4521 58. The exicyclog utility now does better if the number of log files to keep
4522 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4524 59. Two changes related to the smtp_active_hostname option:
4526 (1) $smtp_active_hostname is now available as a variable.
4527 (2) The default for smtp_banner uses $smtp_active_hostname instead
4528 of $primary_hostname.
4530 60. The host_aton() function is supposed to be passed a string that is known
4531 to be a valid IP address. However, in the case of IPv6 addresses, it was
4532 not checking this. This is a hostage to fortune. Exim now panics and dies
4533 if the condition is not met. A case was found where this could be provoked
4534 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4535 components; fortuitously, this particular loophole had already been fixed
4536 by change 4.50/55 above.
4538 If there are any other similar loopholes, the new check in host_aton()
4539 itself should stop them being exploited. The report I received stated that
4540 data on the command line could provoke the exploit when Exim was running as
4541 exim, but did not say which command line option was involved. All I could
4542 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4543 running as the user.
4546 61. There was a buffer overflow vulnerability in the SPA authentication code
4547 (which came originally from the Samba project). I have added a test to the
4548 spa_base64_to_bits() function which I hope fixes it.
4551 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4552 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4554 63. The daemon start-up calls getloadavg() while still root for those OS that
4555 need the first call to be done as root, but it missed one case: when
4556 deliver_queue_load_max is set with deliver_drop_privilege. This is
4557 necessary for the benefit of the queue runner, because there is no re-exec
4558 when deliver_drop_privilege is set.
4560 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4561 This has been fixed.
4563 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4564 was in use, was not putting the data itself into the right store pool;
4565 consequently, it could be overwritten for a subsequent message in the same
4566 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4569 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4571 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4572 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4573 the function were treating the return as a boolean value, which happened to
4574 work because 0=false and not-0=true, but is not correct code.
4576 68. The host_aton() function was not handling scoped IPv6 addresses (those
4577 with, for example, "%eth0" on the end) correctly.
4579 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4580 negated items (that is, ~something) in unsigned ints. Some compilers
4581 apparently mutter when there is no cast.
4583 70. If an address verification called from an ACL failed, and did not produce a
4584 user-specific message (i.e. there was only a "system" message), nothing was
4585 put in $acl_verify_message. In this situation, it now puts the system
4588 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4589 session; change 4.31/43 added the unwanted input to the log line - except
4590 that it did not do this in the start of session case. It now does.
4592 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4593 This is wrong and can cause the other end to generate a synchronization
4594 error if it is another Exim or anything else that does the synchronization
4595 check. A QUIT command is no longer sent after a timeout.
4597 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4598 during host lookups.
4600 74. The defer_ok option of callout verification was not working if it was used
4601 when verifying addresses in header lines, that is, for this case:
4603 verify = header_sender/callout=defer_ok
4605 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4606 those file descriptors could be used for SMTP connections. If anything
4607 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4608 could be sent to the SMTP client, causing chaos. The daemon now opens
4609 stdin, stdout, and stderr to /dev/null when it puts itself into the
4612 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4613 log by default. The user can override this with suitable Perl magic.
4615 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4616 the log message when discard triggers, was not working for the DATA ACL or
4617 for the non-SMTP ACL.
4619 78. Error message wording change in sieve.c.
4621 79. If smtp_accept_max_per_host was set, the number of connections could be
4622 restricted to fewer than expected, because the daemon was trying to set up
4623 a new connection before checking whether the processes handling previous
4624 connections had finished. The check for completed processes is now done
4625 earlier. On busy systems, this bug wouldn't be noticed because something
4626 else would have woken the daemon, and it would have reaped the completed
4629 80. If a message was submitted locally by a user whose login name contained one
4630 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4631 It caused a spool format error. I have fixed the spool reading code. A
4632 related problem was that the "from" clause in the Received: line became
4633 illegal because of the space(s). It is now covered by ${quote_local_part.
4635 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4638 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4640 83. Patch from Sieve maintainer to fix unterminated string problem in
4641 "vacation" handling.
4643 84. Some minor changes to the Linux configuration files to help with other
4644 OS variants using glibc.
4646 85. One more patch for Sieve to update vacation handling to latest spec.
4649 ----------------------------------------------------
4650 See the note above about the 4.44 and 4.50 releases.
4651 ----------------------------------------------------
4657 1. Change 4.43/35 introduced a bug that caused file counts to be
4658 incorrectly computed when quota_filecount was set in an appendfile
4661 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4662 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4665 3. Give more explanation in the error message when the command for a transport
4666 filter fails to execute.
4668 4. There are several places where Exim runs a non-Exim command in a
4669 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4670 was being done only for the command run by the queryprogram router. It is
4671 now done for all such subprocesses. The other cases are: ${run, transport
4672 filters, and the commands run by the lmtp and pipe transports.
4674 5. Some older OS have a limit of 256 on the maximum number of file
4675 descriptors. Exim was using setrlimit() to set 1000 as a large value
4676 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4677 systems. I've change it so that if it can't get 1000, it tries for 256.
4679 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4680 was an oversight, and furthermore, ever since the addition of extra
4681 controls (e.g. 4.43/32), the checks on when to allow different forms of
4682 "control" were broken. There should now be diagnostics for all cases when a
4683 control that does not make sense is encountered.
4685 7. $recipients is now available in the predata ACL (oversight).
4687 8. Tidy the search cache before the fork to do a delivery from a message
4688 received from the command line. Otherwise the child will trigger a lookup
4689 failure and thereby defer the delivery if it tries to use (for example) a
4690 cached ldap connection that the parent has called unbind on.
4692 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4693 of $address_data from the recipient verification was clobbered by the
4694 sender verification.
4696 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4697 was its contents. (It was OK if the option was not defined at all.)
4699 11. A "Completed" log line is now written for messages that are removed from
4700 the spool by the -Mrm option.
4702 12. $host_address is now set to the target address during the checking of
4703 ignore_target_hosts.
4705 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4706 being passed; this would have caused $sender_host_name to have been used if
4707 matching the list had actually called for a host name (not very likely,
4708 since this list is usually IP addresses). A host name is now passed as
4711 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4712 code that specifies a non-restarting handler (typically sigaction() in
4713 modern systems) in an attempt to fix a rare and obscure crash bug.
4715 15. Narrowed the window for a race in the daemon that could cause it to ignore
4716 SIGCHLD signals. This is not a major problem, because they are used only to
4717 wake it up if nothing else does.
4719 16. A malformed maildirsize file could cause Exim to calculate negative values
4720 for the mailbox size or file count. Odd effects could occur as a result.
4721 The maildirsize information is now recalculated if the size or filecount
4724 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4725 support for a long time. Removed HAVE_SYS_VFS_H.
4727 18. Updated exipick to current release from John Jetmore.
4729 19. Allow an empty sender to be matched against a lookup in an address list.
4730 Previously the only cases considered were a regular expression, or an
4733 20. Exim went into a mad DNS lookup loop when doing a callout where the
4734 host was specified on the transport, if the DNS lookup yielded more than
4737 21. The RFC2047 encoding function was originally intended for short strings
4738 such as real names; it was not keeping to the 75-character limit for
4739 encoded words that the RFC imposes. It now respects the limit, and
4740 generates multiple encoded words if necessary. To be on the safe side, I
4741 have increased the buffer size for the ${rfc2047: expansion operator from
4744 22. Failure to deliver a bounce message always caused it to be frozen, even if
4745 there was an errors_to setting on the router. The errors_to setting is now
4748 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4749 canonical form (fully expanded) before being placed in
4750 $sender_host_address.
4752 24. Updated eximstats to version 1.33
4754 25. Include certificate and key file names in error message when GnuTLS fails
4755 to set them up, because the GnuTLS error message doesn't include the name
4756 of the failing file when there is a problem reading it.
4758 26. Expand error message when OpenSSL has problems setting up cert/key files.
4761 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4762 (this can affect the format of dates).
4764 28. exim_tidydb, when checking for the continued existence of a message for
4765 which it has found a message-specific retry record, was not finding
4766 messages that were in split spool directories. Consequently, it was
4767 deleting retry records that should have stayed in existence.
4769 29. eximstats updated to version 1.35
4770 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4771 1.35 - bugfix such that pie charts by volume are generated correctly
4773 30. The SPA authentication driver was not abandoning authentication and moving
4774 on to the next authenticator when an expansion was forced to fail,
4775 contradicting the general specification for all authenticators. Instead it
4776 was generating a temporary error. It now behaves as specified.
4778 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4779 (the order specifies the preference for clients). The order is now AES256,
4780 AES128, 3DES, ARCFOUR128.
4782 31. Small patch to Sieve code - explicitly set From: when generating an
4785 32. Exim crashed if a remote delivery caused a very long error message to be
4786 recorded - for instance if somebody sent an entire SpamAssassin report back
4787 as a large number of 550 error lines. This bug was coincidentally fixed by
4788 increasing the size of one of Exim's internal buffers (big_buffer) that
4789 happened as part of the Exiscan merge. However, to be on the safe side, I
4790 have made the code more robust (and fixed the comments that describe what
4793 33. Some experimental protocols are using DNS PTR records for new purposes. The
4794 keys for these records are domain names, not reversed IP addresses. The
4795 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4796 leaves it alone. Component reversal etc. now happens only for IP addresses.
4799 34. The host_aton() function is supposed to be passed a string that is known
4800 to be a valid IP address. However, in the case of IPv6 addresses, it was
4801 not checking this. This is a hostage to fortune. Exim now panics and dies
4802 if the condition is not met. A case was found where this could be provoked
4803 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4804 components; fortuitously, this particular loophole had already been fixed
4805 by change 4.50/55 or 4.44/33 above.
4807 If there are any other similar loopholes, the new check in host_aton()
4808 itself should stop them being exploited. The report I received stated that
4809 data on the command line could provoke the exploit when Exim was running as
4810 exim, but did not say which command line option was involved. All I could
4811 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4812 running as the user.
4815 35. There was a buffer overflow vulnerability in the SPA authentication code
4816 (which came originally from the Samba project). I have added a test to the
4817 spa_base64_to_bits() function which I hope fixes it.
4820 36. The daemon start-up calls getloadavg() while still root for those OS that
4821 need the first call to be done as root, but it missed one case: when
4822 deliver_queue_load_max is set with deliver_drop_privilege. This is
4823 necessary for the benefit of the queue runner, because there is no re-exec
4824 when deliver_drop_privilege is set.
4826 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4827 was in use, was not putting the data itself into the right store pool;
4828 consequently, it could be overwritten for a subsequent message in the same
4829 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4832 38. Sometimes the final signoff response after QUIT could fail to get
4833 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4834 < 0 before doing a fflush(). This bug looks as though it goes back to the
4835 introduction of TLS in release 3.20, but "sometimes" must have been rare
4836 because the tests only now provoked it.
4842 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4843 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4844 Now it returns an integer. A number of calls were still expecting a T/F
4845 return. Fortuitously, in all cases, the tests worked in OK situations,
4846 which is the norm. However, things would have gone wrong on any write
4847 failures on the smtp file descriptor. This function is used when sending
4848 messages over SMTP and also when doing verify callouts.
4850 2. When Exim is called to do synchronous delivery of a locally submitted
4851 message (the -odf or -odi options), it no longer closes stderr before doing
4854 3. Implemented the mua_wrapper option.
4856 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4858 5. Implemented the functions header_remove(), header_testname(),
4859 header_add_at_position(), and receive_remove_recipient(), and exported them
4862 6. If an ACL "warn" statement specified the addition of headers, Exim already
4863 inserted X-ACL-Warn: at the start if there was no header name. However, it
4864 was not making this test for the second and subsequent header lines if
4865 there were newlines in the string. This meant that an invalid header could
4866 be inserted if Exim was badly configured.
4868 7. Allow an ACL "warn" statement to add header lines at the start or after all
4869 the Received: headers, as well as at the end.
4871 8. Added the rcpt_4xx retry error code.
4873 9. Added postmaster_mailfrom=xxx to callout verification option.
4875 10. Added mailfrom=xxxx to the callout verification option, for verify=
4878 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4879 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4881 12. Inserted some casts to stop certain compilers warning when using pointer
4882 differences as field lengths or precisions in printf-type calls (mostly
4883 affecting debugging statements).
4885 13. Added optional readline() support for -be (dynamically loaded).
4887 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4888 same clock tick as a message's arrival, so that its received time was the
4889 same as the "first fail" time on the retry record, and that message
4890 remained on the queue past the ultimate address timeout, every queue runner
4891 would try a delivery (because it was past the ultimate address timeout) but
4892 after another failure, the ultimate address timeout, which should have then
4893 bounced the address, did not kick in. This was a "< instead of <=" error;
4894 in most cases the first failure would have been in the next clock tick
4895 after the received time, and all would be well.
4897 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4898 being recognized when the domain list was tested by the match_domain
4899 condition in an expansion string.
4901 16. Added the ${str2b64: operator.
4903 17. Exim was always calling setrlimit() to set a large limit for the number of
4904 processes, without checking whether the existing limit was already
4905 adequate. (It did check for the limit on file descriptors.) Furthermore,
4906 errors from getrlimit() and setrlimit() were being ignored. Now they are
4907 logged to the main and panic logs, but Exim does carry on, to try to do its
4908 job under whatever limits there are.
4910 18. Imported PCRE 5.0.
4912 19. Trivial typo in log message " temporarily refused connection" (the leading
4915 20. If the log selector return_path_on_delivery was set and an address was
4916 redirected to /dev/null, the delivery process crashed because it assumed
4917 that a return path would always be set for a "successful" delivery. In this
4918 case, the whole delivery is bypassed as an optimization, and therefore no
4921 21. Internal re-arrangement: the function for sending a challenge and reading
4922 a response while authentication was assuming a zero-terminated challenge
4923 string. It's now changed to take a pointer and a length, to allow for
4924 binary data in such strings.
4926 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4928 23. Exim was not respecting finduser_retries when seeking the login of the
4929 uid under which it was called; it was always trying 10 times. (The default
4930 setting of finduser_retries is zero.) Also, it was sleeping after the final
4931 failure, which is pointless.
4933 24. Implemented tls_on_connect_ports.
4935 25. Implemented acl_smtp_predata.
4937 26. If the domain in control=submission is set empty, Exim assumes that the
4938 authenticated id is a complete email address when it generates From: or
4939 Sender: header lines.
4941 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4942 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4943 chown and chgrp in /bin and hostname in /usr/bin.
4945 28. Exim was keeping the "process log" file open after each use, just as it
4946 does for the main log. This opens the possibility of it remaining open for
4947 long periods when the USR1 signal hits a daemon. Occasional processlog
4948 errors were reported, that could have been caused by this. Anyway, it seems
4949 much more sensible not to leave this file open at all, so that is what now
4952 29. The long-running daemon process does not normally write to the log once it
4953 has entered its main loop, and it closes the log before doing so. This is
4954 so that log files can straightforwardly be renamed and moved. However,
4955 there are a couple of unusual error situations where the daemon does write
4956 log entries, and I had neglected to close the log afterwards.
4958 30. The text of an SMTP error response that was received during a remote
4959 delivery was being truncated at 512 bytes. This is too short for some of
4960 the long messages that one sometimes sees. I've increased the limit to
4963 31. It is now possible to make retry rules that apply only when a message has a
4964 specific sender, in particular, an empty sender.
4966 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4967 it possible to be selective about when SMTP synchronization is enforced.
4969 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4971 32. Implemented hosts_connection_nolog.
4973 33. Added an ACL for QUIT.
4975 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4978 35. Added mailbox_size and mailbox_filecount to appendfile.
4980 36. Added control = no_multiline_responses to ACLs.
4982 37. There was a bug in the logic of the code that waits for the clock to tick
4983 in the case where the clock went backwards by a substantial amount such
4984 that the microsecond fraction of "now" was more than the microsecond
4985 fraction of "then" (but the whole seconds number was less).
4987 38. Added support for the libradius Radius client library this is found on
4988 FreeBSD (previously only the radiusclient library was supported).
4994 1. When certain lookups returned multiple values in the form name=value, the
4995 quoting of the values was not always being done properly. Specifically:
4996 (a) If the value started with a double quote, but contained no whitespace,
4998 (b) If the value contained whitespace other than a space character (i.e.
4999 tabs or newlines or carriage returns) it was not quoted.
5000 This fix has been applied to the mysql and pgsql lookups by writing a
5001 separate quoting function and calling it from the lookup code. The fix
5002 should probably also be applied to nisplus, ibase and oracle lookups, but
5003 since I cannot test any of those, I have not disturbed their existing code.
5005 2. A hit in the callout cache for a specific address caused a log line with no
5006 reason for rejecting RCPT. Now it says "Previous (cached) callout
5007 verification failure".
5009 3. There was an off-by-one bug in the queryprogram router. An over-long
5010 return line was truncated at 256 instead of 255 characters, thereby
5011 overflowing its buffer with the terminating zero. As well as fixing this, I
5012 have increased the buffer size to 1024 (and made a note to document this).
5014 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5015 when Exim is waiting for an SMTP response from a remote server, Exim
5016 restarts its select() call on the socket, thereby resetting its timeout.
5017 This is not a problem when such interrupts are rare. Somebody set up a cron
5018 job to run exiwhat every 2 minutes, which is less than the normal select()
5019 timeout (5 or 10 minutes). This meant that the select() timeout never
5020 kicked in because it was always reset. I have fixed this by comparing the
5021 time when an interrupt arrives with the time at the start of the first call
5022 to select(). If more time than the timeout has elapsed, the interrupt is
5023 treated as a timeout.
5025 5. Some internal re-factoring in preparation for the addition of Sieve
5026 extensions (by MH). In particular, the "personal" test is moved to a
5027 separate function, and given an option for scanning Cc: and Bcc: (which is
5028 not set for Exim filters).
5030 6. When Exim created an email address using the login of the caller as the
5031 local part (e.g. when creating a From: or Sender: header line), it was not
5032 quoting the local part when it contained special characters such as @.
5034 7. Installed new OpenBSD configuration files.
5036 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5037 try to make them clearer.
5039 9. Callout options, other than the timeout value, were being ignored when
5040 verifying sender addresses in header lines. For example, when using
5042 verify = header_sender/callout=no_cache
5044 the cache was (incorrectly) being used.
5046 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5047 only the Cygwin environment.
5049 11. When return_path_on_delivery was set as a log selector, if different remote
5050 addresses in the same message used different return paths and parallel
5051 remote delivery occurred, the wrong values would sometimes be logged.
5052 (Whenever a remote delivery process finished, the return path value from
5053 the most recently started remote delivery process was logged.)
5055 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5056 lines when AUTH and/or TLS are in use. This is the "received protocol"
5057 field. Exim used to use "asmtp" for authenticated SMTP, without any
5058 indication (in the protocol name) for TLS use. Now it follows the RFC and
5059 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5060 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5061 names appear in log lines as well as in Received: header lines.
5063 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5064 extensions, and comparison tests, and to fix some bugs.
5066 14. Changes to the "personal" filter test:
5068 (1) The test was buggy in that it was just doing the equivalent of
5069 "contains" tests on header lines. For example, if a user's address was
5070 anne@some.where, the "personal" test would incorrectly be true for
5072 To: susanne@some.where
5074 This test is now done by extracting each address from the header in turn,
5075 and checking the entire address. Other tests that are part of "personal"
5076 are now done using regular expressions (for example, to check local parts
5077 of addresses in From: header lines).
5079 (2) The list of non-personal local parts in From: addresses has been
5080 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5081 taken from the Sieve specification recommendations.
5083 (3) If the message contains any header line starting with "List-" it is
5084 treated as non-personal.
5086 (4) The test for "circular" in the Subject: header line has been removed
5087 because it now seems ill-conceived.
5089 15. Minor typos in src/EDITME comments corrected.
5091 16. Installed latest exipick from John Jetmore.
5093 17. If headers_add on a router specified a text string that was too long for
5094 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5095 of string_sprintf() is now avoided.
5097 18. $message_body_size was not set (it was always zero) when running the DATA
5098 ACL and the local_scan() function.
5100 19. For the "mail" command in an Exim filter, no default was being set for
5101 the once_repeat time, causing a random time value to be used if "once" was
5102 specified. (If the value happened to be <= 0, no repeat happened.) The
5103 default is now 0s, meaning "never repeat". The "vacation" command was OK
5104 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5105 (I found it when inspecting the code).
5107 20. There is now an overall timeout for performing a callout verification. It
5108 defaults to 4 times the callout timeout, which applies to individual SMTP
5109 commands during the callout. The overall timeout applies when there is more
5110 than one host that can be tried. The timeout is checked before trying the
5111 next host. This prevents very long delays if there are a large number of
5112 hosts and all are timing out (e.g. when the network connections are timing
5113 out). The value of the overall timeout can be changed by specifying an
5114 additional sub-option for "callout", called "maxwait". For example:
5116 verify = sender/callout=5s,maxwait=20s
5118 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5119 to the end before writing, but this should make it even safer).
5121 22. Exim was forgetting that it had advertised PIPELINING for the second and
5122 subsequent messages on an SMTP connection. It was also not resetting its
5123 memory on STARTTLS and an internal HELO.
5125 23. When Exim logs an SMTP synchronization error within a session, it now
5126 records whether PIPELINING has been advertised or not.
5128 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5129 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5130 rather than long int.
5132 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5134 26. Added the never_mail option to autoreply.
5140 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5141 crash if the getsockname() call failed; this can happen if a connection is
5142 closed very soon after it is established. The problem was simply in the
5143 order in which certain operations were done, causing Exim to try to write
5144 to the SMTP stream before it had set up the file descriptor. The bug has
5145 been fixed by making things happen in the correct order.
5151 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5152 before the connection was closed, thus losing the rejection response.
5154 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5155 some early Solaris releases, but causes trouble in current releases where
5156 socklen_t is defined.
5158 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5161 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5164 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5165 that releasing the top of it at the end releases what was used for sub-
5166 expansions (unless the block got too big). However, discard this block if
5167 the first thing is a variable or header, so that we can use its block when
5168 it is dynamic (useful for very large $message_headers, for example).
5170 6. Lookups now cache *every* query, not just the most recent. A new, separate
5171 store pool is used for this. It can be recovered when all lookup caches are
5172 flushed. Lookups now release memory at the end of their result strings.
5173 This has involved some general refactoring of the lookup sources.
5175 7. Some code has been added to the store_xxx() functions to reduce the amount
5176 of flapping under certain conditions.
5178 8. log_incoming_interface used to affect only the <= reception log lines. Now
5179 it causes the local interface and port to be added to several more SMTP log
5180 lines, for example "SMTP connection from", and rejection lines.
5182 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5184 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5186 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5187 could be overwritten at the end of the current message (or the start of a
5188 new message if it was set in a HELO ACL). The value is now preserved for
5189 the duration of the SMTP connection.
5191 12. If a transport had a headers_rewrite setting, and a matching header line
5192 contained an unqualified address, that address was qualified, even if it
5193 did not match any rewriting rules. The underlying bug was that the values
5194 of the flags that permit the existence of unqualified sender and recipient
5195 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5196 non-local messages, and by -bnq for local messages) were not being
5197 preserved with the message after it was received.
5199 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5200 "next input=" as part of the text comprising the host identity instead of
5201 the correct text. The code was using the same buffer for two different
5202 strings. However, depending on which order the printing function evaluated
5203 its arguments, the bug did not always show up. Under Linux, for example, my
5204 test suite worked just fine.
5206 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5207 doesn't work with some older versions of Perl. It has been changed to "my",
5208 which in any case is probably the better facility to use.
5210 15. A really picky compiler found some instances of statements for creating
5211 error messages that either had too many or two few arguments for the format
5214 16. The size of the buffer for calls to the DNS resolver has been increased
5215 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5216 for addresses that have a lot of PTR records. This alleviates a problem; it
5217 does not fully solve it.
5219 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5220 buffer now truncates the list and logs the incident, which is the same
5221 action as happens when Exim is looking up a host name and its aliases.
5222 Previously in this situation something unpredictable would happen;
5223 sometimes it was "internal error: store_reset failed".
5225 18. If a server dropped the connection unexpectedly when an Exim client was
5226 using GnuTLS and trying to read a response, the client delivery process
5227 crashed while trying to generate an error log message.
5229 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5230 string, for example:
5232 warn message = H1: something\nH2: something
5234 the text was added as a single header line from Exim's point of view
5235 though it ended up OK in the delivered message. However, searching for the
5236 second and subsequent header lines using $h_h2: did not work. This has been
5237 fixed. Similarly, if a system filter added multiple headers in this way,
5238 the routers could not see them.
5240 20. Expanded the error message when iplsearch is called with an invalid key to
5241 suggest using net-iplsearch in a host list.
5243 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5244 ACLs are no longer actually imposed (and a message to that effect is
5247 22. If a "gecos" field in a passwd entry contained escaped characters, in
5248 particular, if it contained a \" sequence, Exim got it wrong when building
5249 a From: or a Sender: header from that name. A second bug also caused
5250 incorrect handling when an unquoted " was present following a character
5251 that needed quoting.
5253 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5254 was not being matched caselessly.
5256 24. Arranged for all hyphens in the exim.8 source to be escaped with
5259 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5260 properly. Recipient callout cache records were still being keyed to include
5261 the sender, even when use_sender was set false. This led to far more
5262 callouts that were necessary. The sender is no longer included in the key
5263 when use_sender is false.
5265 26. Added "control = submission" modifier to ACLs.
5267 27. Added the ${base62d: operator to decode base 62 numbers.
5269 28. dnsdb lookups can now access SRV records.
5271 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5272 the configuration file.
5274 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5275 (-v) mode. This makes the output for a verbose queue run more intelligible.
5277 31. Added a use_postmaster feature to recipient callouts.
5279 32. Added the $body_zerocount variable, containing the number of binary zero
5280 bytes in the message body.
5282 33. The time of last modification of the "new" subdirectory is now used as the
5283 "mailbox time last read" when there is a quota error for a maildir
5286 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5288 35. Added +ignore_unknown as a special item in host lists.
5290 36. Code for decoding IPv6 addresses in host lists is now included, even if
5291 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5292 address was recognized as an IP address, but was then not correctly decoded
5293 into binary, causing unexpected and incorrect effects when compared with
5300 1. Very minor rewording of debugging text in manualroute to say "list of
5301 hosts" instead of "hostlist".
5303 2. If verify=header_syntax was set, and a header line with an unqualified
5304 address (no domain) and a large number of spaces between the end of the
5305 name and the colon was received, the reception process suffered a buffer
5306 overflow, and (when I tested it) crashed. This was caused by some obsolete
5307 code that should have been removed. The fix is to remove it!
5309 3. When running in the test harness, delay a bit after writing a bounce
5310 message to get a bit more predictability in the log output.
5312 4. Added a call to search_tidyup() just before forking a reception process. In
5313 theory, someone could use a lookup in the expansion of smtp_accept_max_
5314 per_host which, without the tidyup, could leave open a database connection.
5316 5. Added the variables $recipient_data and $sender_data which get set from a
5317 lookup success in an ACL "recipients" or "senders" condition, or a router
5318 "senders" option, similar to $domain_data and $local_part_data.
5320 6. Moved the writing of debug_print from before to after the "senders" test
5323 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5324 problems for message scanning, either using a data ACL, or using
5325 local_scan() because the Received: header was not generated till after they
5326 were called (in order to set the time as the time of reception completion).
5327 I have revised the way this works. The header is now generated after the
5328 body is received, but before the ACL or local_scan() are called. After they
5329 are run, the timestamp in the header is updated.
5335 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5336 before starting a queue runner without re-exec. This happened only when
5337 deliver_drop_privilege was set or when the Exim user was set to root. The
5338 effect of the bug was that timeouts during subsequent deliveries caused
5339 crashes instead of being properly handled. The handler is now left at its
5340 default (and expected) setting.
5342 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5343 message, again when deliver_drop_privilege is set or Exim is run as root.
5344 The bug described in (1) was not present in this case, but the tidying up
5345 of the other signals was missing. I have made the two cases consistent.
5347 3. The ignore_target_hosts setting on a manualroute router was being ignored
5348 for hosts that were looked up using the /MX notation.
5350 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5353 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5354 operated on the sender address. After changing the $sender_address to <>
5355 for the sender address verify, Exim was re-instated it as the original
5356 (before rewriting) address, but remembering that it had rewritten it, so it
5357 wasn't rewriting it again. This bug also had the effect of breaking the
5358 sender address verification caching when the sender address was rewritten.
5360 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5361 This has been changed so that if the ip literal address matches
5362 ignore_target_hosts, the router declines.
5364 7. Added expansion conditions match_domain, match_address, and match_local_
5365 part (NOT match_host).
5367 8. The placeholder for the Received: header didn't have a length field set.
5369 9. Added code to Exim itself and to exim_lock to test for a specific race
5370 condition that could lead to file corruption when using MBX delivery. The
5371 issue is with the lockfile that is created in /tmp. If this file is removed
5372 after a process has opened it but before that process has acquired a lock,
5373 there is the potential for a second process to recreate the file and also
5374 acquire a lock. This could lead to two Exim processes writing to the file
5375 at the same time. The added code performs the same test as UW imapd; it
5376 checks after acquiring the lock that its file descriptor still refers to
5377 the same named file.
5379 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5380 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5383 11. Added the smtp_active_hostname option. If used, this will typically be made
5384 to depend on the incoming interface address. Because $interface_address is
5385 not set up until the daemon has forked a reception process, error responses
5386 that can happen earlier (such as "too many connections") no longer contain
5389 12. If an expansion in a condition on a "warn" statement fails because a lookup
5390 defers, the "warn" statement is abandoned, and the next ACL statement is
5391 processed. Previously this caused the whole ACL to be aborted.
5393 13. Added the iplsearch lookup type.
5395 14. Added ident_timeout as a log selector.
5397 15. Added tls_certificate_verified as a log selector.
5399 16. Added a global option tls_require_ciphers (compare the smtp transport
5400 option of the same name). This controls incoming TLS connections.
5402 17. I finally figured out how to make tls_require_ciphers do a similar thing
5403 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5404 before starting the TLS session.
5406 18. Tabs are now shown as \t in -bP output.
5408 19. If the log selector return_path_on_delivery was set, Exim crashed when
5409 bouncing a message because it had too many Received: header lines.
5411 20. If two routers both had headers_remove settings, and the first one included
5412 a superfluous trailing colon, the final name in the first list and the
5413 first name in the second list were incorrectly joined into one item (with a
5414 colon in the middle).
5420 1. Added -C and -D options to the exinext utility, mainly to make it easier
5421 to include in the automated testing, but these could be helpful when
5422 multiple configurations are in use.
5424 2. The exinext utility was not formatting the output nicely when there was
5425 an alternate port involved in the retry record key, nor when there was a
5426 message id as well (for retries that were specific to a specific message
5427 and a specific host). It was also confused by IPv6 addresses, because of
5428 the additional colons they contain. I have fixed the IPv4 problem, and
5429 patched it up to do a reasonable job for IPv6.
5431 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5432 delivery, the log line now contains "pipelined" if PIPELINING was used.
5434 4. An SMTP transport process used to panic and die if the bind() call to set
5435 an explicit outgoing interface failed. This has been changed; it is now
5436 treated in the same way as a connect() failure.
5438 5. A reference to $sender_host_name in the part of a conditional expansion
5439 that was being skipped was still causing a DNS lookup. This no longer
5442 6. The def: expansion condition was not recognizing references to header lines
5443 that used bh_ and bheader_.
5445 7. Added the _cache feature to named lists.
5447 8. The code for checking quota_filecount in the appendfile transport was
5448 allowing one more file than it should have been.
5450 9. For compatibility with Sendmail, the command line option
5458 and sets the incoming protocol and host name (for trusted callers). The
5459 host name and its colon can be omitted when only the protocol is to be set.
5460 Note the Exim already has two private options, -pd and -ps, that refer to
5461 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5462 "s", but I don't think that's a major issue.
5464 10. A number of refactoring changes to the code, none of which should affect
5467 (a) The number of logging options was getting close to filling up the
5468 32-bit word that was used as a bit map. I have split them into two classes:
5469 those that are passed in the argument to log_write(), and those that are
5470 only ever tested independently outside of that function. These are now in
5471 separate 32-bit words, so there is plenty of room for expansion again.
5472 There is no change in the user interface or the logging behaviour.
5474 (b) When building, for example, log lines, the code previously used a
5475 macro that called string_cat() twice, in order to add two strings. This is
5476 not really sufficiently general. Furthermore, there was one instance where
5477 it was actually wrong because one of the argument was used twice, and in
5478 one call a function was used. (As it happened, calling the function twice
5479 did not affect the overall behaviour.) The macro has been replaced by a
5480 function that can join an arbitrary number of extra strings onto a growing
5483 (c) The code for expansion conditions now uses a table and a binary chop
5484 instead of a serial search (which was left over from when there were very
5485 few conditions). Also, it now recognizes conditions like "pam" even when
5486 the relevant support is not compiled in: a suitably worded error message is
5487 given if an attempt is made to use such a condition.
5489 11. Added ${time_interval:xxxxx}.
5491 12. A bug was causing one of the ddress fields not to be passed back correctly
5492 from remote delivery subprocesses. The field in question was not being
5493 subsequently used, so this caused to problems in practice.
5495 13. Added new log selectors queue_time and deliver_time.
5497 14. Might have fixed a bug in maildirsizefile handling that threw up
5498 "unexpected character" debug warnings, and recalculated the data
5499 unnecessarily. In any case, I expanded the warning message to give more
5502 15. Added the message "Restricted characters in address" to the statements in
5503 the default ACL that block characters like @ and % in local parts.
5505 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5506 Three changes have been made:
5508 (a) There was a serious bug; a negative response to MAIL caused the whole
5509 recipient domain to be cached as invalid, thereby blocking all messages
5510 to all local parts at the same domain, from all senders. This bug has
5511 been fixed. The domain is no longer cached after a negative response to
5512 MAIL if the sender used is not empty.
5514 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5517 (c) A new callout option, "use_sender" has been added for people who want
5518 the modified behaviour.
5524 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5527 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5528 indeed breaks things for older releases.
5530 3. Added additional logging to the case where there is a problem reading data
5531 from a filter that is running in a subprocess using a pipe, in order to
5532 try to track down a specific problem.
5534 4. Testing facility fudge: when running in the test harness and attempting
5535 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5536 getting "No route to host". Convert this to a timeout.
5538 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5541 6. Some OS don't have socklen_t but use size_t instead. This affects the
5542 fifth argument of getsockopt() amongst other things. This is now
5543 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5544 can be set for individual OS. I have set it for SunOS5, OSF1, and
5545 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5546 some earlier ones do not.
5548 7. Change 4.30/15 was not doing the test caselessly.
5550 8. The standard form for an IPv6 address literal was being rejected by address
5551 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5552 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5553 this, as well as the form without the "IPv6" on the front (but only when
5554 address literals are enabled, of course).
5556 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5558 10. Exim crashed if a message with an empty sender address specified by -f
5559 encountered a router with an errors_to setting. This could be provoked only
5560 by a command such as
5564 where an empty string was supplied; "<>" did not hit this bug.
5566 11. Installed PCRE release 4.5.
5568 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5569 remained set. It is now erased.
5571 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5572 times from message ids (which are base 36 rather than the normal 62).
5574 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5575 were being counted as actual protocol errors, and logged if the log
5576 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5577 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5578 and DATA following a set of rejected RCPTs do not count as protocol errors.
5579 In other words, Exim assumes they were pipelined, though this may not
5580 actually be the case. Of course, in all cases the client gets an
5581 appropriate error code.
5583 15. If a lookup fails in an ACL condition, a message about the failure may
5584 be available; it is used if testing the ACL cannot continue, because most
5585 such messages specify what the cause of the deferral is. However, some
5586 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5587 that caused an old message to be retained and used if a later statement
5588 caused a defer, replacing the real cause of the deferral.
5590 16. If an IP address had so many PTR records that the DNS lookup buffer
5591 was not large enough to hold them, Exim could crash while trying to process
5592 the truncated data. It now detects and logs this case.
5594 17. Further to 4.21/58, another change has been made: if (and only if) the
5595 first line of a message (the first header line) ends with CRLF, a bare LF
5596 in a subsequent header line has a space inserted after it, so as not to
5597 terminate the header.
5599 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5600 unnecessarily, used atoi() instead of strtol(), and didn't check the
5601 termination when getting file sizes from file names by regex.
5603 19. Completely re-implemented the support for maildirsize files, in the light
5604 of a number of problems with the previous contributed implementation
5605 (4.30/29). In particular:
5607 . If the quota is zero, the maildirsize file is maintained, but no quota is
5610 . If the maildir directory does not exist, it is created before any attempt
5611 to write a maildirsize file.
5613 . The quota value in the file is just a cache; if the quota is changed in
5614 the transport, the new value overrides.
5616 . A regular expression is available for excluding directories from the
5619 20. The autoreply transport checks the characters in options that define the
5620 message's headers; it allows continued headers, but it was checking with
5621 isspace() after an embedded newline instead of explicitly looking for a
5624 21. If all the "regular" hosts to which an address was routed had passed their
5625 expiry times, and had not reached their retry times, the address was
5626 bounced, even if fallback hosts were defined. Now Exim should go on to try
5629 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5630 equivalent code in the SMTP transport. Some hosts send humungous responses
5631 to HELO/EHLO, more than 1024 it seems.
5633 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5634 gives compiler warnings in some environments. I've now done it "properly",
5637 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5638 (because of problems with the built-in one) was declared to return uschar *
5639 instead of char *, causing compiler failure.
5641 25. Fixed a file descriptor leak when processing alias/forward files.
5643 26. Fixed a minor format string issue in dbfn.c.
5645 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5647 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5648 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5651 29. When a sender address is verified, it is cached, to save repeating the test
5652 when there is more than one recipient in a message. However, when the
5653 verification involves a callout, it is possible for different callout
5654 options to be set for different recipients. It is too complicated to keep
5655 track of this in the cache, so now Exim always runs a verification when a
5656 callout is required, relying on the callout cache for the optimization.
5657 The overhead is duplication of the address routing, but this should not be
5660 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5661 to be verified with callout=postmaster, and the main callout worked but the
5662 postmaster check failed, the verification correctly failed. However, if a
5663 subsequent RCPT command asked for sender verification *without* the
5664 postmaster check, incorrect caching caused this verification also to fail,
5667 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5668 it was not caching the DNS options (qualify_single, search_parents) that
5669 were used when the lookup failed. A subsequent lookup with different
5670 options therefore always gave the same answer, though there were cases
5671 where it should not have. (Example: a "domains = !$mx_any" option on a
5672 dnslookup router: the "domains" option is always processed without any
5673 widening, but the router might have qualify_single set.) Now Exim uses the
5674 cached value only when the same options are set.
5676 32. Added John Jetmore's "exipick" utility to the distribution.
5678 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5679 than a timeout (e.g. a certificate is required, and is not provided), an
5680 Exim server now closes the connection immediately. Previously it waited for
5681 the client to close - but if the client is SSL, it seems that they each
5682 wait for each other, leading to a delay before one of them times out.
5684 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5685 maintained 0.8.x compatibility because I don't think many are using it, and
5686 it is clearly obsolete.
5688 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5691 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5692 client certificate was expired. A simple patch fixes this, though I don't
5693 understand the full logic of why the verify callback is called multiple
5696 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5697 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5698 which causes problems with some clients (such as the Certicom SSL Plus
5699 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5700 disables the coutermeasure allowing Eudora to connect."
5702 38. Exim was not checking that a write() to a log file succeeded. This could
5703 lead to Bad Things if a log got too big, in particular if it hit a file
5704 size limit. Exim now panics and dies if it cannot write to a log file, just
5705 as it does if it cannot open a log file.
5707 39. Modified OS/Makefile-Linux so that it now contains
5709 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5711 The two -D definitions ensure that Exim is compiled with large file
5712 support, which makes it possible to handle log files that are bigger than
5715 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5716 instance) a domain was checked against a named list that involved a lookup,
5717 causing $domain_data to be set, then another domain was checked against the
5718 same list, then the first domain was re-checked, the value of $domain_data
5719 after the final check could be wrong. In particular, if the second check
5720 failed, it could be set empty. This bug probably also applied to
5723 41. The strip_trailing_dot option was not being applied to the address given
5724 with the -f command-line option.
5726 42. The code for reading a message's header from the spool was incrementing
5727 $received_count, but never initializing it. This meant that the value was
5728 incorrect (doubled) while delivering a message in the same process in which
5729 it was received. In the most common configuration of Exim, this never
5730 happens - a fresh exec is done - but it can happen when
5731 deliver_drop_privilege is set.
5733 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5734 it now includes up to 150 characters of the unexpected data in the log
5737 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5738 and building data strings. The size of both of these buffers was 10 000
5739 bytes - far larger than anybody would *ever* want, thought I. Needless to
5740 say, somebody hit the limit. I have increased the maximum line length to
5741 20 000 and the maximum data length of concatenated lines to 100 000. I have
5742 also fixed two bugs, because there was no checking on these buffers. Tsk,
5743 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5744 buffer is too small.
5746 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5747 lsearch lookups. Now it does.
5749 46. When parsing a route_list item in a manualroute router, a fixed-length
5750 buffer was used for the list of hosts. I made this 1024 bytes long,
5751 thinking that nobody would ever have a list of hosts that long. Wrong.
5752 Somebody had a whole pile of complicated expansion conditions, and the
5753 string was silently truncated, leading to an expansion error. It turns out
5754 that it is easier to change to an unlimited length (owing to other changes
5755 that have happened since this code was originally written) than to build
5756 structure for giving a limitation error. The length of the item that
5757 expands into the list of hosts is now unlimited.
5759 47. The lsearch lookup could not handle data where the length of text line was
5760 more than 4095 characters. Such lines were truncated, leading to shortened
5761 data being returned. It should now handle lines of any length.
5763 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5764 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5767 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5768 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5769 no longer happens in comment lines. A list of replacements is now placed
5770 at the head of all of the source files, except those whose only change is
5771 to replace PERL_COMMAND in the very first #! line.
5773 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5774 messages on the queue, with a bottom-up merge sort, using code contributed
5775 by Michael Haardt. This should make operations like -bp somewhat faster on
5776 large queues. It won't affect queue runners, except when queue_run_in_order
5779 51. Installed eximstats 1.31 in the distribution.
5781 52. Added support for SRV lookups to the dnslookup router.
5783 53. If an ACL referred to $message_body or $message_body_end, the value was not
5784 reset for any messages that followed in the same SMTP session.
5786 54. The store-handling optimization for building very long strings was not
5787 differentiating between the different store pools. I don't think this
5788 actually made any difference in practice, but I've tidied it.
5790 55. While running the routers to verify a sender address, $sender_address
5791 was still set to the sender address. This is wrong, because when routing to
5792 send a bounce to the sender, it would be empty. Therefore, I have changed
5793 it so that, while verifying a sender address, $sender_address is set to <>.
5794 (There is no change to what happens when verifying a recipient address.)
5796 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5797 target A or AAAA records (if not already returned) without resetting the
5798 qualify_single or search_parents options of the DNS resolver. These are
5799 inappropriate in this case because the targets of MX and SRV records must
5800 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5801 target that, when qualified, matched something in the local domain. These
5802 two options are now turned off when doing these lookups.
5804 57. It seems that at least some releases of Reiserfs (which does not have the
5805 concept of a fixed number of inodes) returns zero and not -1 for the
5806 number of available inodes. This interacted badly with check_spool_inodes,
5807 which assumed that -1 was the "no such thing" setting. What I have done is
5808 to check that the total number of inodes is greater than zero before doing
5809 the test of how many are available.
5811 58. When a "warn" ACL statement has a log_message modifier, the message is
5812 remembered, and not repeated. This is to avoid a lot of repetition when a
5813 message has many recipients that cause the same warning to be written.
5814 However, Exim was preserving the list of already written lines for an
5815 entire SMTP session, which doesn't seem right. The memory is now reset if a
5816 new message is started.
5818 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5819 debugging output unless log_rewrite was also set.
5821 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5822 of (int)(handle) when we know that handle contains (void *)(-1).
5824 61. The Exim daemon panic-logs an error return when it closes the incoming
5825 connection. However "connection reset by peer" seems to be common, and
5826 isn't really an error worthy of noting specially, so that particular error
5829 62. When Exim is trying to find all the local interfaces, it used to panic and
5830 die if the ioctl to get the interface flags failed. However, it seems that
5831 on at least one OS (Solaris 9) it is possible to have an interface that is
5832 included in the list of interfaces, but for which you get a failure error
5833 for this call. This happens when the interface is not "plumbed" into a
5834 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5835 failure of the "get flags" call assumes that the interface is down.
5837 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5838 makes life easier for people who are doing arithmetic on fields extracted
5839 from dates, where you often get leading zeros that should not be
5840 interpreted as octal.
5842 64. Added qualify_domain to the redirect router, to override the global
5845 65. If a pathologically long header line contained very many addresses (the
5846 report of this problem mentioned 10 000) and each of them was rewritten,
5847 Exim could use up a very large amount of memory. (It kept on making new
5848 copies of the header line as it rewrote, and never released the old ones.)
5849 At the expense of a bit more processing, the header rewriting function has
5850 been changed so that it no longer eats memory in this way.
5852 66. The generation of the Received: header has been moved from the time that a
5853 message starts to be received, to the time that it finishes. The timestamp
5854 in the Received: header should now be very close to that of the <= log
5855 line. There are two side-effects of this change:
5857 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5858 logged header lines no longer include the local Received: line, because
5859 it has not yet been created. The same applies to a copy of the message
5860 that is returned to a non-SMTP sender when a message is rejected.
5862 (b) When a filter file is tested using -bf, no additional Received: header
5863 is added to the test message. After some thought, I decided that this
5866 This change does not affect the value of $received_for. It is still set
5867 after address rewriting, but before local_scan() is called.
5869 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5871 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5872 gave an unhelpful panic error message, and a defer error. I have managed to
5873 change this behaviour so that it now rejects any supplied certificate,
5874 which seems right, as the list of acceptable certificates is empty.
5876 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5877 gave an unhelpful defer error. I have not managed to make this reject any
5878 supplied certificates, but the error message it gives is "no certificate
5879 supplied", which is not helpful.
5881 70. exigrep's output now also includes lines that are not associated with any
5882 message, but which match the given pattern. Implemented by a patch from
5883 Martin Sluka, which also tidied up the Perl a bit.
5885 71. Recipient callout verification, like sender verification, was using <> in
5886 the MAIL FROM command. This isn't really the right thing, since the actual
5887 sender may affect whether the remote host accepts the recipient or not. I
5888 have changed it to use the actual sender in the callout; this means that
5889 the cache record is now keyed on a recipient/sender pair, not just the
5890 recipient address. There doesn't seem to be a real danger of callout loops,
5891 since a callout by the remote host to check the sender would use <>.
5892 [SEE ABOVE: changed after hitting problems.]
5894 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5895 temporary errors. However, in the case of such a code being given after
5896 the end of a data transmission (i.e. after ".") Exim was failing to write
5897 a retry record for the message. (Yes, there was some broken host that was
5898 actually sending 8xx at this point.)
5900 73. An unknown lookup type in a host list could cause Exim to panic-die when
5901 the list was checked. (An example that provoked this was putting <; in the
5902 middle of a list instead of at the start.) If this happened during a DATA
5903 ACL check, a -D file could be left lying around. This kind of configuration
5904 error no longer causes Exim to die; instead it causes a defer error. The
5905 incident is still logged to the main and panic logs.
5907 74. Buglet left over from Exim 3 conversion. The message "too many messages
5908 in one connection" was written to the rejectlog but not the mainlog, except
5909 when address rewriting (yes!) was being logged.
5911 75. Added write_rejectlog option.
5913 76. When a system filter was run not as root (that is, when system_filter_user
5914 was set), the values of the $n variables were not being returned to the
5915 main process; thus, they were not subsequently available in the $sn
5918 77. Added +return_path_on_delivery log selector.
5920 78. A connection timeout was being treated differently from recipients deferred
5921 when testing hosts_max_try with a message that was older than the host's
5922 retry timeout. (The host should not be counted, thus allowing all hosts to
5923 be tried at least once before bouncing.) This may have been the cause of an
5924 occasionally reported bug whereby a message would remain on the queue
5925 longer than the retry timeout, but would be bounced if a delivery was
5926 forced. I say "may" because I never totally pinned down the problem;
5927 setting up timeout/retry tests is difficult. See also the next item.
5929 79. The ultimate address timeout was not being applied to errors that involved
5930 a combination of host plus message (for example, a timeout on a MAIL
5931 command). When an address resolved to a number of possible hosts, and they
5932 were not all tried for each delivery (e.g. because of hosts_max_try), a
5933 message could remain on the queue longer than the retry timeout.
5935 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5938 81. Fixed an obscure SMTP outgoing bug which required at least the following
5939 conditions: (a) there was another message waiting for the same server;
5940 (b) the server returned 5xx to all RCPT commands in the first message so
5941 that the message was not completed; (c) the server dropped the connection
5942 or gave a negative response to the RSET that Exim sends to abort the
5943 transaction. The observed case was a dropped connection after DATA that had
5944 been sent in pipelining mode. That is, the server had advertised PIPELINING
5945 but was not implementing it correctly. The effect of the bug was incorrect
5946 behaviour, such as trying another host, and this could lead to a crash.
5952 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5953 and daemon.c were passed as pointers to ints; they should have been
5954 pointers to socklen_t variables (which are typically unsigned ints).
5956 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5959 3. Fixed a really odd bug that affected only the testing scheme; patching a
5960 certain fixed string in the binary changed the value of another string that
5961 happened to be identical to the end of the original first string.
5963 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5964 name", it returns that address as the IP address. On some operating
5965 systems (e.g. Solaris), it also passes back the IP address string as the
5966 "host name". However, on others (e.g. Linux), it passes back an empty
5967 string. Exim wasn't checking for this, and was changing the host name to an
5968 empty string, assuming it had been canonicalized.
5970 5. Although rare, it is permitted to have more than one PTR record for a given
5971 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5972 all the names associated with an address, because they do in Solaris.
5973 However, it seems that they do not in Linux for data that comes from the
5974 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5975 I found this out when I moved to a new Linux workstation and tried to run
5976 the Exim test suite.
5978 To get round this problem I have changed the code so that it now does its
5979 own call to the DNS to look up PTR records when searching for a host name.
5980 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5981 addresses that are only in /etc/hosts are still found.
5983 This behaviour is, however, controlled by an option called host_lookup_
5984 order, which defaults to "bydns:byaddr". If people want to use the other
5985 order, or indeed, just use one or the other means of lookup, they can
5986 specify it in this variable.
5988 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5989 some operating systems, this comes back from gethostbyaddr() as an empty
5990 string, and this is what Exim used to test for. However, it seems that in
5991 other systems, "." is yielded. Exim now tests for this case too.
5993 7. The values of check_spool_space and check_log_space are now held internally
5994 as a number of kilobytes instead of an absolute number of bytes. If a
5995 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5996 kilobyte. This means that much larger values can be stored.
5998 8. Exim monitor: an attempt to get the action menu when not actually pointing
5999 at a message produces an empty menu entitled "No message selected". This
6000 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6001 no entries in it ("Shell widget menu has zero width and/or height"). So I
6002 have added a single, blank menu entry in this case.
6004 9. Added ${quote_local_part.
6006 10. MIME decoding is now applied to the contents of Subject: header lines when
6009 11. Now that a reference to $sender_host_address automatically causes a reverse
6010 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6011 host lookup before query-style lookups in lists that might use this
6012 variable. This has therefore been abolished, and the "net-" prefix is no
6013 longer necessary for query-style lookups.
6015 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6016 have been a typo for LFLAGS, so it has been changed.
6018 13. The install script calls Exim with "-C /dev/null" in order to find the
6019 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6020 to be output. However, since Exim outputs its version number before the
6021 error, it didn't break the script. It just looked ugly. I fixed this by
6022 always allowing "-C /dev/null" if the caller is root.
6024 14. Ignore overlarge ACL variable number when reading spool file - insurance
6025 against a later release with more variables having written the file.
6027 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6028 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6029 this, as well as the form without the "IPv6" on the front.
6031 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6032 OS/Makefile-Darwin file.
6034 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6035 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6037 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6038 variable. However, some OS use ints for inodes. I've added cast to long int
6039 to get rid of the compiler warning.
6041 19. I had forgotten to lock out "/../" in configuration file names when
6042 ALT_CONFIG_PREFIX was set.
6044 20. Routers used for verification do not need to specify transports. However,
6045 if such a router generated a host list, and callout was configured, Exim
6046 crashed, because it could not find a port number from the (non-existent)
6047 transport. It now assumes port 25 in this circumstance.
6049 21. Added the -t option to exigrep.
6051 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6052 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6053 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6054 them is set, LOOKUP_LSEARCH is forced.
6056 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6058 24. Added sender and host information to the "rejected by local_scan()" log
6059 line; previously there was no indication of these.
6061 25. Added .include_if_exists.
6063 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6064 Linux. It turns out that not all file systems support this. Apparently some
6065 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6066 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6067 not supported on the file descriptor, is now ignored when Exim is trying to
6068 sync a directory. This applies only to Linux.
6070 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6072 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6073 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6076 29. Added support for maildirsize files from supplied patch (modified a bit).
6078 30. The use of :fail: followed by an empty string could lead Exim to respond to
6079 sender verification failures with (e.g.):
6081 550 Verification failed for <xxx>
6082 550 Sender verify failed
6084 where the first response line was missing the '-' that indicates it is not
6085 the final line of the response.
6087 31. The loop for finding the name of the user that called Exim had a hardwired
6088 limit of 10; it now uses the value of finduser_retries, which is used for
6089 all other user lookups.
6091 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6094 33. Exim was neglecting to zero errno before one call of strtol() when
6095 expanding a string and expecting an integer value. On some systems this
6096 resulted in spurious "integer overflow" errors. Also, it was casting the
6097 result into an int without checking.
6099 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6100 did not work. The code looks as if it has *never* worked, though it appears
6101 to have been documented since at least release 1.62. I have made it work.
6103 35. The "timeout_DNS" error in retry rules, also documented since at least
6104 1.62, also never worked. As it isn't clear exactly what this means, and
6105 clearly it isn't a major issue, I have abolished the feature by treating it
6106 as "timeout", and writing a warning to the main and panic logs.
6108 36. The display of retry rules for -brt wasn't always showing the error code
6111 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6112 timeout_connect_A, timeout_connect_MX.
6114 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6115 to the empty sender.
6117 39. The daemon was not analysing the content of -oX till after it had closed
6118 stderr and disconnected from the controlling terminal. This meant that any
6119 syntax errors were only noted on the panic log, and the return code from
6120 the command was 0. By re-arranging the code a little, I've made the
6121 decoding happen first, so such errors now appear on stderr, and the return
6122 code is 1. However, the actual setting up of the sockets still happens in
6123 the disconnected process, so errors there are still only recorded on the
6126 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6127 connections (as happens on some IP stacks) was logged at start up time as
6128 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6129 it from "IPv6 and IPv4", which means that two separate sockets are being
6132 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6133 says whether AF_INET or AF_INET6 was passed as an argument.
6135 42. Exiwhat output was messed up when time zones were included in log
6138 43. Exiwhat now gives more information about the daemon's listening ports,
6139 and whether -tls-on-connect was used.
6141 44. The "port" option of the smtp transport is now expanded.
6143 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6144 silently ignored. Now an error message is written to the main and panic
6147 46. There's a new ACL modifier called "logwrite" which writes to a log file
6148 as soon as it is encountered.
6150 47. Added $local_user_uid and $local_user_gid at routing time.
6152 48. Exim crashed when trying to verify a sender address that was being
6155 49. Exim was recognizing only a space character after ".include". It now also
6156 recognizes a tab character.
6158 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6159 extracting the relevant information from the specification. The man page no
6160 longer contains scrambled data for the -d option, and I've added a section
6161 at the front about calling Exim under different names.
6163 51. Added "extra_headers" argument to the "mail" command in filter files.
6165 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6168 53. Installed eximstats 1.29.
6170 54. Added transport_filter_timeout as a generic transport option.
6172 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6173 Cc: header lines. This was required by RFC 822, but it not required by RFC
6176 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6177 incoming messages that did not have them. Now it does so only if the
6178 message originates locally, that is, if there is no associated remote host
6179 address. When Resent- header lines are present, this applies to the Resent-
6180 lines rather than the non-Resent- lines.
6182 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6183 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6185 58. Messages for configuration errors now include the name of the main
6186 configuration file - useful now that there may be more than one file in a
6187 list (.included file names were always shown).
6189 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6190 for those rare installations that do not start the daemon as root or run it
6191 setuid root. I've cut out the call to initgroups() if the daemon is not
6194 60. The Exim user and group can now be bound into the binary as text strings
6195 that are looked up at the start of Exim's processing.
6197 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6199 62. Added $mailstore_basename variable.
6201 63. Installed patch to sieve.c from Michael Haardt.
6203 64. When Exim failed to open the panic log after failing to open the main log,
6204 the original message it was trying to log was written to stderr and debug
6205 output, but if they were not available (the usual case in production), it
6206 was lost. Now it is written to syslog before the two lines that record the
6207 failures to open the logs.
6209 65. Users' Exim filters run in subprocesses under the user's uid. It is
6210 possible for a "deliver" command or an alias in a "personal" command to
6211 provoke an address rewrite. If logging of address rewriting is configured,
6212 this fails because the process is not running as root or exim. There may be
6213 a better way of dealing with this, but for the moment (because 4.30 needs
6214 to be released), I have disabled address rewrite logging when running a
6215 filter in a non-root, non-exim process.
6221 1. The buildconfig auxiliary program wasn't quoting the value set for
6222 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6223 not defined. This bug was masked in 4.22 by the effect that was fixed in
6226 2. Some messages that were rejected after a message id was allocated were
6227 shown as "incomplete" by exigrep. It no longer does this for messages that
6228 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6230 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6231 have allow_domain_literals set, the ID did not get logged in the <= line.
6232 Domain literals are now always recognized in Message-ID: header lines.
6234 4. The first argument for a ${extract expansion item is the key name or field
6235 number. Leading and trailing spaces in this item were not being ignored,
6236 causing some misleading effects.
6238 5. When deliver_drop_privilege was set, single queue runner processes started
6239 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6240 same command in the process it spins off) were not dropping privilege.
6242 6. When the daemon running as "exim" started a queue runner, it always
6243 re-executed Exim in the spun-off process. This is a waste of effort when
6244 deliver_drop_privilege is set. The new process now just calls the
6245 queue-runner function directly.
6251 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6254 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6255 ignored. Though the use of -oP was forcing the writing of a pid file, it
6256 was always written to the default place.
6258 3. If the message "no IP address found for host xxxx" is generated during
6259 incoming verification, it is now followed by identification of the incoming
6260 connection (so you can more easily find what provoked it).
6262 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6264 5. Added some features to "harden" Exim a bit more against certain attacks:
6266 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6267 be put in Local/Makefile. This is like the never_users runtime option,
6268 but it cannot be overridden. The default setting is "root".
6270 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6271 prefix string with which any file named in a -C command line option
6274 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6275 is retained for -C and -D only if the caller of Exim is root. Without
6276 it, the exim user may also use -C and -D and retain privilege.
6278 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6279 command line option is disabled.
6281 6. Macro names set by the -D option must start with an upper case letter, just
6282 like macro names defined in the configuration file.
6284 7. Added "dereference=" facility to LDAP.
6286 8. Two instances of the typo "uknown" in the source files are fixed.
6288 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6289 the Configure-Makefile script screwed up while processing it.
6291 10. Incorporated PCRE 4.4.
6293 11. The SMTP synchronization check was not operating right at the start of an
6294 SMTP session. For example, it could not catch a HELO sent before the client
6295 waited for the greeting. There is now a check for outstanding input at the
6296 point when the greeting is written. Because of the duplex, asynchronous
6297 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6298 way, but not yet received, when the check is performed.
6300 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6301 on TCP/IP sockets, because this apparently causes some broken clients to
6304 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6305 unchanged) from the Cygwin maintainer.
6307 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6308 was supported instead of testing for mbx. Effectively a typo.
6310 15. The spa authenticator server code was not checking that the input it
6311 received was valid base64.
6313 16. The debug output line for the "set" modifier in ACLs was not showing the
6314 name of the variable that was being set.
6316 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6318 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6319 lookup on its own. Something else was needed to trigger the lookup. For
6320 example, a match in host_lookup or the need for a host name in a host list.
6321 Now, if $sender_host_name is referenced and the host name has not yet been
6322 looked up, a lookup is performed. If the lookup fails, the variable remains
6323 empty, and $host_lookup_failed is set to "1".
6325 19. Added "eqi" as a case-independent comparison operator.
6327 20. The saslauthd authentication condition could segfault if neither service
6328 nor realm was specified.
6330 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6331 error message that was logged was misleading, and incoming SMTP
6332 connections were dropped. The message is now more accurate, and temporary
6333 errors are given to SMTP connections.
6335 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6336 (except RSET) with a 421 error, until QUIT is received. However, it was
6337 failing to send a response to QUIT.
6339 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6340 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6341 ACL is now run after the helo_try_verify_hosts code.
6343 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6344 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6345 case-independent, so other case variants are also recognized). Apparently
6346 some systems use these upper case variants.
6348 25. If more than two messages were waiting for the same host, and a transport
6349 filter was specified for the transport, Exim sent two messages over the
6350 same TCP/IP connection, and then failed with "socket operation on non-
6351 socket" when it tried to send the third.
6353 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6355 27. The extern definition of crypt16() in expand.c was not being excluded when
6356 the OS had its own crypt16() function.
6358 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6359 as a preferred synonym for return_size_limit, both as an option and as an
6362 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6364 30. Changed the default configuration ACL to relax the local part checking rule
6365 for addresses that are not in any local domains. For these addresses,
6366 slashes and pipe symbols are allowed within local parts, but the sequence
6367 /../ is explicitly forbidden.
6369 31. SPA server authentication was not clearing the challenge buffer before
6372 32. log_message in a "warn" ACL statement was writing to the reject log as
6373 well as to the main log, which contradicts the documentation and doesn't
6374 seem right (because no rejection is happening). So I have stopped it.
6376 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6377 However, I am unable to do any testing of this.
6379 34. Fixed an infelicity in the appendfile transport. When checking directories
6380 for a mailbox, to see if any needed to be created, it was accidentally
6381 using path names with one or more superfluous leading slashes; tracing
6382 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6384 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6385 contents are added to the log line that is written for every discarded
6386 recipient. (Previously a log_message setting was ignored.)
6388 36. The ${quote: operator now quotes the string if it is empty.
6390 37. The install script runs exim in order to find its version number. If for
6391 some reason other than non-existence or emptiness, which it checks, it
6392 could not run './exim', it was installing it with an empty version number,
6393 i.e. as "exim-". This error state is now caught, and the installation is
6396 38. An argument was missing from the function that creates an error message
6397 when Exim fails to connect to the socket for saslauthd authentication.
6398 This could cause Exim to crash, or give a corrupted message.
6400 39. Added isip, isip4, and isip6 to ${if conditions.
6402 40. The ACL variables $acl_xx are now saved with the message, and can be
6403 accessed later in routers, transports, and filters.
6405 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6406 strings in the file are not string-expanded.
6408 42. If a MAIL command specified a SIZE value that was too large to fit into an
6409 int variable, the check against message_size_limit failed. Such values are
6410 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6411 day this will have to be increased, but I don't think I want to be around
6412 when emails are that large.
6419 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6420 iconv() is not standard in FreeBSD.
6422 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6423 IPv6 enabled. The observed symptom was a segmentation fault on return from
6424 the function os_common_find_running_interfaces() in src/os.c.
6426 3. In the check_special_case() function in daemon.c I had used "errno" as an
6427 argument name, which causes warnings on some systems. This was basically a
6428 typo, since it was named "eno" in the comments!
6430 4. The code that waits for the clock to tick (at a resolution of some fraction
6431 of a second) so as to ensure message-id uniqueness was always waiting for
6432 at least one whole tick, when it could have waited for less. [This is
6433 almost certainly not relevant at current processor speeds, where it is
6434 unlikely to ever wait at all. But we try to future-proof.]
6436 5. The function that sleeps for a time interval that includes fractions of a
6437 second contained a race. It did not block SIGALRM between setting the
6438 timer, and suspending (a couple of lines later). If the interval was short
6439 and the sigsuspend() was delayed until after it had expired, the suspension
6440 never ended. On busy systems this could lead to processes getting stuck for
6443 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6444 process, before it forks any delivery processes. The open lookup caching
6445 mechanism meant that the open file or database connection was passed into
6446 the delivery process. The problem was that delivery processes always tidy
6447 up cached lookup data. This could cause a problem for the next delivery
6448 process started by the queue runner, because the external queue runner
6449 process does not know about the closure. So the next delivery process
6450 still has data in the lookup cache. In the case of a file lookup, there was
6451 no problem because closing a file descriptor in a subprocess doesn't affect
6452 the parent. However, if the lookup was caching a connection to a database,
6453 the connection was closed, and the second delivery process was likely to
6454 see errors such as "PGSQL: query failed: server closed the connection
6455 unexpectedly". The problem has been fixed by closing all cached lookups
6456 in a queue runner before running a delivery process.
6458 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6459 seem to have the "const" qualifier which it has on other OS. I've
6462 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6463 *of the same type* (client or server) with the same public name that an
6464 error should be diagnosed.
6466 9. When Exim looked up a host name for an IP address, but failed to find the
6467 original IP address when looking up the host name (a safety check), it
6468 output the message "<ip address> does not match any IP for NULL", which was
6469 confusing, to say the least. The bug was that the host name should have
6470 appeared instead of "NULL".
6472 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6473 user that is built into the binary, and the -C or -D options is used, root
6474 privilege is dropped before the configuration file is read. In addition,
6475 logging is switched to stderr instead of the normal log files. If the
6476 configuration then re-defines the Exim user, the unprivileged environment
6477 is probably not what is expected, so Exim logs a panic warning message (but
6480 However, if deliver_drop_privilege is set, the unprivileged state may well
6481 be exactly what is intended, so the warning has been cut out in that case,
6482 and Exim is allowed to try to write to its normal log files.
6488 1. smtp_return_error_details was not giving details for temporary sender
6489 or receiver verification errors.
6491 2. Diagnose a configuration error if two authenticators have the same public
6494 3. Exim used not to create the message log file for a message until the first
6495 delivery attempt. This could be confusing when incoming messages were held
6496 for policy or load reasons. The message log file is now created at the time
6497 the message is received, and an initial "Received" line is written to it.
6499 4. The automatically generated man page for command line options had a minor
6500 bug that caused no ill effects; however, a more serious problem was that
6501 the procedure for building the man page automatically didn't always
6502 operate. Consequently, release 4.20 contains an out-of-date version. This
6503 shouldn't happen again.
6505 5. When building Exim with embedded Perl support, the script that builds the
6506 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6507 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6509 6. The freeze_tell option was not being used for messages that were frozen on
6510 arrival, either by an ACL or by local_scan().
6512 7. Added the smtp_incomplete_transaction log selector.
6514 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6515 was accepting AUTH without a new EHLO.
6517 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6518 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6521 10. Logging of TCP/IP connections (when configured) now happens in the main
6522 daemon process instead of the child process, so that the TCP/IP connection
6523 count is more accurate (but it can never be perfect).
6525 11. The use of "drop" in a nested ACL was not being handled correctly in the
6526 outer ACL. Now, if condition failure induced by the nested "drop" causes
6527 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6528 or "require"), the connection is dropped.
6530 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6531 that yield "discard" can now be used with an "accept" or a "discard" verb,
6532 but an error is generated for any others (because I can't see a useful way
6533 to define what should happen).
6535 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6536 are now processed in the same way as lines in the Exim configuration file.
6537 In particular, continuation lines are supported.
6539 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6541 15. Added -ti meaning -t -i.
6543 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6544 domains, and warn by logging if others are found.
6546 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6547 returned from GIFCONF when Exim is trying to find the list of interfaces on
6548 a host. The code in os.c has been modified to copy each ifreq to an aligned
6549 structure in all cases.
6551 Also, in some cases, the returned ifreq's were being copied to a 'struct
6552 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6553 means the last couple of bytes of an IPv6 address could be chopped if the
6554 ifreq contained only a normal sockaddr (14 bytes storage).
6556 18. Named domain lists were not supported in the hosts_treat_as_local option.
6557 An entry such as +xxxx was not recognized, and was treated as a literal
6560 19. Ensure that header lines added by a DATA ACL are included in the reject log
6561 if the ACL subsequently rejects the message.
6563 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6564 MD5 (which is deprecated).
6566 21. When testing a filter file using -bf, Exim was writing a message when it
6567 took the sender from a "From " line in the message, but it was not doing so
6568 when it took $return_path from a Return-Path: header line. It now does.
6570 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6571 with a valid header line field name (a series of printing characters
6572 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6574 23. Changed "disc" in the source to "disk" to conform to the documentation and
6575 the book and for uniformity.
6577 24. Ignore Sendmail's -Ooption=value command line item.
6579 25. When execve() failed while trying to run a command in a pipe transport,
6580 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6581 could be confused with a return value of 69 from the command itself. This
6582 has been changed to 127, the value the shell returns if it is asked to run
6583 a non-existent command. The wording for the related log line suggests a
6584 non-existent command as the problem.
6586 26. If received_header_text expands to an empty string, do not add a Received:
6587 header line to the message. (Well, it adds a token one on the spool, but
6588 marks it "old" so that it doesn't get used or transmitted.)
6590 27. Installed eximstats 1.28 (addition of -nt option).
6592 28. There was no check for failure on the call to getsockname() in the daemon
6593 code. This can fail if there is a shortage of resources on the system, with
6594 ENOMEM, for example. A temporary error is now given on failure.
6596 29. Contrary to the C standard, it seems that in some environments, the
6597 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6598 program. Exim now does this explicitly; it affects the formatting of
6599 timestamps using strftime().
6601 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6602 complaints. I've now initialized all the variables, to avoid this.
6604 32. Header lines added by a system filter were not being "seen" during
6605 transport-time rewrites.
6607 33. The info_callback() function passed to OpenSSL is set up with type void
6608 (*)(SSL *, int, int), as described somewhere. However, when calling the
6609 function (actually a macro) that sets it up, the type void(*)() is
6610 expected. I've put in a cast to prevent warnings from picky compilers.
6612 34. If a DNS black list lookup found a CNAME record, but there were no A
6613 records associated with the domain it pointed at, Exim crashed.
6615 35. If a DNS black list lookup returned more than one A record, Exim ignored
6616 all but the first. It now scans all returned addresses if a particular IP
6617 value is being sought. In this situation, the contents of the
6618 $dnslist_value variable are a list of all the addresses, separated by a
6621 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6622 to accept a host name and all its aliases if the forward lookup for any of
6623 them yielded the IP address of the incoming connection. Now it accepts only
6624 those names whose forward lookup yields the correct IP address. Any other
6625 names are discarded. This closes a loophole whereby a rogue DNS
6626 administrator could create reverse DNS records to break through a
6627 wildcarded host restriction in an ACL.
6629 37. If a user filter or a system filter that ran in a subprocess used any of
6630 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6631 the wrong values were passed to the pipe command ($thisaddress had the
6632 value of $0, $0 had the value of $1, etc). This bug was introduced by
6633 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6635 38. Improved the line breaking for long SMTP error messages from ACLs.
6636 Previously, if there was no break point between 40 and 75 characters, Exim
6637 left the rest of the message alone. Two changes have been made: (a) I've
6638 reduced the minimum length to 35 characters; (b) if it can't find a break
6639 point between 35 and 75 characters, it looks ahead and uses the first one
6640 that it finds. This may give the occasional overlong line, but at least the
6641 remaining text gets split now.
6643 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6644 file descriptors might be low, and that setting 1000 would always raise it.
6645 It turns out that in some environments, the limit is already over 1000 and
6646 that lowering it causes trouble. So now Exim takes care not to decrease it.
6648 40. When delivering a message, the value of $return_path is set to $sender_
6649 address at the start of routing (routers may change the value). By an
6650 oversight, this default was not being set up when an address was tested by
6651 -bt or -bv, which affected the outcome if any router or filter referred to
6654 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6655 log only when "message" or "log_message" are set. However, if one of the
6656 conditions was an address verification, or a call to a nested ACL, the
6657 messages generated by the underlying test were being passed through. This
6658 no longer happens. The underlying message is available in $acl_verify_
6659 message for both "message" and "log_message" expansions, so it can be
6660 passed through if needed.
6662 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6663 new expansion $bh_ to give the encoded byte string without charset
6664 translation. Translation happens only if iconv() is available; HAVE_ICONV
6665 indicates this at build time. HEADERS_CHARSET gives the charset to
6666 translate to; headers_charset can change it in the configuration, and
6667 "headers charset" can change it in an individual filter file.
6669 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6670 that creates RFC 2047 encoded "words" labels them as that charset instead
6671 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6672 expansion operator; (ii) when Exim creates a From: line for a local
6673 message; (iii) when a header line is rewritten to include a "phrase" part.
6675 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6676 buggy, causing it to skip the first lines of messages whose message ID
6677 ended in 'D'. This would not have bitten before Exim release 4.14, because
6678 message IDs were unlikely to end in 'D' before then. The effect was to have
6679 incorrect size information for certain domains.
6681 45. #include "config.h" was missing at the start of the crypt16.c module. This
6682 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6685 46. If there was a timeout during a "random" callout check, Exim treated it as
6686 a failure of the random address, and carried on sending RSET and the real
6687 address. If the delay was just some slowness somewhere, the response to the
6688 original RCPT would be taken as a response to RSET and so on, causing
6689 mayhem of various kinds.
6691 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6692 when I implemented it. It didn't allow for the fact that some option values
6693 may legitimately be negative (e.g. size_addition), and it didn't even do
6694 the right test for positive values.
6696 48. Domain names in DNS records are case-independent. Exim always looks them up
6697 in lower case. Some resolvers return domain names in exactly the case they
6698 appear in the zone file, that is, they may contain uppercase letters. Not
6699 all resolvers do this - some return always lower case. Exim was treating a
6700 change of case by a resolver as a change of domain, similar to a widening
6701 of a domain abbreviation. This triggered its re-routing code and so it was
6702 trying to route what was effectively the same domain again. This normally
6703 caused routing to fail (because the router wouldn't handle the domain
6704 twice). Now Exim checks for this case specially, and just changes the
6705 casing of the domain that it ultimately uses when it transmits the message
6708 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6711 50. If a filter generated a file delivery with a non-absolute name (possible if
6712 no home directory exists for the router), the forbid_file option was not
6715 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6716 the existing equality matching.
6718 52. Exim was using ints instead of ino_t variables in some places where it was
6719 dealing with inode numbers.
6721 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6722 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6723 called TMPDIR, and if it finds it is different, it changes its value.
6725 54. The smtp_printf() function is now made available to local_scan() so
6726 additional output lines can be written before returning. There is also an
6727 smtp_fflush() function to enable the detection of a dropped connection.
6728 The variables smtp_input and smtp_batched_input are exported to
6731 55. Changed the default runtime configuration: the message "Unknown user"
6732 has been removed from the ACL, and instead placed on the localuser router,
6733 using the cannot_route_message feature. This means that any verification
6734 failures that generate their own messages won't get overridden. Similarly,
6735 the "Unrouteable address" message that was in the ACL for unverifiable
6736 relay addresses has also been removed.
6738 56. Added hosts_avoid_esmtp to the smtp transport.
6740 57. The exicyclog script was not checking for the esoteric option
6741 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6742 will work only if exicyclog is run under the appropriate euid.
6744 58. Following a discussion on the list, the rules by which Exim recognises line
6745 endings on incoming messages have been changed. The -dropcr and drop_cr
6746 options are now no-ops, retained only for backwards compatibility. The
6747 following line terminators are recognized: LF CRLF CR. However, special
6748 processing applies to CR:
6750 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6751 nor a local message in the state where . is a terminator.
6753 (ii) If a bare CR is encountered in a header line, an extra space is added
6754 after the line terminator so as not to end the header. The reasoning
6755 behind this is that bare CRs in header lines are most likely either
6756 to be mistakes, or people trying to play silly games.
6758 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6759 was being incorrectly given as 18 bytes larger than it should have been.
6760 This is a VOB (very old bug).
6762 60. This may never have affected anything current, but just in case it has:
6763 When the local host is found other than at the start of a list of hosts,
6764 the local host, those with the same MX, and any that follow, are discarded.
6765 When the list in question was part of a longer list of hosts, the following
6766 hosts (not currently being processed) were also being discarded. This no
6767 longer happens. I'm not sure if this situation could ever has previously
6770 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6773 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6774 header. This is something that is recommended in a new Internet Draft, and
6775 is something that is documented as being done by Sendmail. There are two
6776 possible values. For messages generated by the autoreply transport, Exim
6779 Auto-Submitted: auto-replied
6781 whereas for all other generated messages (e.g. bounces) it adds
6783 Auto-Submitted: auto-generated
6785 63. The "personal" condition in filters now includes a test for the
6786 Auto-Submitted: header. If it contains the string "auto-" the message it
6787 not considered personal.
6789 64. Added rcpt_include_affixes as a generic transport option.
6791 65. Added queue_only_override (default true).
6793 66. Added the syslog_duplication option.
6795 67. If what should have been the first header line of a message consisted of
6796 a space followed by a colon, Exim was mis-interpreting it as a header line.
6797 It isn't of course - it is syntactically invalid and should therefore be
6798 treated as the start of the message body. The misbehaviour could have
6799 caused a number of strange effects, including loss of data in subsequent
6800 header lines, and spool format errors.
6802 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6803 client host had authenticated. This control can now be exercised by an ACL
6804 for more flexibility.
6806 69. By default, callouts do not happen when testing with -bh. There is now a
6807 variant, -bhc, which does actually run the callout code, including
6808 consulting and updating the callout cache.
6810 70. Added support for saslauthd authentication, courtesy of Alexander
6813 71. If statvfs() failed on the spool or log directories while checking their
6814 size for availability, Exim confusingly gave the error "space shortage".
6815 Furthermore, in debugging mode it crashed with a floating point exception.
6816 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6817 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6818 serious problem, Exim now writes to the main and panic logs when this
6819 happens, with details of the failure. It then refuses to accept the
6820 incoming message, giving the message "spool directory problem" or "log
6821 directory problem" with a 421 code for SMTP messages.
6823 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6824 0, 1, and 2 exist, because some OS complain for execs without them (see
6825 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6826 descriptors. However, the code omitted to check that the open succeeded,
6827 causing mysterious errors if for some reason the permissions on /dev/null
6828 got screwed. Now Exim writes a message to the main and panic logs, and
6829 bombs out if it can't open /dev/null.
6831 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6832 interact so that it is all more flexible. It is supposed to remain
6833 backwards compatible. Also added extra_local_interfaces.
6835 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6836 to bomb out with an assertion failure - to the client this appears as a
6837 connection drop. This problem occurs in the part of the code that was taken
6838 from the Samba project. Fortunately, the assertion is in a very simple
6839 function, so I have fixed this by reproducing the function inline in the
6840 one place where it is called, and arranging for authentication to fail
6841 instead of killing the process with assert().
6843 75. The SPA client code was not working when the server requested OEM rather
6844 than Unicode encoding.
6846 76. Added code to make require_files with a specific uid setting more usable in
6847 the case where statting the file as root fails - usually a non-root-mounted
6848 NFS file system. When this happens and the failure is EACCES, Exim now
6849 forks a subprocess and does the per-uid checking as the relevant uid.
6851 77. Added process_log_path.
6853 78. If log_file_path was not explicitly set, a setting of check_log_space or
6854 check_log_inodes was ignored.
6856 79. If a space check for the spool or log partitions fails, the incident is now
6857 logged. Of course, in the latter case the data may get lost...
6859 80. Added the %p formatting code to string_format() so that it can be used to
6860 print addresses in debug_print(). Adjusted all the address printing in the
6861 debugging in store.c to use %p rather than %d.
6863 81. There was a concern that a line of code in smtp_in.c could overflow a
6864 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6865 initially expressed, the concern was not well-founded, because trailing
6866 spaces are removed early. However, if the trailing spaces were followed by
6867 a NULL, they did not get removed, so the overflow was possible. Two fixes
6870 (a) I re-wrote the offending code in a cleaner fashion.
6871 (b) If an incoming SMTP command contains a NULL character, it is rejected
6874 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6875 runs initgroups(), so that if the Exim user is in any additional groups,
6876 they will be used during message reception.
6882 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob