1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
40 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
41 Disable the session-cache too, which might reduce our load. Since we
42 currrectly use a new context for every connection, both as server and
43 client, there is no benefit for these.
44 GnuTLS appears to not support tickets server-side by default (we don't
45 call gnutls_session_ticket_enable_server()) but client side is enabled
46 by default on recent versions (3.1.3 +) unless the PFS priority string
49 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
50 <https://reproducible-builds.org/specs/source-date-epoch/>.
52 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
53 the check for any unsuccessful recipients did not notice the limit, and
54 erroneously found still-pending ones.
56 JH/08 Pipeline CHUNKING command and data together, on kernels that support
57 MSG_MORE. Only in-clear (not on TLS connections).
59 JH/09 Avoid using a temporary file during transport using dkim. Unless a
60 transport-filter is involved we can buffer the headers in memory for
61 creating the signature, and read the spool data file once for the
62 signature and again for transmission.
64 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
65 4.77 as the kernel support then wasn't solid, having issues in 64bit
66 mode. Now, it's been long enough. Add support for FreeBSD also.
68 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
69 case where the routing stage had gathered several addresses to send to
70 a host before calling the transport for the first, we previously failed
71 to close down TLS in the old transport process before passing the TCP
72 connection to the new process. The new one sent a STARTTLS command
73 which naturally failed, giving a failed delivery and bloating the retry
74 database. Investigation and fix prototype from Wolfgang Breyha.
76 JH/12 Fix check on SMTP command input synchronisation. Previously there were
77 false-negatives in the check that the sender had not preempted a response
78 or prompt from Exim (running as a server), due to that code's lack of
79 awareness of the SMTP input buffering.
81 PP/04 Add commandline_checks_require_admin option.
82 Exim drops privileges sanely, various checks such as -be aren't a
83 security problem, as long as you trust local users with access to their
84 own account. When invoked by services which pass untrusted data to
85 Exim, this might be an issue. Set this option in main configuration
86 AND make fixes to the calling application, such as using `--` to stop
89 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
90 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
91 responses to those, into a single TLS record each way (this usually means
92 a single packet). As a side issue, smtp_enforce_sync now works on TLS
95 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
96 affects you only if you're dancing at the edge of the param size limits.
97 If you are, and this message makes sense to you, then: raise the
98 configured limit or use OpenSSL 1.1. Nothing we can do for older
101 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
102 cmdline element to get the documented default one. Previously it was
105 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
106 are unsuable for non-message-specific reasons - eg. connection timeout,
107 banner-time rejection.
109 JH/16 Fix logging of delivery remote port, when specified by router, under
112 PP/06 Repair manualroute's ability to take options in any order, even if one
113 is the name of a transport.
116 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
118 JH/17 Change the list-building routines interface to use the expanding-string
119 triplet model, for better allocation and copying behaviour.
121 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
122 Previously it was constructed the first time a possibly-matching string
123 was met in the configuration file input during startup; now it is done
126 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
127 compatible one, to avoid the (poorly documented) possibility of a config
128 file in the working directory redirecting the DB files, possibly correpting
131 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
132 cache-hot. Previously, although the result was properly cached, the
133 initial verify call returned a defer.
139 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
140 than -2003 did; needs libidn2 in addition to libidn.
142 JH/02 The path option on a pipe transport is now expanded before use.
144 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
145 Patch provided by "Björn", documentation fix added too.
147 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
148 missing a wire-to-host endian conversion.
150 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
151 close after a BDAT command line could be taken as a following command,
152 giving a synch failure. Fix by only checking for synch immediately
153 before acknowledging the chunk.
155 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
156 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
157 macro. Patches provided by Josh Soref.
159 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
160 Previously we did not; the RFC seems ambiguous and VRFY is not listed
161 by IANA as a service extension. However, John Klensin suggests that we
164 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
165 the dkim code may be unix-mode line endings rather than smtp wire-format
166 CRLF, so prepend a CR to any bare LF.
168 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
169 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
171 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
172 the first were themselves being wrongly included in the feed into dkim
173 processing; with most chunk sizes in use this resulted in an incorrect
174 body hash calculated value.
176 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
177 DKIM signature block, for verification. Although advised against by
178 standards it is specifically not ruled illegal.
180 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
182 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
183 missing a body hash (the bh= tag).
185 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
186 It seems that HAProxy sends the Proxy Protocol information in clear and
187 only then does a TLS startup, so do the same.
189 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
190 TCP connections (such as for Spamd) unless the daemon successfully set
191 Fast Open mode on its listening sockets. This fixes breakage seen on
192 too-old kernels or those not configured for Fast Open, at the cost of
193 requiring both directions being enabled for TFO, and TFO never being used
194 by non-daemon-related Exim processes.
196 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
197 endings, at least on the first header line. Try to canonify any that get
198 past that check, despite the cost.
200 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
201 now limited to an arbitrary five deep, while parsing addresses with the
202 strip_excess_angle_brackets option enabled.
204 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
205 instead leave the unprompted TLS handshake in socket buffer for the
206 TLS library to consume.
208 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
210 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
212 JH/16 Drop variables when they go out of scope. Memory management drops a whole
213 region in one operation, for speed, and this leaves assigned pointers
214 dangling. Add checks run only under the testsuite which checks all
215 variables at a store-reset and panics on a dangling pointer; add code
216 explicitly nulling out all the variables discovered. Fixes one known
217 bug: a transport crash, where a dangling pointer for $sending_ip_address
218 originally assigned in a verify callout, is re-used.
220 PP/06 Drop '.' from @INC in various Perl scripts.
222 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
224 PP/08 Reduce a number of compilation warnings under clang; building with
225 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
226 should be warning-free.
228 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
230 HS/01 Fix portability problems introduced by PP/08 for platforms where
231 realloc(NULL) is not equivalent to malloc() [SunOS et al].
233 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
234 chunk. This allows us to accept broken chunked messages. We need a more
235 general solution here.
237 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
238 already-broken messages in the queue.
240 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
242 JH/19 Fix reference counting bug in routing-generated-address tracking.
248 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
249 supports it and a size is available (ie. the sending peer gave us one).
251 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
252 years of being deprecated). The replacements are the ACLs
253 acl_smtp_mime and acl_not_smtp_mime.
255 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
256 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
257 or even an in-clear connection were permitted. Now, if the host lookup
258 was dnssec and dane was requested then the host is only used if the
259 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
260 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
261 if one fails this test.
262 This means that a poorly-configured remote DNS will make it incommunicado;
263 but it protects against a DNS-interception attack on it.
265 JH/04 Bug 1810: make continued-use of an open smtp transport connection
266 non-noisy when a race steals the message being considered.
268 JH/05 If main configuration option tls_certificate is unset, generate a
269 self-signed certificate for inbound TLS connections.
271 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
272 in rewrites and routers.
274 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
275 and logged a warning sing 4.83; now they are a configuration file error.
277 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
278 (lacking @domain). Apply the same qualification processing as RCPT.
280 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
282 JH/10 Support ${sha256:} applied to a string (as well as the previous
285 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
286 a cutthrough deliver is pending, as we always want to make a connection.
287 This also avoids re-routing the message when later placing the cutthrough
288 connection after a verify cache hit.
289 Do not update it with the verify result either.
291 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
292 when routing results in more than one destination address.
294 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
295 signing (which inhibits the cutthrough capability). Previously only
296 the presence of an option was tested; now an expansion evaluating as
297 empty is permissible (obviously it should depend only on data available
298 when the cutthrough connection is made).
300 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
301 the relevant preceding SMTP command did not note the pipelining mode.
303 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
304 Previously they were not counted.
306 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
307 as one having no matching records. Previously we deferred the message
308 that needed the lookup.
310 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
311 distinguished as "(=".
313 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
314 for missing MX records. Previously it only worked for missing A records.
316 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
318 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
319 after the data-go-ahead and data-ack. Patch from Jason Betts.
321 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
322 even for a "none" policy. Patch from Tony Meyer.
324 JH/22 Fix continued use of a connection for further deliveries. If a port was
325 specified by a router, it must also match for the delivery to be
328 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
329 When one of the recipients of a message was unsuitable for the connection
330 (has no matching addresses), we lost track of needing to mark it
331 deferred. As a result mail would be lost.
333 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
335 JH/25 Decoding ACL controls is now done using a binary search; the source code
336 takes up less space and should be simpler to maintain. Merge the ACL
337 condition decode tables also, with similar effect.
339 JH/26 Fix problem with one_time used on a redirect router which returned the
340 parent address unchanged. A retry would see the parent address marked as
341 delivered, so not attempt the (identical) child. As a result mail would
344 JH/27 Fix a possible security hole, wherein a process operating with the Exim
345 UID can gain a root shell. Credit to http://www.halfdog.net/ for
346 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
349 JH/28 Enable {spool,log} filesystem space and inode checks as default.
350 Main config options check_{log,spool}_{inodes,space} are now
351 100 inodes, 10MB unless set otherwise in the configuration.
353 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
354 Previously it only applied to the main-section connection policy
357 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
359 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
360 by me. Added RFC7919 DH primes as an alternative.
362 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
363 are not in the system include path.
365 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
366 GnuTLS, when a session startup failed (eg because the client disconnected)
367 Exim did stdio operations after fclose. This was exposed by a recent
368 change which nulled out the file handle after the fclose.
370 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
371 signed directly by the cert-signing cert, rather than an intermediate
372 OCSP-signing cert. This is the model used by LetsEncrypt.
374 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
376 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
377 an incoming connection.
379 HS/02 Bug 1802: Do not half-close the connection after sending a request
382 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
383 fallback to "prime256v1".
385 JH/34 SECURITY: Use proper copy of DATA command in error message.
386 Could leak key material. Remotely exploitable. CVE-2016-9963.
392 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
393 and 3.4.4 - once the server is enabled to respond to an OCSP request
394 it does even when not requested, resulting in a stapling non-aware
395 client dropping the TLS connection.
397 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
398 support variable-length bit vectors. No functional change.
400 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
401 The I= interface field on outgoing lines is now after the H= remote
402 host field, same as incoming lines. There is a separate
403 outgoing_interface log selector which allows you to disable the
406 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
407 If not running log_selector +smtp_connection the mainlog would be held
408 open indefinitely after a "too many connections" event, including to a
409 deleted file after a log rotate. Leave the per net connection logging
410 leaving it open for efficiency as that will be quickly detected by the
411 check on the next write.
413 HS/01 Bug 1671: Fix post transport crash.
414 Processing the wait-<transport> messages could crash the delivery
415 process if the message IDs didn't exist for some reason. When
416 using 'split_spool_directory=yes' the construction of the spool
417 file name failed already, exposing the same netto behaviour.
419 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
420 mime_regex ACL conditions.
422 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
423 to DSN fail messages (bounces): remote IP, remote greeting, remote response
424 to HELO, local diagnostic string.
426 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
427 log line to debug. Even when configured with a tls authenticator many
428 client connections are expected to not authenticate in this way, so
429 an authenticate fail is not an error.
431 HS/02 Add the Exim version string to the process info. This way exiwhat
432 gives some more detail about the running daemon.
434 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
435 matter for fast-change records such as DNSBLs.
437 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
438 retry record, even if constant. There may be multiple transports with
439 different interface settings and the retry behaviour needs to be kept
442 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
444 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
446 JH/10 Bug 840: fix log_defer_output option of pipe transport
448 JH/11 Bug 830: use same host for all RCPTS of a message, even under
449 hosts_randomize. This matters a lot when combined with mua_wrapper.
451 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
452 ${quote_pgsql:<string>} operator.
454 JH/13 Bug 1708: avoid misaligned access in cached lookup.
456 JH/14 Change header file name for freeradius-client. Relevant if compiling
457 with Radius support; from the Gentoo tree and checked under Fedora.
459 JH/15 Bug 1712: Introduce $prdr_requested flag variable
461 JH/16 Bug 1714: Permit an empty string as expansion result for transport
462 option transport_filter, meaning no filtering.
464 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
466 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
467 defaults to "*" (all hosts). The variable is now available when not built
468 with TLS, default unset, mainly to enable keeping the testsuite sane.
469 If a server certificate is not supplied (via tls_certificate) an error is
470 logged, and clients will find TLS connections fail on startup. Presumably
471 they will retry in-clear.
472 Packagers of Exim are strongly encouraged to create a server certificate
473 at installation time.
475 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
476 with the $config_file variable.
478 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
479 in transport context, after the attempt, and per-recipient. The latter type
480 is per host attempted. The event data is the error message, and the errno
481 information encodes the lookup type (A vs. MX) used for the (first) host,
482 and the trailing two digits of the smtp 4xx response.
484 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
485 to write to mainlog (or rejectlog, paniclog) in the window between file
486 creation and permissions/ownership being changed. Particularly affects
487 installations where exicyclog is run as root, rather than exim user;
488 result is that the running daemon panics and dies.
490 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
492 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
493 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
494 "pri" and "weight". Note that the previous implicit priority given by the
495 list order is no longer honoured.
497 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
500 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
501 by defining SUPPORT_SOCKS.
503 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
504 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
505 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
506 variables are renamed to proxy_{local,external}_{address,port}.
508 JH/27 Move Internationalisation support from Experimental to mainline, enabled
509 for a build by defining SUPPORT_I18N
511 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
512 of the query string, and make ${quote_redis:} do that quoting.
514 JH/29 Move Events support from Experimental to mainline, enabled by default
515 and removable for a build by defining DISABLE_EVENT.
517 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
519 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
520 cached by the daemon.
522 JH/32 Move Redis support from Experimental to mainline, enabled for a build
523 by defining LOOKUP_REDIS. The libhiredis library is required.
525 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
526 keys are given for lookup.
528 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
529 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
530 only supported when built with TLS support. The PolarSSL SHA routines
531 are still used when the TLS library is too old for convenient support.
533 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
534 openssl_options), for security. OpenSSL forces this from version 1.1.0
535 server-side so match that on older versions.
537 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
538 allocation for $value could be released as the expansion processing
539 concluded, but leaving the global pointer active for it.
541 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
542 and to use the domains and local_parts ACL conditions.
544 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
545 incorrectly not doubled on cutthrough transmission, hence seen as a
546 body-termination at the receiving system - resulting in truncated mails.
547 Commonly the sender saw a TCP-level error, and retransmitted the message
548 via the normal store-and-forward channel. This could result in duplicates
549 received - but deduplicating mailstores were liable to retain only the
550 initial truncated version.
552 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
554 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
556 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
557 we're in there, support oversigning also; bug 1309.
559 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
561 HS/04 Add support for keep_environment and add_environment options.
563 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
564 either intentional arithmetic overflow during PRNG, or testing config-
567 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
568 delivery resulted in actual delivery. Cancel cutthrough before DATA
571 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
572 rejects a recipient: pass the reject to the originator.
574 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
575 Many were false-positives and ignorable, but it's worth fixing the
578 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
579 for the new environment-manipulation done at startup. Move the routines
580 from being local to tls.c to being global via the os.c file.
582 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
583 an extract embedded as result-arg for a map, the first arg for extract
584 is unavailable so we cannot tell if this is a numbered or keyed
585 extraction. Accept either.
591 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
594 JH/02 The smtp transport option "multi_domain" is now expanded.
596 JH/03 The smtp transport now requests PRDR by default, if the server offers
599 JH/04 Certificate name checking on server certificates, when exim is a client,
600 is now done by default. The transport option tls_verify_cert_hostnames
601 can be used to disable this per-host. The build option
602 EXPERIMENTAL_CERTNAMES is withdrawn.
604 JH/05 The value of the tls_verify_certificates smtp transport and main options
605 default to the word "system" to access the system default CA bundle.
606 For GnuTLS, only version 3.0.20 or later.
608 JH/06 Verification of the server certificate for a TLS connection is now tried
609 (but not required) by default. The verification status is now logged by
610 default, for both outbound TLS and client-certificate supplying inbound
613 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
616 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
617 Status Notification (bounce) messages are now MIME format per RFC 3464.
618 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
619 under the control of the dsn_advertise_hosts option, and routers may
620 have a dsn_lasthop option.
622 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
623 default, modifiable by a malware= option. The list separator for
624 the options can now be changed in the usual way. Bug 68.
626 JH/10 The smtp_receive_timeout main option is now expanded before use.
628 JH/11 The incoming_interface log option now also enables logging of the
629 local interface on delivery outgoing connections.
631 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
632 if the interface and destination host and port all match.
634 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
637 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
638 Patch from Andrew Lewis.
640 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
641 now supports optional time-restrictions, weighting, and priority
642 modifiers per server. Patch originally by <rommer@active.by>.
644 JH/16 The spamd_address main option now supports a mixed list of local
645 and remote servers. Remote servers can be IPv6 addresses, and
646 specify a port-range.
648 JH/17 Bug 68: The spamd_address main option now supports an optional
649 timeout value per server.
651 JH/18 Bug 1581: Router and transport options headers_add/remove can
652 now have the list separator specified.
654 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
657 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
660 JH/21 Support for the A6 type of dns record is withdrawn.
662 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
663 rather than the verbs used.
665 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
666 from 255 to 1024 chars.
668 JH/24 Verification callouts now attempt to use TLS by default.
670 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
671 are generic router options now. The defaults didn't change.
673 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
674 Original patch from Alexander Shikoff, worked over by JH.
676 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
677 files if ZCAT_COMMAND is not executable.
679 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
681 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
683 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
684 Normally benign, it bites when the pair was led to by a CNAME;
685 modern usage is to not canonicalize the domain to a CNAME target
686 (and we were inconsistent anyway for A-only vs AAAA+A).
688 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
690 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
691 when evaluating $sender_host_dnssec.
693 JH/31 Check the HELO verification lookup for DNSSEC, adding new
694 $sender_helo_dnssec variable.
696 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
698 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
700 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
702 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
703 documented as working, but never had. Support all but $spam_report.
705 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
706 added for tls authenticator.
708 HS/03 Add perl_taintmode main config option
714 TL/01 When running the test suite, the README says that variables such as
715 no_msglog_check are global and can be placed anywhere in a specific
716 test's script, however it was observed that placement needed to be near
717 the beginning for it to behave that way. Changed the runtest perl
718 script to read through the entire script once to detect and set these
719 variables, reset to the beginning of the script, and then run through
720 the script parsing/test process like normal.
722 TL/02 The BSD's have an arc4random API. One of the functions to induce
723 adding randomness was arc4random_stir(), but it has been removed in
724 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
725 function when detected.
727 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
728 cause callback expansion.
730 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
731 syntax errors in an expansion can be treated as a string instead of
732 logging or causing an error, due to the internal use of bool_lax
733 instead of bool when processing it.
735 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
736 server certificates when making smtp deliveries.
738 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
740 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
742 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
744 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
745 Merged patch from Sebastian Wiedenroth.
747 JH/05 Fix results-pipe from transport process. Several recipients, combined
748 with certificate use, exposed issues where response data items split
749 over buffer boundaries were not parsed properly. This eventually
750 resulted in duplicates being sent. This issue only became common enough
751 to notice due to the introduction of connection certificate information,
752 the item size being so much larger. Found and fixed by Wolfgang Breyha.
754 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
755 size buffer was used, resulting in syntax errors when an expansion
758 JH/07 Add support for directories of certificates when compiled with a GnuTLS
759 version 3.3.6 or later.
761 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
762 is EXPERIMENTAL_EVENT, the main-configuration and transport options
763 both become "event_action", the variables become $event_name, $event_data
764 and $event_defer_errno. There is a new variable $verify_mode, usable in
765 routers, transports and related events. The tls:cert event is now also
766 raised for inbound connections, if the main configuration event_action
769 TL/06 In test suite, disable OCSP for old versions of openssl which contained
770 early OCSP support, but no stapling (appears to be less than 1.0.0).
772 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
773 server certificate names available under the smtp transport option
774 "tls_verify_cert_hostname" now do not permit multi-component wildcard
777 JH/10 Time-related extraction expansions from certificates now use the main
778 option "timezone" setting for output formatting, and are consistent
779 between OpenSSL and GnuTLS compilations. Bug 1541.
781 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
782 encoded parameter in the incoming message. Bug 1558.
784 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
785 include certificate info, eximon was claiming there were spoolfile
788 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
790 JH/14 Log delivery-related information more consistently, using the sequence
791 "H=<name> [<ip>]" wherever possible.
793 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
794 are problematic for Debian distribution, omit them from the release
797 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
799 JH/16 Fix string representation of time values on 64bit time_t architectures.
802 JH/17 Fix a null-indirection in certextract expansions when a nondefault
803 output list separator was used.
808 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
809 checkers that were complaining about end of non-void function with no
812 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
813 This was a regression introduced in 4.83 by another bugfix.
815 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
817 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
818 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
824 TF/01 Correctly close the server side of TLS when forking for delivery.
826 When a message was received over SMTP with TLS, Exim failed to clear up
827 the incoming connection properly after forking off the child process to
828 deliver the message. In some situations the subsequent outgoing
829 delivery connection happened to have the same fd number as the incoming
830 connection previously had. Exim would try to use TLS and fail, logging
831 a "Bad file descriptor" error.
833 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
834 utilities have not been installed.
836 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
837 temporary space as the ACL may create new global variables.
839 TL/01 LDAP support uses per connection or global context settings, depending
840 upon the detected version of the libraries at build time.
842 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
843 to extract and use the src ip:port in logging and expansions as if it
844 were a direct connection from the outside internet. PPv2 support was
845 updated based on HAProxy spec change in May 2014.
847 JH/02 Add ${listextract {number}{list}{success}{fail}}.
849 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
850 Properly escape header and check for NULL return.
852 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
855 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
857 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
858 characters in header names, implemented as a verify condition.
859 Contributed by Michael Fischer v. Mollard.
861 TL/05 Rename SPF condition results err_perm and err_temp to standardized
862 results permerror and temperror. Previous values are deprecated but
863 still accepted. In a future release, err_perm and err_temp will be
864 completely removed, which will be a backward incompatibility if the
865 ACL tests for either of these two old results. Patch contributed by
866 user bes-internal on the mailing list.
868 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
870 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
871 selectors, in both main and reject logs.
873 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
876 JH/07 Add malware type "sock" for talking to simple daemon.
878 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
880 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
881 routers/transports under cutthrough routing.
883 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
884 numbers. Touch up "bool" conditional to keep the same definition.
886 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
888 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
890 JH/12 Expand items in router/transport headers_add or headers_remove lists
891 individually rather than the list as a whole. Bug 1452.
893 Required for reasonable handling of multiple headers_ options when
894 they may be empty; requires that headers_remove items with embedded
895 colons must have them doubled (or the list-separator changed).
897 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
898 view the policy declared in the DMARC record. Currently, $dmarc_status
899 is a combined value of both the record presence and the result of the
902 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
904 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
905 dnslookup router and the smtp transport (applying to the forward
908 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
909 of ldap servers used for a specific lookup. Patch provided by Heiko
912 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
913 New variable $lookup_dnssec_authenticated for observability.
915 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
916 Patch submitted by Lars Timman.
918 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
920 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
921 Requires trusted mode and valid format message id, aborts otherwise.
922 Patch contributed by Heiko Schlichting.
924 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
925 certextract with support for various fields. Bug 1358.
927 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
928 is requested by default, modifiable by smtp transport option
931 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
932 operate on certificate variables to give certificate fingerprints
933 Also new ${sha256:cert_variable}.
935 JH/23 The PRDR feature is moved from being Experimental into the mainline.
937 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
938 Christian Aistleitner.
940 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
942 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
943 file. Patch from Wolfgang Breyha.
945 JH/25 Expand the coverage of the delivery $host and $host_address to
946 client authenticators run in verify callout. Bug 1476.
948 JH/26 Port service names are now accepted for tls_on_connect_ports, to
949 align with daemon_smtp_ports. Bug 72.
951 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
952 support and error reporting did not work properly.
954 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
955 and is readable. Patch from Andrew Colin Kissa.
957 TL/14 Enhance documentation of ${run expansion and how it parses the
958 commandline after expansion, particularly in the case when an
959 unquoted variable expansion results in an empty value.
961 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
963 PP/02 Fix internal collision of T_APL on systems which support RFC3123
964 by renaming away from it. Addresses GH issue 15, reported by
967 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
969 TL/15 SECURITY: prevent double expansion in math comparison functions
970 (can expand unsanitized data). Not remotely exploitable.
977 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
979 PP/02 Make -n do something, by making it not do something.
980 When combined with -bP, the name of an option is not output.
982 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
985 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
986 $sender_host_name and config options to manage this, and basic check
989 PP/05 DSCP support for outbound connections and control modifier for inbound.
991 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
992 (Only plugin which currently uses this is kerberos4, which nobody should
993 be using, but we should make it available and other future plugins might
994 conceivably use it, even though it would break NAT; stuff *should* be
995 using channel bindings instead).
997 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
998 name; added for Sendmail compatibility; requires admin caller.
999 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1000 just ignore it); requires trusted caller.
1001 Also parse but ignore: -Ac -Am -X<logfile>
1004 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1006 TL/02 Add +smtp_confirmation as a default logging option.
1008 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1009 Patch by Magnus Holmgren from 2007-02-20.
1011 TL/04 Bugzilla 1281 - Spec typo.
1012 Bugzilla 1283 - Spec typo.
1013 Bugzilla 1290 - Spec grammar fixes.
1015 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1017 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1019 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1020 mailing list by by Dmitry Isaikin.
1022 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1024 JH/02 Support "G" suffix to numbers in ${if comparisons.
1026 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1028 NM/01 Bugzilla 1197 - Spec typo
1029 Bugzilla 1196 - Spec examples corrections
1031 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1033 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1034 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1037 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1038 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1040 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1041 "acl {{name}{arg}...}", and optional args on acl condition
1044 JH/05 Permit multiple router/transport headers_add/remove lines.
1046 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1048 JH/07 Avoid using a waiting database for a single-message-only transport.
1049 Performance patch from Paul Fisher. Bugzilla 1262.
1051 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1054 JH/09 Add $headers_added variable, with content from use of ACL modifier
1055 add_header (but not yet added to the message). Bugzilla 199.
1057 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1058 Pulled from Bugzilla 817 by Wolfgang Breyha.
1060 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1062 (nb: this is the same fix as in Exim 4.80.1)
1064 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1067 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1070 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1072 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1073 advertise SMTP AUTH mechanism to us, instead of a generic
1074 protocol violation error. Also, make Exim more robust to bad
1075 data from the Dovecot auth socket.
1077 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1079 When a queue runner is handling a message, Exim first routes the
1080 recipient addresses, during which it prunes them based on the retry
1081 hints database. After that it attempts to deliver the message to
1082 any remaining recipients. It then updates the hints database using
1085 So if a recipient address works intermittently, it can get repeatedly
1086 deferred at routing time. The retry hints record remains fresh so the
1087 address never reaches the final cutoff time.
1089 This is a fairly common occurrence when a user is bumping up against
1090 their storage quota. Exim had some logic in its local delivery code
1091 to deal with this. However it did not apply to per-recipient defers
1092 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1094 This change adds a proper retry rule check during routing so that the
1095 final cutoff time is checked against the message's age. We only do
1096 this check if there is an address retry record and there is not a
1097 domain retry record; this implies that previous attempts to handle
1098 the address had the retry_use_local_parts option turned on. We use
1099 this as an approximation for the destination being like a local
1100 delivery, as in LMTP.
1102 I suspect this new check makes the old local delivery cutoff check
1103 redundant, but I have not verified this so I left the code in place.
1105 TF/02 Correct gecos expansion when From: is a prefix of the username.
1107 Test 0254 submits a message to Exim with the header
1111 When I ran the test suite under the user fanf2, Exim expanded
1112 the header to contain my full name, whereas it should have added
1113 a Resent-Sender: header. It erroneously treats any prefix of the
1114 username as equal to the username.
1116 This change corrects that bug.
1118 GF/01 DCC debug and logging tidyup
1119 Error conditions log to paniclog rather than rejectlog.
1120 Debug lines prefixed by "DCC: " to remove any ambiguity.
1122 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1124 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1125 Bug spotted by Jeremy Harris; was flawed since initial commit.
1126 Would have resulted in OCSP responses post-SNI triggering an Exim
1127 NULL dereference and crash.
1129 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1131 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1132 Bug detection, analysis and fix by Samuel Thibault.
1133 Bugzilla 1331, Debian bug #698092.
1135 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1137 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1138 Server implementation by Todd Lyons, client by JH.
1139 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1140 config variable "prdr_enable" controls whether the server
1141 advertises the facility. If the client requests PRDR a new
1142 acl_data_smtp_prdr ACL is called once for each recipient, after
1143 the body content is received and before the acl_smtp_data ACL.
1144 The client is controlled by both of: a hosts_try_prdr option
1145 on the smtp transport, and the server advertisement.
1146 Default client logging of deliveries and rejections involving
1147 PRDR are flagged with the string "PRDR".
1149 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1150 fclose(). Diagnosis by Todd Lyons.
1152 PP/17 Update configure.default to handle IPv6 localhost better.
1153 Patch by Alain Williams (plus minor tweaks).
1156 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1157 This is now consistent with GnuTLS, and is now documented: the
1158 previous undocumented portable approach to treating the option as
1159 unset was to force an expansion failure. That still works, and
1160 an empty string is now equivalent.
1162 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1163 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1164 not performing validation itself.
1166 PP/20 Added force_command boolean option to pipe transport.
1167 Patch from Nick Koston, of cPanel Inc.
1169 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1172 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1174 PP/21 Fix eximon continuous updating with timestamped log-files.
1175 Broken in a format-string cleanup in 4.80, missed when I repaired the
1176 other false fix of the same issue.
1177 Report and fix from Heiko Schlichting.
1180 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1181 Report from Prashanth Katuri.
1183 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1184 It's SecureTransport, so affects any MacOS clients which use the
1185 system-integrated TLS libraries, including email clients.
1187 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1188 using a MIME ACL for non-SMTP local injection.
1189 Report and assistance in diagnosis by Warren Baker.
1191 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1193 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1195 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1196 last id that failed so it may be referenced in subsequent ACL's.
1198 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1201 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1202 ldap library initialization, allowing self-signed CA's to be
1203 used. Also properly sets require_cert option later in code by
1204 using NULL (global ldap config) instead of ldap handle (per
1205 session). Bug diagnosis and testing by alxgomz.
1207 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1208 the src/util/ subdirectory.
1210 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1211 renamed to Transport Post Delivery Action by Jeremy Harris, as
1214 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1215 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1216 redis_servers = needs to be configured which will be used by the redis
1217 lookup. Patch from Warren Baker, of The Packet Hub.
1219 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1221 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1222 hostname or reverse DNS when processing a host list. Used suggestions
1223 from multiple comments on this bug.
1225 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1227 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1228 Missed a few lines, added it to make the runtest require no keyboard
1231 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1232 contains upper case chars. Make router use caseful_local_part.
1234 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1235 support when GnuTLS has been built with p11-kit.
1241 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1243 This, or similar/improved, will also be change PP/11 of 4.82.
1249 PP/01 Handle short writes when writing local log-files.
1250 In practice, only affects FreeBSD (8 onwards).
1251 Bugzilla 1053, with thanks to Dmitry Isaikin.
1253 NM/01 Bugzilla 949 - Documentation tweak
1255 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1258 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1260 PP/02 Implemented gsasl authenticator.
1262 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1264 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1265 `pkg-config foo` for cflags/libs.
1267 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1268 with rest of GSASL and with heimdal_gssapi.
1270 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1271 `pkg-config foo` for cflags/libs for the TLS implementation.
1273 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1274 properties get this fed in as external SSF. A number of robustness
1275 and debugging improvements to the cyrus_sasl authenticator.
1277 PP/08 cyrus_sasl server now expands the server_realm option.
1279 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1280 Patch by Jeremy Harris.
1282 PP/10 Added dbmjz lookup type.
1284 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1286 PP/12 MAIL args handles TAB as well as SP, for better interop with
1287 non-compliant senders.
1288 Analysis and variant patch by Todd Lyons.
1290 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1291 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1292 Patch from Dirk Mueller <dmueller@suse.com>
1294 PP/13 tls_peerdn now print-escaped for spool files.
1295 Observed some $tls_peerdn in wild which contained \n, which resulted
1296 in spool file corruption.
1298 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1299 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1300 or write after TLS renegotiation, which otherwise led to messages
1303 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1304 as a tracking header (ie: a signed header comes before the signature).
1305 Patch from Wolfgang Breyha.
1307 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1308 comma-sep list; embedded commas doubled.
1310 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1312 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1314 Report and patch from Dmitry Banschikov.
1316 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1317 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1318 failures. We appear to now support TLS1.1+ with Exim.
1320 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1321 lets Exim select keys and certificates based upon TLS SNI from client.
1322 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1323 before an outbound SMTP session. New log_selector, +tls_sni.
1325 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1326 NULL dereference. Report and patch from Alun Jones.
1328 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1329 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1330 Not seeing resolver debug output on NetBSD, but suspect this is a
1331 resolver implementation change.
1333 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1334 Left warnings. Added "eximon gdb" invocation mode.
1336 PP/21 Defaulting "accept_8bitmime" to true, not false.
1338 PP/22 Added -bw for inetd wait mode support.
1340 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1341 locate the relevant includes and libraries. Made this the default.
1343 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1344 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1346 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1347 This may cause build issues on older platforms.
1349 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1350 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1351 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1352 Added SNI support via GnuTLS too.
1353 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1355 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1357 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1358 Applied second patch from Janne, implementing suggestion to default
1359 multiple-strings-in-record handling to match SPF spec.
1361 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1363 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1364 read-only, out of scope).
1365 Patch from Wolfgang Breyha, report from Stuart Northfield.
1367 PP/29 Fix three issues highlighted by clang analyser static analysis.
1368 Only crash-plausible issue would require the Cambridge-specific
1369 iplookup router and a misconfiguration.
1370 Report from Marcin Mirosław.
1372 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1374 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1375 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1376 As part of this, removing so much warning spew let me fix some minor
1377 real issues in debug logging.
1379 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1380 assignment on my part. Fixed.
1382 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1383 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1384 Janne Snabb (who went above and beyond: thank you).
1386 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1387 string otherwise requires a connection and a bunch more work and it's
1388 relatively easy to get wrong. Should also expose TLS library linkage
1391 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1392 64-bit ${eval} (JH/03).
1394 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1395 GNU libc to support some of the 64-bit stuff, should not lead to
1396 conflicts. Defined before os.h is pulled in, so if a given platform
1397 needs to override this, it can.
1399 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1400 protection layer was required, which is not implemented.
1401 Bugzilla 1254, patch from Wolfgang Breyha.
1403 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1404 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1405 tls_dhparam take prime identifiers. Also unbreak combination of
1406 OpenSSL+DH_params+TLSSNI.
1408 PP/39 Disable SSLv2 by default in OpenSSL support.
1414 PP/01 Solaris build fix for Oracle's LDAP libraries.
1415 Bugzilla 1109, patch from Stephen Usher.
1417 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1419 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1422 TF/02 Fix a couple more cases where we did not log the error message
1423 when unlink() failed. See also change 4.74-TF/03.
1425 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1426 lock up or crash if it happened to be inside a call to libc when it
1427 got a SIGUSR1 from exiwhat.
1429 The SIGUSR1 handler appends the current process status to the process
1430 log which is later printed by exiwhat. It used to use the general
1431 purpose logging code to do this, but several functions it calls are
1432 not safe for signals.
1434 The new output code in the SIGUSR1 handler is specific to the process
1435 log, and simple enough that it's easy to inspect for signal safety.
1436 Removing some special cases also simplifies the general logging code.
1437 Removing the spurious timestamps from the process log simplifies
1440 TF/04 Improved ratelimit ACL condition.
1442 The /noupdate option has been deprecated in favour of /readonly which
1443 has clearer semantics. The /leaky, /strict, and /readonly update modes
1444 are mutually exclusive. The update mode is no longer included in the
1445 database key; it just determines when the database is updated. (This
1446 means that when you upgrade Exim will forget old rate measurements.)
1448 Exim now checks that the per_* options are used with an update mode that
1449 makes sense for the current ACL. For example, when Exim is processing a
1450 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1451 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1452 must specify per_mail/readonly. If you omit the update mode it defaults to
1453 /leaky where that makes sense (as before) or /readonly where required.
1455 The /noupdate option is now undocumented but still supported for
1456 backwards compatibility. It is equivalent to /readonly except that in
1457 ACLs where /readonly is required you may specify /leaky/noupdate or
1458 /strict/noupdate which are treated the same as /readonly.
1460 A useful new feature is the /count= option. This is a generalization
1461 of the per_byte option, so that you can measure the throughput of other
1462 aggregate values. For example, the per_byte option is now equivalent
1463 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1465 The per_rcpt option has been generalized using the /count= mechanism
1466 (though it's more complicated than the per_byte equivalence). When it is
1467 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1468 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1469 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1470 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1471 Note that using per_rcpt with a non-readonly update mode in more than
1472 one ACL will cause the recipients to be double-counted. (The per_mail
1473 and per_byte options don't have this problem.)
1475 The handling of very low rates has changed slightly. If the computed rate
1476 is less than the event's count (usually one) then this event is the first
1477 after a long gap. In this case the rate is set to the same as this event's
1478 count, so that the first message of a spam run is counted properly.
1480 The major new feature is a mechanism for counting the rate of unique
1481 events. The new per_addr option counts the number of different
1482 recipients that someone has sent messages to in the last time period. It
1483 behaves like per_rcpt if all the recipient addresses are different, but
1484 duplicate recipient addresses do not increase the measured rate. Like
1485 the /count= option this is a general mechanism, so the per_addr option
1486 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1487 example, measure the rate that a client uses different sender addresses
1488 with the options per_mail/unique=$sender_address. There are further
1489 details in the main documentation.
1491 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1493 TF/06 Removed a few PCRE remnants.
1495 TF/07 Automatically extract Exim's version number from tags in the git
1496 repository when doing development or release builds.
1498 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1499 Bugzilla 879. Patch from Paul Fisher.
1501 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1502 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1505 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1507 PP/05 Variable $av_failed, true if the AV scanner deferred.
1508 Bugzilla 1078. Patch from John Horne.
1510 PP/06 Stop make process more reliably on build failure.
1511 Bugzilla 1087. Patch from Heiko Schlittermann.
1513 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1514 Bugzilla 1089. Patch from Heiko Schlittermann.
1516 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1517 Bugzilla 1131. Patch from Holger Weiß.
1519 PP/09 Handle IPv6 addresses with SPF.
1520 Bugzilla 860. Patch from Wolfgang Breyha.
1522 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1524 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1527 PP/11 match_* no longer expand right-hand-side by default.
1528 New compile-time build option, EXPAND_LISTMATCH_RHS.
1529 New expansion conditions, "inlist", "inlisti".
1531 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1533 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1535 PP/14 fix log_write() format string regression from TF/03.
1536 Bugzilla 1152. Patch from Dmitry Isaikin.
1542 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1544 PP/02 Harmonised TLS library version reporting; only show if debugging.
1545 Layout now matches that introduced for other libraries in 4.74 PP/03.
1547 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1549 PP/04 New "dns_use_edns0" global option.
1551 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1554 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1555 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1557 TK/01 Updated PolarSSL code to 0.14.2.
1558 Bugzilla 1097. Patch from Andreas Metzler.
1560 PP/07 Catch divide-by-zero in ${eval:...}.
1561 Fixes bugzilla 1102.
1563 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1566 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1567 format-string attack -- SECURITY: remote arbitrary code execution.
1569 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1570 time unintentionally subject to list matching rules, letting the header
1571 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1572 arbitrary string expansion). This allowed for information disclosure.
1574 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1575 INT_MIN/-1 -- value coerced to INT_MAX.
1581 NM/01 Workaround for PCRE version dependency in version reporting
1584 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1585 This fixes portability to compilers other than gcc, notably
1586 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1588 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1589 makefiles for portability to HP-UX and POSIX correctness.
1591 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1592 Also via indented variable definition in the Makefile.
1593 (Debugging by Oliver Heesakkers).
1595 PP/02 Restore caching of spamd results with expanded spamd_address.
1596 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1598 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1599 Improves build reliability. Fix from: Frank Elsner
1601 NM/02 Fix wide character breakage in the rfc2047 coding
1602 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1604 NM/03 Allow underscore in dnslist lookups
1605 Fixes bug 1026. Patch from Graeme Fowler
1607 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1608 Code patches from Adam Ciarcinski of NetBSD.
1610 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1613 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1614 is logged, to avoid truncation. Patch from John Horne.
1616 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1617 Patch from Jakob Hirsch.
1619 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1620 SQL string expansion failure details.
1621 Patch from Andrey Oktyabrski.
1623 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1624 Patch from Simon Arlott.
1626 PP/09 New lookups functionality failed to compile on old gcc which rejects
1627 extern declarations in function scope.
1628 Patch from Oliver Fleischmann
1630 PP/10 Use sig_atomic_t for flags set from signal handlers.
1631 Check getgroups() return and improve debugging.
1632 Fixed developed for diagnosis in bug 927 (which turned out to be
1635 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1636 Patch from Mark Zealey.
1638 PP/12 Bugzilla 1056: Improved spamd server selection.
1639 Patch from Mark Zealey.
1641 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1642 Based on patch from Heiko Schlittermann.
1644 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1645 Patch from Uwe Doering, sign-off by Michael Haardt.
1647 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1648 variable declaration deep within a block. Bug and patch from
1651 PP/15 lookups-Makefile IRIX compatibility coercion.
1653 PP/16 Make DISABLE_DKIM build knob functional.
1655 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1656 Patch by Simon Arlott
1658 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1659 variable argument macros. Our copy now differs from upstream.
1665 TF/01 Failure to get a lock on a hints database can have serious
1666 consequences so log it to the panic log.
1668 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1669 controlled using the smtp_confirmation log selector.
1671 TF/03 Include the error message when we fail to unlink a spool file.
1673 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1674 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1675 for maintaining out-of-tree patches for some time.
1677 PP/01 Bugzilla 139: Documentation and portability issues.
1678 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1679 Handle per-OS dynamic-module compilation flags.
1681 PP/02 Let /dev/null have normal permissions.
1682 The 4.73 fixes were a little too stringent and complained about the
1683 permissions on /dev/null. Exempt it from some checks.
1684 Reported by Andreas M. Kirchwitz.
1686 PP/03 Report version information for many libraries, including
1687 Exim version information for dynamically loaded libraries. Created
1688 version.h, now support a version extension string for distributors
1689 who patch heavily. Dynamic module ABI change.
1691 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1692 privilege escalation vulnerability whereby the Exim run-time user
1693 can cause root to append content of the attacker's choosing to
1696 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1699 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1700 If dropping privileges for untrusted macros, we disabled normal logging
1701 on the basis that it would fail; for the Exim run-time user, this is not
1702 the case, and it resulted in successful deliveries going unlogged.
1703 Fixed. Reported by Andreas Metzler.
1709 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1710 only prepend for the Resent-* case. Fixes regression introduced in
1711 Exim 4.70 by NM/22 for Bugzilla 607.
1713 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1714 increasing numbers of administrators be bitten by this.
1716 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1718 PP/03 Bugzilla 994: added openssl_options main configuration option.
1720 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1722 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1724 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1726 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1727 without a peer certificate, leading to a segfault because of an
1728 assumption that peers always have certificates. Be a little more
1729 paranoid. Problem reported by Martin Tscholak.
1731 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1732 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1733 NB: ClamAV planning to remove STREAM in "middle of 2010".
1734 CL also introduces -bmalware, various -d+acl logging additions and
1735 more caution in buffer sizes.
1737 PP/09 Implemented reverse_ip expansion operator.
1739 PP/10 Bugzilla 937: provide a "debug" ACL control.
1741 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1743 PP/12 Bugzilla 973: Implement --version.
1745 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1747 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1749 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1751 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1752 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1753 ignore trailing whitespace.
1755 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1757 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1760 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1761 Notification from Dr Andrew Aitchison.
1763 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1764 ExtendedDetectionInfo response format.
1765 Notification from John Horne.
1767 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1770 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1771 XSL and documented dependency on system catalogs, with examples of how
1774 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1777 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1778 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1779 configuration file which is writeable by the Exim user or group.
1781 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1782 of configuration files to cover files specified with the -C option if
1783 they are going to be used with root privileges, not just the default
1786 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1787 option (effectively making it always true).
1789 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1790 files to be used while preserving root privileges.
1792 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1793 that rogue child processes cannot use them.
1795 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1796 run-time user, instead of root.
1798 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1799 Exim run-time user without dropping privileges.
1801 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1802 result string, instead of calling string_vformat() twice with the same
1805 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1806 for other users. Others should always drop root privileges if they use
1807 -C on the command line, even for a whitelisted configure file.
1809 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1811 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1817 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1818 $data_path, and $header_path variables; fixed documentation bugs and
1821 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1822 exipick to access non-standard spools, including the "frozen" queue
1825 NM/01 Bugzilla 965: Support mysql stored procedures.
1826 Patch from Alain Williams
1828 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1830 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1831 Patch from Andreas Metzler
1833 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1834 Patch from Kirill Miazine
1836 NM/05 Bugzilla 671: Added umask to procmail example.
1838 JJ/03 installed exipick 20100323.0, fixing doc bug
1840 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1841 directory. Notification and patch from Dan Rosenberg.
1843 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1845 TK/02 Improve log output when DKIM signing operation fails.
1847 MH/01 Treat the transport option dkim_domain as a colon separated
1848 list, not as a single string, and sign the message with each element,
1849 omitting multiple occurences of the same signer.
1851 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1852 Bugzilla 985, 986. Patch by Simon Arlott
1854 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1855 Patch by Simon Arlott
1857 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1858 MBX locking. Notification from Dan Rosenberg.
1864 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1866 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1868 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1870 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1872 NM/04 Bugzilla 915: Flag broken perl installation during build.
1878 TK/01 Added patch by Johannes Berg that expands the main option
1879 "spamd_address" if it starts with a dollar sign.
1881 TK/02 Write list of recipients to X-Envelope-Sender header when building
1882 the mbox-format spool file for content scanning (suggested by Jakob
1885 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1886 (http://www.dcc-servers.net/) support via dccifd. Activated by
1887 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1889 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1890 by Mark Daniel Reidel <mr@df.eu>.
1892 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1893 When building exim an external PCRE library is now needed -
1894 PCRE is a system library on the majority of modern systems.
1895 See entry on PCRE_LIBS in EDITME file.
1897 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1898 conversation. Added nologin parameter to request.
1899 Patch contributed by Kirill Miazine.
1901 TF/01 Do not log submission mode rewrites if they do not change the address.
1903 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1905 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1906 log files in place. Contributed by Roberto Lima.
1908 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1910 TF/03 Bugzilla 615: When checking the local_parts router precondition
1911 after a local_part_suffix or local_part_prefix option, Exim now
1912 does not use the address's named list lookup cache, since this
1913 contains cached lookups for the whole local part.
1915 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1916 Robert Millan. Documentation is in experimental-spec.txt.
1918 TF/04 Bugzilla 668: Fix parallel build (make -j).
1920 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1922 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1923 Patch provided by Jan Srzednicki.
1925 TF/05 Leading white space used to be stripped from $spam_report which
1926 wrecked the formatting. Now it is preserved.
1928 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1929 that they are available at delivery time.
1931 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1933 TF/08 TLS error reporting now respects the incoming_interface and
1934 incoming_port log selectors.
1936 TF/09 Produce a more useful error message if an SMTP transport's hosts
1937 setting expands to an empty string.
1939 NM/06 Bugzilla 744: EXPN did not work under TLS.
1940 Patch provided by Phil Pennock.
1942 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1943 Patch provided by Richard Godbee.
1945 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1946 acl_smtp_notquit, added index entry.
1948 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1949 Patch provided by Eugene Bujak.
1951 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1952 accept(). Patch provided by Maxim Dounin.
1954 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1955 Patch provided by Phil Pennock.
1957 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1959 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1960 Patch provided by Brad "anomie" Jorsch.
1962 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1963 Patch provided by Dean Brooks.
1965 TK/05 Add native DKIM support (does not depend on external libraries).
1967 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1968 Patch provided by Graeme Fowler.
1970 NM/16 Bugzilla 851: Documentation example syntax fix.
1972 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1974 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1977 NM/19 Bugzilla 745: TLS version reporting.
1978 Patch provided by Phil Pennock.
1980 NM/20 Bugzilla 167: bool: condition support.
1981 Patch provided by Phil Pennock.
1983 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1984 clients. Patch provided by Phil Pennock.
1986 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1987 Patch provided by Brad "anomie" Jorsch.
1989 NM/23 Bugzilla 687: Fix misparses in eximstats.
1990 Patch provided by Heiko Schlittermann.
1992 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1993 Patch provided by Heiko Schlittermann.
1995 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1996 plus update to original patch.
1998 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2000 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2001 Patch provided by David Brownlee.
2003 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2005 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2007 NM/30 Bugzilla 888: TLS documentation bugfixes.
2009 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2011 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2012 Unlike the original bugzilla I have changed all shell scripts in src tree.
2014 NM/33 Bugzilla 898: Transport filter timeout fix.
2015 Patch by Todd Rinaldo.
2017 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2018 Patch by Serge Demonchaux.
2020 NM/35 Bugzilla 39: Base64 decode bug fixes.
2021 Patch by Jakob Hirsch.
2023 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2025 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2027 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2029 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2035 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2036 ALT-N's libdkim that I have put here:
2037 http://duncanthrax.net/exim-experimental/
2039 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2040 were called 'true' and it seems that C99 defines that as a reserved
2041 keyword to be used with 'bool' variable types. That means you could
2042 not include C99-style headers which use bools without triggering
2043 build errors in sieve.c.
2045 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2046 as mailq or other aliases. Changed the --help handling significantly
2047 to do whats expected. exim_usage() emits usage/help information.
2049 SC/01 Added the -bylocaldomain option to eximstats.
2051 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2053 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2055 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2061 PH/01 Another patch from the Sieve maintainer.
2063 PH/02 When an IPv6 address is converted to a string for single-key lookup
2064 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2065 dots are used instead of colons so that keys in lsearch files need not
2066 contain colons. This was done some time before quoting was made available
2067 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2068 (notated using the quote facility) so as to distinguish them from IPv4
2069 keys. This meant that lookups for IP addresses in host lists did not work
2070 for iplsearch lookups.
2072 This has been fixed by arranging for IPv6 addresses to be expressed with
2073 colons if the lookup type is iplsearch. This is not incompatible, because
2074 previously such lookups could never work.
2076 The situation is now rather anomalous, since one *can* have colons in
2077 ordinary lsearch keys. However, making the change in all cases is
2078 incompatible and would probably break a number of configurations.
2080 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2083 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2084 conversion specification without a maximum field width, thereby enabling
2085 a rogue spamd server to cause a buffer overflow. While nobody in their
2086 right mind would setup Exim to query an untrusted spamd server, an
2087 attacker that gains access to a server running spamd could potentially
2088 exploit this vulnerability to run arbitrary code as the Exim user.
2090 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2091 $primary_hostname instead of what libspf2 thinks the hosts name is.
2093 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2094 a directory entry by the name of the lookup key. Previously, if a
2095 symlink pointed to a non-existing file or a file in a directory that
2096 Exim lacked permissions to read, a lookup for a key matching that
2097 symlink would fail. Now it is enough that a matching directory entry
2098 exists, symlink or not. (Bugzilla 503.)
2100 PH/03 The body_linecount and body_zerocount variables are now exported in the
2103 PH/04 Added the $dnslist_matched variable.
2105 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2106 This means they are set thereafter only if the connection becomes
2109 PH/06 Added the client_condition to authenticators so that some can be skipped
2110 by clients under certain conditions.
2112 PH/07 The error message for a badly-placed control=no_multiline_responses left
2113 "_responses" off the end of the name.
2115 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2117 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2118 (without spaces) instead of just copying the configuration text.
2120 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2122 PH/11 Added $max_received_linelength.
2124 PH/12 Added +ignore_defer and +include_defer to host lists.
2126 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2127 way in which PCRE > 7.0 is built.
2129 PH/14 Implemented queue_only_load_latch.
2131 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2132 MAIL command. The effect was to mangle the value on 64-bit systems.
2134 PH/16 Another patch from the Sieve maintainer.
2136 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2138 PH/18 If a system quota error occurred while trying to create the file for
2139 a maildir delivery, the message "Mailbox is full" was not appended to the
2140 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2141 only to a quota excession during the actual writing of the file.
2143 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2144 characters?) which causes problems in log lines. The DN values are now
2145 passed through string_printing() before being added to log lines.
2147 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2148 and InterBase are left for another time.)
2150 PH/21 Added message_body_newlines option.
2152 PH/22 Guard against possible overflow in moan_check_errorcopy().
2154 PH/23 POSIX allows open() to be a macro; guard against that.
2156 PH/24 If the recipient of an error message contained an @ in the local part
2157 (suitably quoted, of course), incorrect values were put in $domain and
2158 $local_part during the evaluation of errors_copy.
2164 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2165 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2168 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2169 issue a MAIL command.
2171 PH/02 In an ACL statement such as
2173 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2175 if a client was not listed at all, or was listed with a value other than
2176 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2177 the condition was not true (as it should be), so access was not denied.
2178 The bug was that the ! inversion was incorrectly passed on to the second
2179 item. This has been fixed.
2181 PH/03 Added additional dnslists conditions == and =& which are different from
2182 = and & when the dns lookup returns more than one IP address.
2184 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2185 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2187 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2188 FSYNC, which compiles an option called disable_fsync that allows for
2189 bypassing fsync(). The documentation is heavily laced with warnings.
2191 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2193 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2194 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2195 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2196 including adding "make clean"; (3) Added -fPIC when compiling the test
2197 dynamically loaded module, to get rid of a warning.
2199 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2200 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2201 = 0s. The bug is otherwise harmless.
2203 PH/07 There was a bug in the dovecot authenticator such that the value of
2204 $auth1 could be overwritten, and so not correctly preserved, after a
2205 successful authentication. This usually meant that the value preserved by
2206 the server_setid option was incorrect.
2208 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2210 PH/09 Installed PCRE release 7.0.
2212 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2213 run for batched SMTP input. It is now run at the start of every message
2214 in the batch. While fixing this I discovered that the process information
2215 (output by running exiwhat) was not always getting set for -bs and -bS
2216 input. This is fixed, and it now also says "batched" for BSMTP.
2218 PH/11 Added control=no_pipelining.
2220 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2221 patch, slightly modified), and move the expansion of helo_data till after
2222 the connection is made in the smtp transport (so it can use these
2225 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2227 PH/14 Added log_selector = +pid.
2229 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2231 PH/16 Add ${if forany and ${if forall.
2233 PH/17 Added dsn_from option to vary the From: line in DSNs.
2235 PH/18 Flush SMTP output before performing a callout, unless control =
2236 no_callout_flush is set.
2238 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2239 was true (the default) a successful delivery failed to delete the retry
2240 item, thus causing premature timeout of the address. The bug is now
2243 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2245 PH/21 Long custom messages for fakedefer and fakereject are now split up
2246 into multiline responses in the same way that messages for "deny" and
2247 other ACL rejections are.
2249 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2250 with slight modification.
2252 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2253 draft, changing the syntax and factoring some duplicate code".
2255 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2256 for deliveries of the second and subsequent messages over the same SMTP
2259 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2260 ${reduce, with only minor "tidies".
2262 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2264 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2265 expansion side effects.
2267 PH/27 When a message times out after an over-quota error from an Exim-imposed
2268 quota, the bounce message says "mailbox is full". This message was not
2269 being given when it was a system quota that was exceeded. It now should
2272 MH/03 Made $recipients available in local_scan(). local_scan() already has
2273 better access to the recipient list through recipients_list[], but
2274 $recipients can be useful in postmaster-provided expansion strings.
2276 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2277 in the case of a MAIL command with additional options following the
2278 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2279 were accidentally chopped off.
2281 PH/29 SMTP synchronization checks are implemented when a command is read -
2282 there is a check that no more input is waiting when there shouldn't be
2283 any. However, for some commands, a delay in an ACL can mean that it is
2284 some time before the response is written. In this time, more input might
2285 arrive, invalidly. So now there are extra checks after an ACL has run for
2286 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2287 pipelining has not been advertised.
2289 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2291 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2292 returned in the SMTP response when smtp_return_error_details was false.
2293 This has been fixed.
2295 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2296 instead of the C I/O that was originally supplied, because problems were
2297 reported on Solaris.
2299 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2300 Exim which did not show up earlier: it was assuming that a call to
2301 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2302 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2303 was a macro that became an assignment, so it seemed to work. This has
2304 changed to a proper function call with a void return, hence the compile
2305 error. Exim's code has been fixed.
2307 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2310 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2312 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2314 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2315 because it turns out to be too big or there is a timeout) there is no
2316 "Completed" line in the log. When some messages of this type were
2317 selected by exigrep, they were listed as "not completed". Others were
2318 picked up by some special patterns. I have improved the selection
2319 criteria to be more general.
2321 PH/38 The host_find_failed option in the manualroute router can now be set
2322 to "ignore", to completely ignore a host whose IP address cannot be
2323 found. If all hosts are ignored, the behaviour is controlled by the new
2324 host_all_ignored option.
2326 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2327 homing or because of multiple MX records with /mx) generated more than
2328 one IP address, and the following item turned out to be the local host,
2329 all the secondary addresses of the first item were incorrectly removed
2330 from the list, along with the local host and any following hosts (which
2331 is what is supposed to happen).
2333 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2334 whoever called Exim into the -H file. In the case of the daemon it was
2335 behaving confusingly. When first started, it used values for whoever
2336 started the daemon, but after a SIGHUP it used the Exim user (because it
2337 calls itself on a restart). I have changed the code so that it now always
2340 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2341 message are rejected with the same error (e.g. no authentication or bad
2342 sender address), and a DATA command is nevertheless sent (as can happen
2343 with PIPELINING or a stupid MUA), the error message that was given to the
2344 RCPT commands is included in the rejection of the DATA command. This is
2345 intended to be helpful for MUAs that show only the final error to their
2348 PH/42 Another patch from the Sieve maintainer.
2350 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2351 Eximstats - Fixed some broken HTML links and added missing column headers
2353 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2354 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2356 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2358 PH/43 Yet another patch from the Sieve maintainer.
2360 PH/44 I found a way to check for a TCP/IP connection going away before sending
2361 the response to the final '.' that terminates a message, but only in the
2362 case where the client has not sent further data following the '.'
2363 (unfortunately, this is allowed). However, in many cases there won't be
2364 any further data because there won't be any more messages to send. A call
2365 to select() can be used: if it shows that the input is "ready", there is
2366 either input waiting, or the socket has been closed. An attempt to read
2367 the next input character can distinguish the two cases. Previously, Exim
2368 would have sent an OK response which the client would never have see.
2369 This could lead to message repetition. This fix should cure that, at
2370 least in a lot of common cases.
2372 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2373 advertised in response to EHLO.
2379 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2380 fixed by 4.65/MH/01 (is this a record?) are fixed:
2382 (i) An empty string was always treated as zero by the numeric comparison
2383 operators. This behaviour has been restored.
2385 (ii) It is documented that the numeric comparison operators always treat
2386 their arguments as decimal numbers. This was broken in that numbers
2387 starting with 0 were being interpreted as octal.
2389 While fixing these problems I realized that there was another issue that
2390 hadn't been noticed. Values of message_size_limit (both the global option
2391 and the transport option) were treated as octal if they started with 0.
2392 The documentation was vague. These values are now always treated as
2393 decimal, and I will make that clear in the documentation.
2399 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2400 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2403 MH/01 Don't check that the operands of numeric comparison operators are
2404 integers when their expansion is in "skipping" mode (fixes bug
2405 introduced by 4.64-PH/07).
2407 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2408 child addresses, Exim now panics and dies. Previously, because the count
2409 is held in a short int, deliveries were likely to be lost. As such a
2410 large number of recipients for a single message is ridiculous
2411 (performance will be very, very poor), I have chosen to impose a limit
2412 rather than extend the field.
2418 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2419 leftover -K file (the existence of which was triggered by #402).
2420 While we were at it, introduced process PID as part of the -K
2421 filename. This should rule out race conditions when creating
2424 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2425 processing considerably. Previous code took too long for large mails,
2426 triggering a timeout which in turn triggers #401.
2428 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2429 in the DK code in transports.c. sendfile() is not really portable,
2430 hence the _LINUX specificness.
2432 TF/01 In the add_headers option to the mail command in an Exim filter,
2433 there was a bug that Exim would claim a syntax error in any
2434 header after the first one which had an odd number of characters
2437 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2438 callout verification, Exim cached a "reject" for the entire domain. This
2439 is correct for most verifications, but it is not correct for a recipient
2440 verification with use_sender or use_postmaster set, because in that case
2441 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2442 case of MAIL FROM:<> rejection from other early rejections (e.g.
2443 rejection of HELO). When verifying a recipient using a non-null MAIL
2444 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2445 Whatever the result of the callout, the value of the domain cache is
2446 left unchanged (for any other kind of callout, getting as far as trying
2447 RCPT means that the domain itself is ok).
2449 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2452 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2453 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2454 session, the daemon ignores ECONNECT errors and logs others; it now
2455 ignores EPIPE as well.
2457 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2458 (quoted-printable decoding).
2460 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2461 later the small subsequent patch to fix an introduced bug.
2463 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2465 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2467 PH/08 An error is now given if message_size_limit is specified negative.
2469 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2470 to be given (somewhat) arbitrary names.
2472 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2475 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2476 miscellaneous code fixes
2478 PH/10 Added the log_reject_target ACL modifier to specify where to log
2481 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2482 hostname. This is wrong, because it relates to the incoming message (and
2483 probably the interface on which it is arriving) and not to the outgoing
2484 callout (which could be using a different interface). This has been
2485 changed to use the value of the helo_data option from the smtp transport
2486 instead - this is what is used when a message is actually being sent. If
2487 there is no remote transport (possible with a router that sets up host
2488 addresses), $smtp_active_hostname is used.
2490 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2491 tweaks were necessary in order to get it to work (see also 21 below):
2492 (a) The code assumed that strncpy() returns a negative number on buffer
2493 overflow, which isn't the case. Replaced with Exim's string_format()
2495 (b) There were several signed/unsigned issues. I just did the minimum
2496 hacking in of casts. There is scope for a larger refactoring.
2497 (c) The code used strcasecmp() which is not a standard C function.
2498 Replaced with Exim's strcmpic() function.
2499 (d) The code set only $1; it now sets $auth1 as well.
2500 (e) A simple test gave the error "authentication client didn't specify
2501 service in request". It would seem that Dovecot has changed its
2502 interface. Fortunately there's a specification; I followed it and
2503 changed what the client sends and it appears to be working now.
2505 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2508 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2509 address A is aliased to B and C, where B exists and C does not. Without
2510 -v the output is "A verified" because verification stops after a
2511 successful redirection if more than one address is generated. However,
2512 with -v the child addresses are also verified. Exim was outputting "A
2513 failed to verify" and then showing the successful verification for C,
2514 with its parentage. It now outputs "B failed to verify", showing B's
2515 parentage before showing the successful verification of C.
2517 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2518 look up a TXT record in a specific list after matching in a combined
2521 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2522 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2523 they consult the DNS. I had assumed they would set it the way they
2524 wanted; and indeed my experiments on Linux seem to show that in some
2525 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2526 To be on the safe side, however, I have now made the interface to
2527 host_find_byname() similar to host_find_bydns(), with an argument
2528 containing the DNS resolver options. The host_find_byname() function now
2529 sets these options at its start, just as host_find_bydns() does. The smtp
2530 transport options dns_qualify_single and dns_search_parents are passed to
2531 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2532 of host_find_byname() use the default settings of RES_DEFNAMES
2533 (qualify_single) but not RES_DNSRCH (search_parents).
2535 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2536 spool_read_header() do less string testing, by means of a preliminary
2537 switch on the second character of optional "-foo" lines. (This is
2538 overdue, caused by the large number of possibilities that now exist.
2539 Originally there were few.) While I was there, I also converted the
2540 str(n)cmp tests so they don't re-test the leading "-" and the first
2541 character, in the hope this might squeeze out yet more improvement.
2543 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2544 flag allowing group syntax was set by the header_syntax check but not
2545 turned off, possible causing trouble later; (2) The flag was not being
2546 set at all for the header_verify test, causing "group"-style headers to
2547 be rejected. I have now set it in this case, and also caused header_
2548 verify to ignore an empty address taken from a group. While doing this, I
2549 came across some other cases where the code for allowing group syntax
2550 while scanning a header line wasn't quite right (mostly, not resetting
2551 the flag correctly in the right place). These bugs could have caused
2552 trouble for malformed header lines. I hope it is now all correct.
2554 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2555 with the "reply" argument non-NULL. The code, however (which originally
2556 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2557 but it didn't always do it. This confused somebody who was copying the
2558 code for some other use. I have removed all the tests.
2560 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2561 feature that was used to support insecure browsers during the U.S. crypto
2562 embargo. It requires special client support, and Exim is probably the
2563 only MTA that supported it -- and would never use it because real RSA is
2564 always available. This code has been removed, because it had the bad
2565 effect of slowing Exim down by computing (never used) parameters for the
2566 RSA_EXPORT functionality.
2568 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2569 authenticator to fail if there's a tab character in the incoming data
2570 (there should never be unless someone is messing about, as it's supposed
2571 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2572 if the connection is using TLS or if the remote IP is the same as the
2573 local IP, and the "valid-client-cert option" if a client certificate has
2576 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2577 authenticators. This can be used for authorization after authentication
2578 succeeds. (In the case of plaintext, it servers for both authentication
2581 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2582 if any retry times were supplied.
2584 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2585 connection, where there is no client IP address to check. In this
2586 situation, the verify now always succeeds.
2588 PH/25 Applied John Jetmore's -Mset patch.
2590 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2592 PH/27 In a string expansion for a processed (not raw) header when multiple
2593 headers of the same name were present, leading whitespace was being
2594 removed from all of them, but trailing whitespace was being removed only
2595 from the last one. Now trailing whitespace is removed from each header
2596 before concatenation. Completely empty headers in a concatenation (as
2597 before) are ignored.
2599 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2600 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2602 PH/29 [Removed. This was a change that I later backed out, and forgot to
2603 correct the ChangeLog entry (that I had efficiently created) before
2604 committing the later change.]
2606 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2607 address errors (4xx response to RCPT) over the same connection as other
2608 messages routed to the same hosts. Such deliveries are always "forced",
2609 so retry times are not inspected. This resulted in far too many retries
2610 for the affected addresses. The effect occurred only when there were more
2611 hosts than the hosts_max_try setting in the smtp transport when it had
2612 the 4xx errors. Those hosts that it had tried were not added to the list
2613 of hosts for which the message was waiting, so if all were tried, there
2614 was no problem. Two fixes have been applied:
2616 (i) If there are any address or message errors in an SMTP delivery, none
2617 of the hosts (tried or untried) are now added to the list of hosts
2618 for which the message is waiting, so the message should not be a
2619 candidate for sending over the same connection that was used for a
2620 successful delivery of some other message. This seems entirely
2621 reasonable: after all the message is NOT "waiting for some host".
2622 This is so "obvious" that I'm not sure why it wasn't done
2623 previously. Hope I haven't missed anything, but it can't do any
2624 harm, as the worst effect is to miss an optimization.
2626 (ii) If, despite (i), such a delivery is accidentally attempted, the
2627 routing retry time is respected, so at least it doesn't keep
2628 hammering the server.
2630 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2631 in ${readsocket because some servers need this prod.
2633 PH/32 Added some extra debug output when updating a wait-xxx database.
2635 PH/33 The hint "could be header name not terminated by colon", which has been
2636 given for certain expansion errors for a long time, was not being given
2637 for the ${if def:h_colon_omitted{... case.
2639 PH/34 The spec says: "With one important exception, whenever a domain list is
2640 being scanned, $domain contains the subject domain." There was at least
2641 one case where this was not true.
2643 PH/35 The error "getsockname() failed: connection reset by peer" was being
2644 written to the panic log as well as the main log, but it isn't really
2645 panic-worthy as it just means the connection died rather early on. I have
2646 removed the panic log writing for the ECONNRESET error when getsockname()
2649 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2650 runs only) independently of the message's sender address. This meant
2651 that, if the 4xx error was in fact related to the sender, a different
2652 message to the same recipient with a different sender could confuse
2653 things. In particular, this can happen when sending to a greylisting
2654 server, but other circumstances could also provoke similar problems.
2655 I have changed the default so that the retry time for these errors is now
2656 based a combination of the sender and recipient addresses. This change
2657 can be overridden by setting address_retry_include_sender=false in the
2660 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2661 remote server are returned as part of bounce messages. This was not
2662 happening for LMTP over a pipe (the lmtp transport), but now it is the
2663 same for both kinds of LMTP.
2665 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2666 in header lines that were in fact CNAMEs. This is no longer the case.
2668 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2669 and queue runs started by the daemon processed all messages. This has
2670 been fixed so that -R and -S can now usefully be given with -q<time>.
2672 PH/40 Import PCRE release 6.7 (fixes some bugs).
2674 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2676 PH/42 Give an error if -q is specified more than once.
2678 PH/43 Renamed the variables $interface_address and $interface_port as
2679 $received_ip_address and $received_port, to make it clear that these
2680 values apply to message reception, and not to the outgoing interface when
2681 a message is delivered. (The old names remain recognized, of course.)
2683 PH/44 There was no timeout on the connect() call when using a Unix domain
2684 socket in the ${readsocket expansion. There now is.
2686 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2687 be meaningful with "accept".
2689 SC/01 Eximstats V1.43
2690 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2692 SC/02 Eximstats V1.44
2693 Use a glob alias rather than an array ref in the generated
2694 parser. This improves both readability and performance.
2696 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2697 Collect SpamAssassin and rejection statistics.
2698 Don't display local sender or destination tables unless
2699 there is data to show.
2700 Added average volumes into the top table text output.
2702 SC/04 Eximstats V1.46
2703 Collect data on the number of addresses (recipients)
2704 as well as the number of messages.
2706 SC/05 Eximstats V1.47
2707 Added 'Message too big' to the list of mail rejection
2708 reasons (thanks to Marco Gaiarin).
2710 SC/06 Eximstats V1.48
2711 Mainlog lines which have GMT offsets and are too short to
2712 have a flag are now skipped.
2714 SC/07 Eximstats V1.49 (Alain Williams)
2715 Added the -emptyok flag.
2717 SC/08 Eximstats V1.50
2718 Fixes for obtaining the IP address from reject messages.
2720 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2721 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2722 whitespace changes from 4.64-PH/27
2724 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2727 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2728 are found, allow negative numbers in numeric criteria)
2730 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2732 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2733 to match changes made in 4.64-PH/43
2735 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2737 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2738 "domain" is actually an IP address, and force a failure. This locks out
2739 those revolvers/nameservers that support "A-for-A" lookups, in
2740 contravention of the specifications.
2742 PH/48 When a host name was looked up from an IP address, and the subsequent
2743 forward lookup of the name timed out, the host name was left in
2744 $sender_host_name, contrary to the specification.
2746 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2747 restricted to single-key lookups, Exim was not diagnosing an error if
2748 * or *@ was used with a query-style lookup.
2750 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2752 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2753 long ago, but noone interested enough thought of it. Let's just say that
2754 the "1.1" means that there are some new functions that weren't there at
2755 some point in the past.
2757 PH/51 Error processing for expansion failure of helo_data from an smtp
2758 transport during callout processing was broken.
2760 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2761 tested/used via the -bh/-bhc/-bs options.
2763 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2764 bug, fixed in subsequent PCRE releases).
2766 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2767 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2769 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2775 SC/01 Use a glob alias rather than an array ref in eximstats generated
2776 parser. This improves both readability and performance.
2778 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2779 Don't display local sender or destination tables in eximstats unless
2780 there is data to show.
2781 Added average volumes into the eximstats top table text output.
2783 SC/03 Collect data on the number of addresses (recipients) as well
2784 as the number of messages in eximstats.
2786 TF/01 Correct an error in the documentation for the redirect router. Exim
2787 does (usually) call initgroups() when daemonizing.
2789 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2790 with consistent privilege compared to when running as a daemon.
2792 TF/03 Note in the spec that $authenticated_id is not set for local
2793 submissions from trusted users.
2795 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2796 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2798 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2799 by adding some example configuration directives to the default
2800 configuration file. A little bit of work is required to uncomment the
2801 directives and define how usernames and passwords are checked, but
2802 there is now a framework to start from.
2804 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2805 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2806 without this. I don't know how relevant this is to other LDAP libraries.
2808 PH/02 Add the verb name to the "unknown ACL verb" error.
2810 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2812 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2814 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2815 directory not expanded when it should be if an expanded home directory
2816 was set for the address (which is overridden by the transport).
2818 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2821 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2822 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2823 because it is too late at that time, and has no effect.
2825 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2826 security issue with \' (bugzilla #107). I could not use the
2827 PQescapeStringConn() function, because it needs a PGconn value as one of
2830 PH/08 When testing addresses using -bt, indicate those final addresses that
2831 are duplicates that would not cause an additional delivery. At least one
2832 person was confused, thinking that -bt output corresponded to deliveries.
2833 (Suppressing duplicates isn't a good idea as you lose the information
2834 about possibly different redirections that led to the duplicates.)
2836 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2837 systems where poll() doesn't work, in particular OS X.
2839 PH/10 Added more information to debugging output for retry time not reached.
2841 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2842 operations in malware.c.
2844 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2847 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2848 syslog_duplication set false, log lines that would normally be written
2849 both the the main log and to the reject log were not written to syslog at
2852 PH/14 In the default configuration, change the use of "message" in ACL warn
2853 statements to "add_header".
2855 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2856 not followed by a command (e.g. "seen endif").
2858 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2859 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2862 PH/17 Added extra conditions to the default value of delay_warning_condition
2866 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2867 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2868 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2871 The Auto-Submitted: and various List- headers are standardised, whereas I
2872 don't think Precedence: ever was.
2874 PH/18 Refactored debugging code in route_finduser() to show more information,
2875 in particular, the error code if getpwnam() issues one.
2877 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2878 This is apparently needed in addition to the PH/07 change above to avoid
2879 any possible encoding problems.
2881 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2882 but not after initializing Perl.
2884 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2885 output them only if debugging. By default they are written stderr,
2886 apparently, which is not desirable.
2888 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2891 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2894 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2896 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2897 authenticated or an ident call has been made. Suppress the default
2898 values for $authenticated_id and $authenticated_sender (but permit -oMai
2899 and -oMas) when testing with -bh.
2901 PH/24 Re-jigged the order of the tests in the default configuration so that the
2902 tests for valid domains and recipients precede the DNS black list and CSA
2903 tests, on the grounds that those ones are more expensive.
2905 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2906 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2907 This bug exists in every version of Exim that I still have, right back to
2910 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2911 However, an attempt to turn on case-sensitivity in a regex key by
2912 including (?-i) didn't work because the subject string was already
2913 lowercased, and the effects were non-intuitive. It turns out that a
2914 one-line patch can be used to allow (?-i) to work as expected.
2920 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2921 other effects) broke the use of negated acl sub-conditions.
2923 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2926 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2927 "Deny" causes Exim to reject the incoming connection with a 554 error.
2928 Unfortunately, if there is a major crisis, such as a disk failure,
2929 tcp-wrappers gives "deny", whereas what one would like would be some
2930 kind of temporary error. A kludge has been added to help with this.
2931 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2932 554 error is used if errno is still zero or contains ENOENT (which occurs
2933 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2936 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2938 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2939 errors. Otherwise a message that provokes a temporary error (when other
2940 messages do not) can cause a whole host to time out.
2942 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2943 addresses were routed directly to files or pipes from a redirect router.
2944 File deliveries just didn't batch; pipe deliveries might have suffered
2947 PH/06 A failure to get a lock for a hints database would erroneously always say
2948 "Failed to get write lock", even when it was really a read lock.
2950 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2951 of 0600. This has been changed to use the value of the lockfile_mode
2952 option (which defaults to 0600).
2954 PH/08 Applied small patch from the Sieve maintainer.
2956 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2957 folder from quota calculations, a direct delivery into this folder messed
2958 up the contents of the maildirsize file. This was because the regex was
2959 used only to exclude .Trash (or whatever) when the size of the mailbox
2960 was calculated. There was no check that a delivery was happening into an
2961 excluded directory. This bug has been fixed by ignoring all quota
2962 processing for deliveries into excluded directories.
2964 PH/10 Added the maildirfolder_create_regex option to appendfile.
2970 PH/01 The code for finding all the local interface addresses on a FreeBSD
2971 system running IPv6 was broken. This may well have applied to all BSD
2972 systems, as well as to others that have similar system calls. The broken
2973 code found IPv4 interfaces correctly, but gave incorrect values for the
2974 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2975 that it would not match correctly against @[] and not recognize the IPv6
2978 PH/02 The ipliteral router was not recognizing addresses of the form user@
2979 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2981 PH/03 Added disable_ipv6.
2983 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2984 decoded form, because it is most often used to construct To: headers
2985 lines in autoreplies, and the decoded form may well be syntactically
2986 invalid. However, $reply_address has leading white space removed, and all
2987 newlines turned into spaces so that the autoreply transport does not
2990 PH/05 If group was specified without a user on a router, and no group or user
2991 was specified on a transport, the group from the router was ignored.
2993 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2994 for visible compile-time settings that can be used to change these
2995 numbers, for those that want even more. Backwards compatibility with old
2996 spool files has been maintained. However, going back to a previous Exim
2997 release will lost any variables that are in spool files.
2999 PH/07 Two small changes when running in the test harness: increase delay when
3000 passing a TCP/IP connection to a new process, in case the original
3001 process has to generate a bounce, and remove special handling of
3002 127.0.0.2 (sic), which is no longer necessary.
3004 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3005 be the same on different OS.
3007 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3010 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3011 whether --show-vars was specified or not
3013 JJ/02 exipick: Added support for new ACL variable spool format introduced
3016 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3017 syntactically invalid From: or Reply-to: line, and a filter used this to
3018 generate an autoreply, and therefore failed to obtain an address for the
3019 autoreply, Exim could try to deliver to a non-existent relative file
3020 name, causing unrelated and misleading errors. What now happens is that
3021 it logs this as a hard delivery error, but does not attempt to create a
3024 PH/11 The exinext utility has a -C option for testing purposes, but although
3025 the given file was scanned by exinext itself; it wasn't being passed on
3026 when Exim was called.
3028 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3029 an end-of-file indication when reading a command response.
3031 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3032 compiled. In many other places in Exim, IPv6 addresses are always
3033 recognized, so I have changed this. It also means that IPv4 domain
3034 literals of the form [IPV4:n.n.n.n] are now always recognized.
3036 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3037 used if the router is not running as root, for example, when verifying at
3038 ACL time, or when using -bh. The debugging output from this situation was
3039 non-existent - all you got was a failure to exec. I have made two
3042 (a) Failures to set uid/gid, the current directory, or a process leader
3043 in a subprocess such as that created by queryprogram now generate
3044 suitable debugging output when -d is set.
3046 (b) The queryprogram router detects when it is not running as root,
3047 outputs suitable debugging information if -d is set, and then runs
3048 the subprocess without attempting to change uid/gid.
3050 PH/15 Minor change to Makefile for building test_host (undocumented testing
3053 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3054 additional section of a DNS packet that returns MX or SRV records.
3055 Instead, it always explicitly searches for A/AAAA records. This avoids
3056 major problems that occur when a DNS server includes only records of one
3057 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3058 fixed another bug: if SRV records were looked up and the corresponding
3059 address records were *not* found in the additional section, the port
3060 values from the SRV records were lost.
3062 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3063 using the correct key (the original address) when searching the retry
3064 rules in order to find which one to use for generating the retry hint.
3066 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3067 adding the default one. Similarly, if it contains a Reply-To: header, the
3068 errors_reply_to option, if set, is not used.
3070 PH/19 When calculating a retry time, Exim used to measure the "time since
3071 failure" by looking at the "first failed" field in the retry record. Now
3072 it does not use this if it is later than than the arrival time of the
3073 message. Instead it uses the arrival time. This makes for better
3074 behaviour in cases where some deliveries succeed, thus re-setting the
3075 "first failed" field. An example is a quota failure for a huge message
3076 when small messages continue to be delivered. Without this change, the
3077 "time since failure" will always be short, possible causing more frequent
3078 delivery attempts for the huge message than are intended.
3079 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3081 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3082 $1, $2, $3) because the numerical variables can be reset during some
3083 expansion items (e.g. "match"), thereby losing the authentication data.
3085 PH/21 Make -bV show the size of off_t variables so that the test suite can
3086 decide whether to run tests for quotas > 2G.
3088 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3089 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3090 filecount value is greater than 2G or if a quota value is greater than 2G
3091 on a system where the size of off_t is not greater than 4, a panic error
3094 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3095 never match. The debug and -bh output now contains an explicit error
3096 message indicating a malformed IPv4 address or mask.
3098 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3099 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3100 PH/23 above applies.
3102 PH/25 Do not write to syslog when running in the test harness. The only
3103 occasion when this arises is a failure to open the main or panic logs
3104 (for which there is an explicit test).
3106 PH/26 Added the /no_tell option to "control=freeze".
3108 PH/27 If a host name lookup failed very early in a connection, for example, if
3109 the IP address matched host_lookup and the reverse lookup yielded a name
3110 that did not have a forward lookup, an error message of the form "no IP
3111 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3112 could be logged. Now it outputs the IP address instead of "NULL".
3114 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3115 allows the sender and the authenticated sender to be set when
3116 submitting a message from within Exim. Since child_open_exim() is
3117 documented for local_scan(), the new function should be too.
3119 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3120 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3121 results in an empty string is now treated as unset.
3123 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3125 PH/31 Added sender_verify_fail logging option.
3127 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3128 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3129 tidied the source and removed it altogether.
3131 PH/33 When a queue run was abandoned because the load average was too high, a
3132 log line was always written; now it is written only if the queue_run log
3133 selector is set. In addition, the log line for abandonment now contains
3134 information about the queue run such as the pid. This is always present
3135 in "start" and "stop" lines but was omitted from the "abandon" line.
3137 PH/34 Omit spaces between a header name and the colon in the error message that
3138 is given when verify = headers_syntax fails (if there are lots of them,
3139 the message gets confusing).
3141 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3142 names, as there are now some PTR records that contain slashes. This check
3143 is only to protect against broken name servers that fall over on strange
3144 characters, so the fact that it applies to all lookups doesn't matter.
3146 PH/36 Now that the new test suite is complete, we can remove some of the
3147 special code in Exim that was needed for the old test suite. For example,
3148 sorting DNS records because real resolvers return them in an arbitrary
3149 order. The new test suite's fake resolver always returns records in the
3152 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3153 bounces) except when queue_only is set, to avoid logging races between
3154 the different processes.
3156 PH/38 Panic-die if .include specifies a non-absolute path.
3158 PH/39 A tweak to the "H" retry rule from its user.
3160 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3161 a label. They prevented compilation on older perls.
3163 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3164 a warning to be raised on newish perls.
3166 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3167 on queue. Changes to match documented behaviour of showing count of
3168 messages matching specified criteria.
3170 PH/40 Changed the default ident timeout from 30s to 5s.
3172 PH/41 Added support for the use of login_cap features, on those BSD systems
3173 that have them, for controlling the resources used by pipe deliveries.
3175 PH/42 The content-scanning code uses fopen() to create files in which to put
3176 message data. Previously it was not paying any attention to the mode of
3177 the files. Exim runs with umask(0) because the rest of the code creates
3178 files with open(), and sets the required mode explicitly. Thus, these
3179 files were ending up world-writeable. This was not a big issue, because,
3180 being within the spool directory, they were not world-accessible. I have
3181 created a function called modefopen, which takes an additional mode
3182 argument. It sets umask(777), creates the file, chmods it to the required
3183 mode, then resets the umask. All the relevant calls to fopen() in the
3184 content scanning code have been changed to use this function.
3186 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3187 to 24 hours. This avoids potential overflow problems when processing G
3188 and H retry rules. I suspect nobody ever tinkers with this value.
3190 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3192 PH/45 When the plaintext authenticator is running as a client, the server's
3193 challenges are checked to ensure they are valid base64 strings. By
3194 default, the authentication attempt is cancelled if an invalid string is
3195 received. Setting client_ignore_invalid_base64 true ignores these errors.
3196 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3197 they are received. Thus, the responses can be made to depend on the
3198 challenges. If an invalid string is ignored, an empty string is placed in
3201 PH/46 Messages that are created by the autoreply transport now contains a
3202 References: header, in accordance with RFCs 2822 and 3834.
3204 PH/47 Added authenticated_sender_force to the smtp transport.
3206 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3208 PH/49 Installed latest patch from the Sieve maintainer.
3210 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3211 was also set, the appendfile transport was unnecessarily scanning a
3212 directory of message files (e.g. for maildir delivery) to find the count
3213 of files (along with the size), even though it did not need this
3214 information. It now does the scan only if it needs to find either the
3215 size of the count of files.
3217 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3219 PH/52 Two bugs concerned with error handling when the smtp transport is
3222 (i) Exim was not creating retry information for temporary errors given
3223 for individual recipients after the DATA command when the smtp transport
3224 was used in LMTP mode. This meant that they could be retried too
3225 frequently, and not timed out correctly.
3227 (ii) Exim was setting the flag that allows error details to be returned
3228 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3229 recipients that were returned after the DATA command.
3231 PH/53 This is related to PH/52, but is more general: for any failing address,
3232 when detailed error information was permitted to be returned to the
3233 sender, but the error was temporary, then after the final timeout, only
3234 "retry timeout exceeded" was returned. Now it returns the full error as
3235 well as "retry timeout exceeded".
3237 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3238 do this, and (what is worse) MTAs that accept it.
3240 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3241 will now be deprecated.
3243 PH/56 New os.c-cygwin from the Cygwin maintainer.
3245 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3246 formats (previously only available in exim formats via -bpr, -bpru,
3247 and -bpra. Now also available in native and exiqgrep formats)
3249 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3250 with very large, slow to parse queues
3252 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3254 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3256 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3257 responses to authentication challenges, though it was showing the
3258 challenges; (ii) I've removed the CR characters from the debug output for
3261 PH/58 Allow for the insertion of a newline as well as a space when a string
3262 is turned into more than one encoded-word during RFC 2047 encoding. The
3263 Sieve code now uses this.
3265 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3266 data_4xx, lost_connection, tls_required.
3268 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3269 message was being sent as an SMTP response.
3271 PH/61 Add -l and -k options to exicyclog.
3273 PH/62 When verifying, if an address was redirected to one new address, so that
3274 verification continued, and the new address failed or deferred after
3275 having set something in $address_data, the value of $address_data was not
3276 passed back to the ACL. This was different to the case when no
3277 redirection occurred. The value is now passed back in both cases.
3279 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3280 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3281 use login_cap.h, so on its own it isn't the distinguishing feature. The
3282 new name refers directly to the setclassresources() function.
3284 PH/65 Added configuration files for NetBSD3.
3286 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3288 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3289 is preferred over IPv4.
3291 PH/68 The bounce_return_message and bounce_return_body options were not being
3292 honoured for bounces generated during the reception of non-SMTP messages.
3293 In particular, this applied to messages rejected by the ACL. This bug has
3294 been fixed. However, if bounce_return_message is true and bounce_return_
3295 body is false, the headers that are returned for a non-SMTP message
3296 include only those that have been read before the error was detected.
3297 (In the case of an ACL rejection, they have all been read.)
3299 PH/69 The HTML version of the specification is now built in a directory called
3300 spec_html instead of spec.html, because the latter looks like a path with
3301 a MIME-type, and this confuses some software.
3303 PH/70 Catch two compiler warnings in sieve.c.
3305 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3306 function verify_get_ident() calls ip_connect() to connect a socket, but
3307 if the "connect()" function timed out, ip_connect() used to close the
3308 socket. However, verify_get_ident() also closes the socket later, and in
3309 between Exim writes to the log, which may get opened at this point. When
3310 the socket was closed in ip_connect(), the log could get the same file
3311 descriptor number as the socket. This naturally causes chaos. The fix is
3312 not to close the socket in ip_connect(); the socket should be closed by
3313 the function that creates it. There was only one place in the code where
3314 this was missing, in the iplookup router, which I don't think anybody now
3315 uses, but I've fixed it anyway.
3317 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3318 well as to direct DNS lookups. Otherwise the handling of names in host
3319 lists is inconsistent and therefore confusing.
3325 PH/01 Two changes to the default runtime configuration:
3327 (1) Move the checks for relay_from_hosts and authenticated clients from
3328 after to before the (commented out) DNS black list checks.
3330 (2) Add control=submission to the relay_from_hosts and authenticated
3331 clients checks, on the grounds that messages accepted by these
3332 statements are most likely to be submissions.
3334 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3336 (1) Generate an error if the third argument for the ${prvs expansion is
3339 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3342 (3) Reset the variables that are obtained from the first argument of
3343 ${prvscheck and used in the second argument before leaving the code,
3344 because their memory is reclaimed, so using them afterwards may do
3347 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3348 one (it's much easier than Tom thought :-).
3350 (5) Because of (4), we can now allow for the use of $prvscheck_result
3351 inside the third argument.
3353 PH/03 For some reason, the default setting of PATH when running a command from
3354 a pipe transport was just "/usr/bin". I have changed it to
3357 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3358 anything to be listed in the output from -bV.
3360 PH/05 When a filter generated an autoreply, the entire To: header line was
3361 quoted in the delivery log line, like this:
3363 => >A.N.Other <ano@some.domain> <original@ddress> ...
3365 This has been changed so that it extracts the operative address. There
3366 may be more than one such address. If so, they are comma-separated, like
3369 => >ano@some.domain,ona@other.domain <original@ddress> ...
3371 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3372 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3373 not being looked up in the rDNS to get a host name, Exim was showing the
3374 IP address twice in Received: lines, even though the IP addresses were
3375 identical. For example:
3377 Received: from [1.2.3.4] (helo=[1.2.3.4])
3379 However, if the real host name was known, it was omitting the HELO data
3380 if it matched the actual IP address. This has been tidied up so that it
3381 doesn't show the same IP address twice.
3383 PH/07 When both +timestamp and +memory debugging was on, the value given by
3384 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3385 called by the debug printing, thereby overwriting the timestamp buffer.
3386 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3388 PH/08 When the original message was included in an autoreply transport, it
3389 always said "this is a copy of the message, including all the headers",
3390 even if body_only or headers_only was set. It now gives an appropriate
3393 PH/09 Applied a patch from the Sieve maintainer which:
3395 o fixes some comments
3396 o adds the (disabled) notify extension core
3397 o adds some debug output for the result of if/elsif tests
3398 o points to the current vacation draft in the documentation
3399 and documents the missing references header update
3403 o fixes a bug in processing the envelope test (when testing
3404 multiple envelope elements, the last element determined the
3407 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3408 Electronic Mail") by including:
3410 Auto-submitted: auto-generated
3412 in the messages that it generates (bounce messages and others, such as
3413 warnings). In the case of bounce messages for non-SMTP messages, there was
3414 also a typo: it was using "Auto_submitted" (underscore instead of
3415 hyphen). Since every message generated by Exim is necessarily in response
3416 to another message, thes have all been changed to:
3418 Auto-Submitted: auto-replied
3420 in accordance with these statements in the RFC:
3422 The auto-replied keyword:
3424 - SHOULD be used on messages sent in direct response to another
3425 message by an automatic process,
3427 - MUST NOT be used on manually-generated messages,
3429 - MAY be used on Delivery Status Notifications (DSNs) and Message
3430 Disposition Notifications (MDNs),
3432 - MUST NOT be used on messages generated by automatic or periodic
3433 processes, except for messages which are automatic responses to
3436 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3437 to the default Received: header definition.
3439 PH/12 Added log selector acl_warn_skipped (default on).
3441 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3442 variables because (a) they are in the wrong storage pool and (b) even if
3443 they were copied, it wouldn't work properly because of the caching.
3445 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3446 checking when decoding. Apparently there are clients that generate
3447 overlong encoded strings. Why am I not surprised?
3449 PH/15 If the first argument of "${if match_address" was not empty, but did not
3450 contain an "@" character, Exim crashed. Now it writes a panic log message
3451 and treats the condition as false.
3453 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3455 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3456 extension "envelope-auth". The code is finished and in agreement with
3457 other implementations, but there is no documentation so far and in fact,
3458 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3459 not changing the active code.
3461 Print executed "if" and "elsif" statements when debugging is used. This
3462 helps a great deal to understand what a filter does.
3464 Document more things not specified clearly in RFC3028. I had all this
3465 sorted out, when out of a sudden new issues came to my mind. Oops."
3467 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3470 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3471 canonical form (as documented). However, after a host name lookup from
3472 the IP address, check_host() was doing a simple string comparison with
3473 addresses acquired from the DNS when checking that the found name did
3474 have the original IP as one of its addresses. Since any found IPv6
3475 addresses are likely to be in abbreviated form, the comparison could
3476 fail. Luckily, there already exists a function for doing the comparison
3477 by converting both addresses to binary, so now that is used instead of
3478 the text comparison.
3480 PH/20 There was another similar case to PH/19, when a complete host name was
3481 given in a host list; looking up its IP address could give an abbreviated
3482 form, whereas the current host's name might or might not be abbreviated.
3483 The same fix has been applied.
3489 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3490 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3493 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3494 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3496 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3498 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3499 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3500 Also, for "Auto-Submitted", treat anything other than "no" as
3501 non-personal, in accordance with RFC 3834. (Previously it treated
3502 anything starting "auto-" as non-personal.)
3504 TF/01 The control=submission/name=... option had a problem with syntax
3505 errors if the name included a slash character. The /name= option
3506 now slurps the rest of the string, so it can include any characters
3507 but it must come last in the list of options (after /sender_retain
3510 PH/05 Some modifications to the interface to the fake nameserver for the new
3518 TK/01 Added the "success_on_redirect" address verification option. See
3519 NewStuff for rationale and an example.
3521 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3523 PH/02 Patch to exigrep to allow it to work on syslog lines.
3525 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3526 fread() to skip over the body file's header line, because in Cygwin the
3527 header line is locked and is inaccessible.
3529 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3530 co-exist for some time) to make it clear that it is the Exim ID that is
3531 referenced, not the Message-ID: header line.
3533 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3534 string_format() function, because snprintf() does not exist on all
3537 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3538 ${stat: expansion item.
3540 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3541 protocol synchronization error", to keep the pedants happy.
3543 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3544 well as for IRIX systems, when gcc is being used. See the host.c source
3547 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3549 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3552 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3553 transport and to the smtp transport in LMTP mode.
3555 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3557 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3558 run a filter in a subprocess. This could lead to confusion in subsequent
3559 lookups in the parent process. There should also be a search_tidyup() at
3560 the end of the subprocess.
3562 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3563 only if the host matched helo_try_verify_hosts, which caused the
3564 verification to occur when the EHLO/HELO command was issued. The ACL just
3565 tested the remembered result. Now, if a previous verification attempt has
3566 not happened, "verify = helo" does it there and then.
3568 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3570 TK/03 Fix log output including CR from clamd.
3572 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3573 exist provoked a memory error which could cause a segfault.
3575 PH/15 Installed PCRE 6.2
3577 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3579 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3580 of the problem. Specifically, suggested +O2 rather than +O1 for the
3583 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3585 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3586 record, the port was not correctly logged when the outgoing_port log
3587 selector was set (it logged the transort's default port).
3589 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3590 fallback_hosts, and "hosts" in the smtp transport.
3592 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3593 host errors such as "Connection refused".
3595 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3596 authentication with radiusclient 0.4.9:
3598 - Error returned from rc_read_config was caught wrongly
3599 - Username/password not passed on to radius server due to wrong length.
3601 The presumption is that some radiusclient API changes for 4.51/PH/17
3602 were not taken care of correctly. The code is still untested by me (my
3603 Linux distribution still has 0.3.2 of radiusclient), but it was
3604 contributed by a Radius user.
3606 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3607 expanding the "port" option of the smtp transport.
3609 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3610 while reading a MIME header. Thanks to Tom Hughes for a patch.
3612 PH/24 Include config.h inside local_scan.h so that configuration settings are
3615 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3616 that in an ACL for RCPT (for example), you can examine exactly what was
3619 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3620 commands, but it was not correctly comparing the address with the actual
3621 client host address. Thus, it would show the EHLO address in Received:
3622 header lines when this was not necessary.
3624 PH/27 Added the % operator to ${eval:}.
3626 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3627 it should be ignoring failures (because with -C, for example, it has lost
3628 privilege). It wasn't ignoring creation failures other than "already
3631 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3634 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3635 because an input line was too long, either on its own, or by virtue of
3636 too many continuations, the temporary file was not being removed, and the
3637 return code was incorrect.
3639 PH/31 Missing "BOOL" in function definition in filtertest.c.
3641 PH/32 Applied Sieve patches from the maintainer.
3643 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3645 PH/33 Added "verify = not_blind".
3647 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3648 Local/Makefile (with some defaults set). These are used in built scripts
3649 such as exicyclog, but they have never been used in the exim_install
3650 script (though there are many overriding facilities there). I have
3651 arranged that the exim_install script now takes note of these two
3654 PH/35 Installed configuration files for Dragonfly.
3656 PH/36 When a locally submitted message by a trusted user did not contain a
3657 From: header, and the sender address was obtained from -f or from an SMTP
3658 MAIL command, and the trusted user did not use -F to supply a sender
3659 name, $originator_name was incorrectly used when constructing a From:
3660 header. Furthermore, $originator_name was used for submission mode
3661 messages from external hosts without From: headers in a similar way,
3662 which is clearly wrong.
3664 PH/37 Added control=suppress_local_fixups.
3666 PH/38 When log_selector = +received_sender was set, and the addition of the
3667 sender made the log line's construction buffer exactly full, or one byte
3668 less than full, an overflow happened when the terminating "\n" was
3671 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3672 when the result of a list match is failure because a DNS lookup failed.
3674 PH/40 RM_COMMAND is now used in the building process.
3676 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3677 the "build-* directories that it finds.
3679 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3680 address in a domain literal was a prefix of an interface address.
3682 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3683 when verifying a sender address, unless rewrite_headers is false.
3685 PH/44 Wrote a long comment about why errors_to addresses are verified as
3686 recipients, not senders.
3688 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3689 the ratelimit ACL was added.
3691 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3693 PH/46 Added extra information about PostgreSQL errors to the error string.
3695 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3696 avoiding the need to install special zones in a real server. This is
3697 backwards compatible; if it can't find the fake resolver, it drops back.
3698 Thus, both old and new test suites can be run.
3700 TF/02 Added util/ratelimit.pl
3702 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3703 clock is set back in time.
3705 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3706 Candler <B.Candler@pobox.com>.
3708 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3709 OK for addresses that are the result of redirections.
3711 PH/48 A number of further additions for the benefit of the new test suite,
3712 including a fake gethostbyname() that interfaces to the fake DNS resolver
3715 TF/06 The fix for widen_domains has also been applied to qualify_single and
3716 search_parents which are the other dnslookup options that can cause
3719 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3722 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3724 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3725 service" if it encountered an MX record with an empty target hostname.
3726 The message is now "an MX or SRV record indicated no SMTP service".
3728 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3729 if the DNS of the sending site is misconfigured. This is quite a
3730 common situation. This change restores the behaviour of treating a
3731 helo verification defer as a failure.
3733 PH/51 If self=fail was set on a router, the bounce message did not include the
3734 actual error message.
3740 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3742 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3743 command itself ended in error, the underlying message about the transport
3744 filter timeout was being overwritten with the pipe command error. Now the
3745 underlying error message should be appended to the second error message.
3747 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3749 PH/02 Reduce the amount of output that "make" produces by default. Full output
3750 can still be requested.
3752 PH/03 The warning log line about a condition test deferring for a "warn" verb
3753 was being output only once per connection, rather than after each
3754 occurrence (because it was using the same function as for successful
3755 "warn" verbs). This seems wrong, so I have changed it.
3757 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3758 it should not have, which might have caused a crash in the right
3759 circumstances, but probably never did.
3761 PH/04 Installed a modified version of Tony Finch's patch to make submission
3762 mode fix the return path as well as the Sender: header line, and to
3763 add a /name= option so that you can make the user's friendly name appear
3766 TF/03 Added the control = fakedefer ACL modifier.
3768 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3769 Mark Lowes for thorough testing.
3771 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3773 TK/03 Merged latest SRS patch from Miles Wilton.
3775 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3776 with the definition in sysexits.h (which is #included earlier).
3777 Fortunately, Exim does not actually use EX_OK. The code used to try to
3778 preserve the sysexits.h value, by assuming that macro definitions were
3779 scanned for macro replacements. I have been disabused of this notion,
3780 so now the code just undefines EX_OK before #including unistd.h.
3782 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3783 in the smtp transport. When a block could not be written in a single
3784 write() function, the timeout was being re-applied to each part-write.
3785 This seems wrong - if the receiver was accepting one byte at a time it
3786 would take for ever. The timeout is now adjusted when this happens. It
3787 doesn't have to be particularly precise.
3789 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3790 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3792 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3793 without a domain if the check to <postmaster@domain> fails.
3795 SC/01 Eximstats: added -xls and the ability to specify output files
3796 (patch written by Frank Heydlauf).
3798 SC/02 Eximstats: use FileHandles for outputting results.
3800 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3802 SC/04 Eximstats: fixed display of large numbers with -nvr option
3804 SC/05 Eximstats: fixed merging of reports with empty tables.
3806 SC/06 Eximstats: added the -include_original_destination flag
3808 SC/07 Eximstats: removed tabs and trailing whitespace.
3810 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3812 TK/06 MBOX spool code: Add real "From " MBOX separator line
3813 so the .eml file is really in mbox format (even though
3814 most programs do not really care). Patch from Alex Miller.
3816 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3817 The latter is generated from $received_to and is only set if the
3818 message has one envelope recipient. SA can use these headers,
3819 obviously out-of-the-box. Patch from Alex Miller.
3821 PH/08 The ${def test on a variable was returning false if the variable's
3822 value was "0", contrary to what the specification has always said!
3823 The result should be true unless the variable is empty.
3825 PH/09 The syntax error of a character other than { following "${if
3826 def:variable_name" (after optional whitespace) was not being diagnosed.
3827 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3828 accidental colon was present, for example, could give incorrect results.
3830 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3831 result is used (not including appendfile, where other changes are about
3834 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3835 This involved changing a lot of size variables from int to off_t. It
3836 should work with maildirs and everything.
3838 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3839 spamd dying while we are connected to it.
3841 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3844 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3845 function for BDB 4.3.
3847 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3849 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3850 the sources for PCRE have been re-arranged and more files are now
3853 PH/15 The code I had for printing potentially long long variables in PH/11
3854 above was not the best (it lost precision). The length of off_t variables
3855 is now inspected at build time, and an appropriate printing format (%ld
3856 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3857 to be "long long int" or "long int". This is needed for the internal
3858 formatting function string_vformat().
3860 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3861 the configuration file to be ":syslog", then the script "guesses" where
3862 the logs files are, rather than using the compiled in default. In our
3863 case the guess is not the same as the compiled default, so the script
3864 suddenly stopped working when I started to use syslog. The patch checks
3865 to see if log_file_path is "". If so, it attempts to read it from exim
3866 with no configuration file to get the compiled in version, before it
3867 falls back to the previous guessing code."
3869 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3870 implementing BATV in an Exim configuration. See NewStuff for the gory
3873 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3874 Makefile that are specific to HP-UX.
3876 PH/18 If the "use_postmaster" option was set for a recipient callout together
3877 with the "random" option, the postmaster address was used as the MAIL
3878 FROM address for the random test, but not for the subsequent recipient
3879 test. It is now used for both.
3881 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3882 patch removes a few documentation additions to RFC 3028, because the
3883 latest draft now contains them. It adds the new en;ascii-case comparator
3884 and a new error check for 8bit text in MIME parts. Comparator and
3885 require names are now matched exactly. I enabled the subaddress
3886 extension, but it is not well tested yet (read: it works for me)."
3888 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3889 rework some of the code of TK/09 above to avoid the hardwired use of
3890 "%lld" and "long long". Replaced the call to snprintf() with a call to
3893 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3894 records point to non-existent hosts", "retry timeout exceeded", and
3895 "retry time not reached for any host after a long failure period".
3897 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3898 experimental DomainKeys support:
3900 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3901 (2) On an error such as an illegally used "control", the wrong name for
3902 the control was given.
3904 These problems did NOT occur unless DomainKeys support was compiled.
3906 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3908 PH/24 Added ${if match_ip condition.
3910 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3911 databases so that it will be absolutely obvious if a crash occurs in the
3912 DB library. This is a regular occurrence (often caused by mis-matched
3915 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3916 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3917 picked up on a user's system that detects such things. There doesn't seem
3918 to be a gcc warning option for this - only an attribute that has to be
3919 put on the function's prototype. It seems that in Fedora Core 4 they have
3920 set this on a number of new functions. No doubt there will be more in due
3923 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3924 specify a transport. However, if an address that was verified by such a
3925 router was the subject of a callout, Exim crashed because it tried to
3926 read the rcpt_include_affixes from the non-existent transport. Now it
3927 just assumes that the setting of that option is false. This bug was
3928 introduced by 4.51/PH/31.
3930 PH/28 Changed -d+all to exclude +memory, because that information is very
3931 rarely of interest, but it makes the output a lot bigger. People tend to
3932 do -d+all out of habit.
3934 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3935 code in os-type was giving problems when libc.so lives in lib64, like on
3938 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3939 aren't the modern standard, and it seems that some systems' include files
3940 don't always have them. Exim was already checking for some of the newer
3941 ones like T_AAAA, and defining it itself. I've added checks for all the
3942 record types that Exim uses.
3944 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3945 not automatically generating a new one, as it is supposed to. This
3946 prevented TLS from working. If the file did exist, but contained invalid
3947 data, a new version was generated, as expected. It was only the case of a
3948 non-existent file that was broken.
3950 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3951 with a change in libdomainkeys > 0.64.
3953 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3954 from DNS. If the selector record carries the flag, it now has
3955 precedence over the domain-wide flag.
3957 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3959 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3960 the use of a port name that isn't defined in /etc/services) occurred, the
3961 message was deferred as in a normal delivery, and thus remained on the
3962 spool, instead of being failed because of the mua_wrapper setting. This
3963 is now fixed, and I tidied up some of the mua_wrapper messages at the
3966 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3967 the messages in a hash of arrays rather than using individual hashes.
3968 This is a bit cleaner and results in dramatic memory savings, albeit
3969 at a slight CPU cost.
3971 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3972 as requested by Marc Sherman.
3974 SC/10 Eximstats: added histograms for user specified patterns as requested
3977 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3979 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3980 fopen() in the content-scanning modules that did not already have it.
3986 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3987 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3989 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3991 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3993 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3994 to test Sieve filters that use "vacation".
3996 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3997 that changes the way the GnuTLS parameters are stored in the cache file.
3998 The new format can be generated externally. For backward compatibility,
3999 if the data in the cache doesn't make sense, Exim assumes it has read an
4000 old-format file, and it generates new data and writes a new file. This
4001 means that you can't go back to an older release without removing the
4004 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4005 work if there are any delivery delays because "one_time" forces the
4006 parent to be marked "delivered", so its unseen clone is never tried
4007 again. For this reason, Exim now forbids the simultaneous setting of
4010 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4011 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4012 entry if you want to know the details. The fix, however, neglected to
4013 consider the case when local delivery batching is involved. The test for
4014 "previously delivered" was not happening when checking to see if an
4015 address could be batched with a previous (undelivered) one; under
4016 certain circumstances this could lead to multiple deliveries to the same
4019 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4020 in its include files, and this causes problems building Exim.
4022 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4023 header_syntax) but Exim was just ignoring anything given after a slash.
4024 In particular, this caused confusion with an attempt to use "verify =
4025 reverse_host_lookup/defer_ok". An error is now given when options are
4026 supplied for verify items that do not have them. (Maybe reverse_host_
4027 lookup should have a defer_ok option, but that's a different point.)
4029 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4030 defined by RFC 821) to 2048, because there were problems with some AUTH
4031 commands, and RFC 1869 says the size should be increased for extended
4032 SMTP commands that take arguments.
4034 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4037 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4038 "unknown" error; now it says that the functionality isn't in the binary.
4040 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4041 an address' error message when a string expansion fails (syntax or
4042 whatever). Otherwise the password may appear in the log. Following change
4043 PH/42 below, there is no longer a chance of it appearing in a bounce
4046 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4048 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4049 of its addresses was ever tried. (Bugzilla bug #2.)
4051 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4052 the result incorrectly in the debug output. (It correctly added a newline
4053 to what was transported.)
4055 TF/01 Added $received_time.
4057 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4058 commented out examples of how to interface to a virus scanner and to
4059 SpamAssassin. Also added commented examples of av_scanner and
4060 spamd_address settings.
4062 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4063 and controls are allowed in which ACLs. There were a couple of minor
4064 errors. Some of the entries in the conditions table (which is a table of
4065 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4066 negation of where the condition IS allowed.
4068 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4070 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4071 header file does not have a version number, so I've had to invent a new
4072 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4073 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4074 radiusclient), but it was contributed by a Radius user.
4076 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4077 files or directories, for OpenSSL.
4079 PH/19 When an Exim process that is running as root has to create an Exim log
4080 file, it does so in a subprocess that runs as exim:exim so as to get the
4081 ownership right at creation (otherwise, other Exim processes might see
4082 the file with the wrong ownership). There was no test for failure of this
4083 fork() call, which would lead to the process getting stuck as it waited
4084 for a non-existent subprocess. Forks do occasionally fail when resources
4085 run out. I reviewed all the other calls to fork(); they all seem to check
4088 PH/20 When checking for unexpected SMTP input at connect time (before writing
4089 the banner), Exim was not dealing correctly with a non-positive return
4090 from the read() function. If the client had disconnected by this time,
4091 the result was a log entry for a synchronization error with an empty
4092 string after "input=" when read() returned zero. If read() returned -1
4093 (an event I could not check), uninitialized data bytes were printed.
4094 There were reports of junk text (parts of files, etc) appearing after
4097 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4099 PH/22 Added support for macro redefinition, and (re)definition in between
4100 driver and ACL definitions.
4102 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4103 forgetting to use the resulting value; it was using the unexpanded value.
4105 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4106 hadn't been configured. The fix is from Juergen Kreileder, who
4107 understands it better than I do:
4109 "Here's what I see happening with three configured cyrus_sasl
4110 authenticators configured (plain, login, cram-md5):
4112 On startup auth_cyrus_sasl_init() gets called for each of these.
4113 This means three calls to sasl_listmech() without a specified mech_list.
4114 => SASL tests which mechs of all available mechs actually work
4115 => three warnings about OTP not working
4116 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4118 With the patch, sasl_listmech() also gets called three times. But now
4119 SASL's mech_list option is set to the server_mech specified in the the
4120 authenticator. Or in other words, the answer from sasl_listmech()
4121 gets limited to just the mech you're testing for (which is different
4123 => the return list contains just 'plain' or 'login', 'cram-md5' or
4124 nothing depending on the value of ob->server_mech.
4126 I've just tested the patch: Authentication still works fine,
4127 unavailable mechs specified in the exim configuration are still
4128 caught, and the auth.log warnings about OTP are gone."
4130 PH/25 When debugging is enabled, the contents of the command line are added
4131 to the debugging output, even when log_selector=+arguments is not
4134 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4135 answer is "GNU", and only if the return is "GNU/something" is the answer
4138 PH/27 $acl_verify_message is now set immediately after the failure of a
4139 verification in an ACL, and so is available in subsequent modifiers. In
4140 particular, the message can be preserved by coding like this:
4142 warn !verify = sender
4143 set acl_m0 = $acl_verify_message
4145 Previously, $acl_verify_message was set only while expanding "message"
4146 and "log_message" when a very denied access.
4148 PH/28 Modified OS/os.c-Linux with
4150 -#ifndef OS_LOAD_AVERAGE
4151 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4153 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4154 nomenclature these days.)
4156 PH/29 Installed patch from the Sieve maintainer that adds the options
4157 sieve_useraddress and sieve_subaddress to the redirect router.
4159 PH/30 In these circumstances:
4160 . Two addresses routed to the same list of hosts;
4161 . First host does not offer TLS;
4162 . First host accepts first address;
4163 . First host gives temporary error to second address;
4164 . Second host offers TLS and a TLS session is established;
4165 . Second host accepts second address.
4166 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4167 and peerdn, if requested) that were in fact used only for the second
4170 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4171 attention to any local part prefix or suffix that was matched by the
4172 router that accepted the address. It now behaves in the same way as it
4173 does for delivery: the affixes are removed from the local part unless
4174 rcpt_include_affixes is set on the transport.
4176 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4177 timeout during the DATA phase of an incoming message.
4179 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4180 applied a patch sanctioned by the Sieve maintainer.
4182 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4183 the uid or gid is negative. A case of a negative gid caused this to be
4184 noticed. The fix allows for either to be negative.
4186 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4187 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4190 PH/36 The change PH/12 above was broken. Fixed it.
4192 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4193 the grounds that routing the same address twice would always produce the
4194 same answer. This might have been true once, but it is certainly no
4195 longer true now. Routing a child address may depend on the previous
4196 routing that produced that child. Some complicated redirection strategies
4197 went wrong when messages had multiple recipients, and made Exim's
4198 behaviour dependent on the order in which the addresses were given.
4200 I have moved the duplicate checking until after the routing is complete.
4201 Exim scans the addresses that are assigned to local and remote
4202 transports, and removes any duplicates. This means that more work will be
4203 done, as duplicates will always all be routed, but duplicates are
4204 presumably rare, so I don't expect this is of any significance.
4206 For deliveries to pipes, files, and autoreplies, the duplicate checking
4207 still happens during the routing process, since they are not going to be
4210 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4211 It corrects a timeout issue with spamd. This is Ian's comment: "The
4212 background is that sometimes spamd either never reads data from a
4213 connection it has accepted, or it never writes response data. The exiscan
4214 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4215 blindly assumes that writes won't block so it may never time out."
4217 PH/39 Allow G after quota size as well as K and M.
4219 PH/40 The value set for $authenticated_id in an authenticator may not contain
4220 binary zeroes or newlines because the value is written to log lines and
4221 to spool files. There was no check on this. Now the value is run through
4222 the string_printing() function so that such characters are converted to
4223 printable escape sequences.
4225 PH/41 $message_linecount is a new variable that contains the total number of
4226 lines in the message. Compare $body_linecount, which is the count for the
4229 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4230 bounce and delay warning messages, except in certain special cases, which
4233 (a) An SMTP error message from a remote host;
4234 (b) A message specified in a :fail: redirection;
4235 (c) A message specified in a "fail" command in a system filter;
4236 (d) A message specified in a FAIL return from the queryprogram router;
4237 (e) A message specified by the cannot_route_message router option.
4239 In these cases only, Exim does include the error details in bounce and
4240 warning messages. There are also a few cases where bland messages such
4241 as "unrouteable address" or "local delivery error" are given.
4243 PH/43 $value is now also set for the "else" part of a ${run expansion.
4245 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4246 being worked on, but at least Exim now implements the latest version to
4249 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4250 process to complete was treated as a delivery failure, a timeout while
4251 writing the message to the pipe was logged, but erroneously treated as a
4252 successful delivery. Such timeouts include transport filter timeouts. For
4253 consistency with the overall process timeout, these timeouts are now
4254 treated as errors, giving rise to delivery failures by default. However,
4255 there is now a new Boolean option for the pipe transport called
4256 timeout_defer, which, if set TRUE, converts the failures into defers for
4257 both kinds of timeout. A transport filter timeout is now identified in
4260 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4261 systems where "make" and "gmake" are different, calling "gmake" at top
4262 level broke things. I've arranged for the value of $(MAKE) to be passed
4263 from the Makefile to this script so that it can call the same version of
4267 A note about Exim versions 4.44 and 4.50
4268 ----------------------------------------
4270 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4271 changes of various kinds. As a consequence, a big documentation update was
4272 needed. This delayed the release for rather longer than seemed good, especially
4273 in the light of a couple of (minor) security issues. Therefore, the changes
4274 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4275 release. So 4.44 and 4.50 are in effect two different branches that both start
4278 I have left the 4.50 change log unchanged; it contains all the changes since
4279 4.43. The change log for 4.44 is below; many of its items are identical to
4280 those for 4.50. This seems to be the most sensible way to preserve the
4281 historical information.
4287 1. Minor wording change to the doc/README.SIEVE file.
4289 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4290 computation of the current number of files was incorrect.
4292 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4293 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4296 4. Give more explanation in the error message when the command for a transport
4297 filter fails to execute.
4299 5. There are several places where Exim runs a non-Exim command in a
4300 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4301 was being done only for the command run by the queryprogram router. It is
4302 now done for all such subprocesses. The other cases are: ${run, transport
4303 filters, and the commands run by the lmtp and pipe transports.
4305 6. Added CONFIGURE_GROUP build-time option.
4307 7. Some older OS have a limit of 256 on the maximum number of file
4308 descriptors. Exim was using setrlimit() to set 1000 as a large value
4309 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4310 systems. I've change it so that if it can't get 1000, it tries for 256.
4312 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4313 was an oversight, and furthermore, ever since the addition of extra
4314 controls (e.g. 4.43/32), the checks on when to allow different forms of
4315 "control" were broken. There should now be diagnostics for all cases when a
4316 control that does not make sense is encountered.
4318 9. Added the /retain_sender option to "control=submission".
4320 10. $recipients is now available in the predata ACL (oversight).
4322 11. Tidy the search cache before the fork to do a delivery from a message
4323 received from the command line. Otherwise the child will trigger a lookup
4324 failure and thereby defer the delivery if it tries to use (for example) a
4325 cached ldap connection that the parent has called unbind on.
4327 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4328 of $address_data from the recipient verification was clobbered by the
4329 sender verification.
4331 13. The value of address_data from a sender verification is now available in
4332 $sender_address_data in subsequent conditions in the ACL statement.
4334 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4336 15. Added a new option "connect=<time>" to callout options, to set a different
4339 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4340 was its contents. (It was OK if the option was not defined at all.)
4342 17. A "Completed" log line is now written for messages that are removed from
4343 the spool by the -Mrm option.
4345 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4346 information about exactly what failed.
4348 19. Added -dd to debug only the daemon process.
4350 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4351 handles timeouts, both on the server side and network timeouts. Renamed the
4352 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4354 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4355 It is now set to "smtps".
4357 22. $host_address is now set to the target address during the checking of
4358 ignore_target_hosts.
4360 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4361 being passed; this would have caused $sender_host_name to have been used if
4362 matching the list had actually called for a host name (not very likely,
4363 since this list is usually IP addresses). A host name is now passed as
4366 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4367 code that specifies a non-restarting handler (typically sigaction() in
4368 modern systems) in an attempt to fix a rare and obscure crash bug.
4370 25. Narrowed the window for a race in the daemon that could cause it to ignore
4371 SIGCHLD signals. This is not a major problem, because they are used only to
4372 wake it up if nothing else does.
4374 26. A malformed maildirsize file could cause Exim to calculate negative values
4375 for the mailbox size or file count. Odd effects could occur as a result.
4376 The maildirsize information is now recalculated if the size or filecount
4379 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4380 support for a long time. Removed HAVE_SYS_VFS_H.
4382 28. Installed the latest version of exipick from John Jetmore.
4384 29. In an address list, if the pattern was not a regular expression, an empty
4385 subject address (from a bounce message) matched only if the pattern was an
4386 empty string. Non-empty patterns were not even tested. This was the wrong
4387 because it is perfectly reasonable to use an empty address as part of a
4388 database query. An empty address is now tested by patterns that are
4389 lookups. However, all the other forms of pattern expect the subject to
4390 contain a local part and a domain, and therefore, for them, an empty
4391 address still always fails if the pattern is not itself empty.
4393 30. Exim went into a mad DNS loop when attempting to do a callout where the
4394 host was specified on an smtp transport, and looking it up yielded more
4395 than one IP address.
4397 31. Re-factored the code for checking spool and log partition space into a
4398 function that finds that data and another that does the check. The former
4399 is then used to implement four new variables: $spool_space, $log_space,
4400 $spool_inodes, and $log_inodes.
4402 32. The RFC2047 encoding function was originally intended for short strings
4403 such as real names; it was not keeping to the 75-character limit for
4404 encoded words that the RFC imposes. It now respects the limit, and
4405 generates multiple encoded words if necessary. To be on the safe side, I
4406 have increased the buffer size for the ${rfc2047: expansion operator from
4409 33. It is now permitted to omit both strings after an "if" condition; if the
4410 condition is true, the result is "true". As before, when the second string
4411 is omitted, a false condition yields an empty string. This makes it less
4412 cumbersome to write custom ACL and router conditions.
4414 34. Failure to deliver a bounce message always caused it to be frozen, even if
4415 there was an errors_to setting on the router. The errors_to setting is now
4418 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4419 canonical form (fully expanded) before being placed in
4420 $sender_host_address.
4422 36. The table in the code that translates DNS record types into text (T_A to
4423 "A" for instance) was missing entries for NS and CNAME. It is just possible
4424 that this could have caused confusion if both these types were looked up
4425 for the same domain, because the text type is used as part of Exim's
4426 per-process caching. But the chance of anyone hitting this buglet seems
4429 37. The dnsdb lookup has been extended in a number of ways.
4431 (1) There is a new type, "zns", which walks up the domain tree until it
4432 finds some nameserver records. It should be used with care.
4434 (2) There is a new type, "mxh", which is like "mx" except that it returns
4435 just the host names, not the priorities.
4437 (3) It is now possible to give a list of domains (or IP addresses) to be
4438 looked up. The behaviour when one of the lookups defers can be
4439 controlled by a keyword.
4441 (4) It is now possible to specify the separator character for use when
4442 multiple records are returned.
4444 38. The dnslists ACL condition has been extended: it is now possible to supply
4445 a list of IP addresses and/or domains to be looked up in a particular DNS
4448 39. Added log_selector=+queue_time_overall.
4450 40. When running the queue in the test harness, wait just a tad after forking a
4451 delivery process, to get repeatability of debugging output.
4453 41. Include certificate and key file names in error message when GnuTLS fails
4454 to set them up, because the GnuTLS error message doesn't include the name
4455 of the failing file when there is a problem reading it.
4457 42. Allow both -bf and -bF in the same test run.
4459 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4461 44. The "Exiscan patch" is now merged into the mainline Exim source.
4463 45. Sometimes the final signoff response after QUIT could fail to get
4464 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4465 < 0 before doing a fflush(). This bug looks as though it goes back to the
4466 introduction of TLS in release 3.20, but "sometimes" must have been rare
4467 because the tests only now provoked it.
4469 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4470 (this can affect the format of dates).
4472 47. exim_tidydb, when checking for the continued existence of a message for
4473 which it has found a message-specific retry record, was not finding
4474 messages that were in split spool directories. Consequently, it was
4475 deleting retry records that should have stayed in existence.
4477 48. Steve fixed some bugs in eximstats.
4479 49. The SPA authentication driver was not abandoning authentication and moving
4480 on to the next authenticator when an expansion was forced to fail,
4481 contradicting the general specification for all authenticators. Instead it
4482 was generating a temporary error. It now behaves as specified.
4484 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4485 (the order specifies the preference for clients). The order is now AES256,
4486 AES128, 3DES, ARCFOUR128.
4488 51. Small patch to Sieve code - explicitly set From: when generating an
4491 52. Exim crashed if a remote delivery caused a very long error message to be
4492 recorded - for instance if somebody sent an entire SpamAssassin report back
4493 as a large number of 550 error lines. This bug was coincidentally fixed by
4494 increasing the size of one of Exim's internal buffers (big_buffer) that
4495 happened as part of the Exiscan merge. However, to be on the safe side, I
4496 have made the code more robust (and fixed the comments that describe what
4499 53. Now that there can be additional text after "Completed" in log lines (if
4500 the queue_time_overall log selector is set), a one-byte patch to exigrep
4501 was needed to allow it to recognize "Completed" as not the last thing in
4504 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4505 patch that reportedly fixes this has been added. I am not expert enough to
4506 create a test for it. This is what the patch creator wrote:
4508 "I found a little strange behaviour of ldap code when working with
4509 Windows 2003 AD Domain, where users was placed in more than one
4510 Organization Units. When I tried to give exim partial DN, the exit code
4511 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4512 But simultaneously result of request was absolutely normal ldap result,
4513 so I produce this patch..."
4515 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4516 REFERENCE, so I have modified the code to exclude the patch when that macro
4519 55. Some experimental protocols are using DNS PTR records for new purposes. The
4520 keys for these records are domain names, not reversed IP addresses. The
4521 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4522 leaves it alone. Component reversal etc. now happens only for IP addresses.
4525 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4527 57. Double the size of the debug message buffer (to 2048) so that more of very
4528 long debug lines gets shown.
4530 58. The exicyclog utility now does better if the number of log files to keep
4531 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4533 59. Two changes related to the smtp_active_hostname option:
4535 (1) $smtp_active_hostname is now available as a variable.
4536 (2) The default for smtp_banner uses $smtp_active_hostname instead
4537 of $primary_hostname.
4539 60. The host_aton() function is supposed to be passed a string that is known
4540 to be a valid IP address. However, in the case of IPv6 addresses, it was
4541 not checking this. This is a hostage to fortune. Exim now panics and dies
4542 if the condition is not met. A case was found where this could be provoked
4543 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4544 components; fortuitously, this particular loophole had already been fixed
4545 by change 4.50/55 above.
4547 If there are any other similar loopholes, the new check in host_aton()
4548 itself should stop them being exploited. The report I received stated that
4549 data on the command line could provoke the exploit when Exim was running as
4550 exim, but did not say which command line option was involved. All I could
4551 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4552 running as the user.
4555 61. There was a buffer overflow vulnerability in the SPA authentication code
4556 (which came originally from the Samba project). I have added a test to the
4557 spa_base64_to_bits() function which I hope fixes it.
4560 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4561 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4563 63. The daemon start-up calls getloadavg() while still root for those OS that
4564 need the first call to be done as root, but it missed one case: when
4565 deliver_queue_load_max is set with deliver_drop_privilege. This is
4566 necessary for the benefit of the queue runner, because there is no re-exec
4567 when deliver_drop_privilege is set.
4569 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4570 This has been fixed.
4572 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4573 was in use, was not putting the data itself into the right store pool;
4574 consequently, it could be overwritten for a subsequent message in the same
4575 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4578 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4580 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4581 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4582 the function were treating the return as a boolean value, which happened to
4583 work because 0=false and not-0=true, but is not correct code.
4585 68. The host_aton() function was not handling scoped IPv6 addresses (those
4586 with, for example, "%eth0" on the end) correctly.
4588 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4589 negated items (that is, ~something) in unsigned ints. Some compilers
4590 apparently mutter when there is no cast.
4592 70. If an address verification called from an ACL failed, and did not produce a
4593 user-specific message (i.e. there was only a "system" message), nothing was
4594 put in $acl_verify_message. In this situation, it now puts the system
4597 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4598 session; change 4.31/43 added the unwanted input to the log line - except
4599 that it did not do this in the start of session case. It now does.
4601 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4602 This is wrong and can cause the other end to generate a synchronization
4603 error if it is another Exim or anything else that does the synchronization
4604 check. A QUIT command is no longer sent after a timeout.
4606 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4607 during host lookups.
4609 74. The defer_ok option of callout verification was not working if it was used
4610 when verifying addresses in header lines, that is, for this case:
4612 verify = header_sender/callout=defer_ok
4614 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4615 those file descriptors could be used for SMTP connections. If anything
4616 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4617 could be sent to the SMTP client, causing chaos. The daemon now opens
4618 stdin, stdout, and stderr to /dev/null when it puts itself into the
4621 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4622 log by default. The user can override this with suitable Perl magic.
4624 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4625 the log message when discard triggers, was not working for the DATA ACL or
4626 for the non-SMTP ACL.
4628 78. Error message wording change in sieve.c.
4630 79. If smtp_accept_max_per_host was set, the number of connections could be
4631 restricted to fewer than expected, because the daemon was trying to set up
4632 a new connection before checking whether the processes handling previous
4633 connections had finished. The check for completed processes is now done
4634 earlier. On busy systems, this bug wouldn't be noticed because something
4635 else would have woken the daemon, and it would have reaped the completed
4638 80. If a message was submitted locally by a user whose login name contained one
4639 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4640 It caused a spool format error. I have fixed the spool reading code. A
4641 related problem was that the "from" clause in the Received: line became
4642 illegal because of the space(s). It is now covered by ${quote_local_part.
4644 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4647 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4649 83. Patch from Sieve maintainer to fix unterminated string problem in
4650 "vacation" handling.
4652 84. Some minor changes to the Linux configuration files to help with other
4653 OS variants using glibc.
4655 85. One more patch for Sieve to update vacation handling to latest spec.
4658 ----------------------------------------------------
4659 See the note above about the 4.44 and 4.50 releases.
4660 ----------------------------------------------------
4666 1. Change 4.43/35 introduced a bug that caused file counts to be
4667 incorrectly computed when quota_filecount was set in an appendfile
4670 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4671 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4674 3. Give more explanation in the error message when the command for a transport
4675 filter fails to execute.
4677 4. There are several places where Exim runs a non-Exim command in a
4678 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4679 was being done only for the command run by the queryprogram router. It is
4680 now done for all such subprocesses. The other cases are: ${run, transport
4681 filters, and the commands run by the lmtp and pipe transports.
4683 5. Some older OS have a limit of 256 on the maximum number of file
4684 descriptors. Exim was using setrlimit() to set 1000 as a large value
4685 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4686 systems. I've change it so that if it can't get 1000, it tries for 256.
4688 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4689 was an oversight, and furthermore, ever since the addition of extra
4690 controls (e.g. 4.43/32), the checks on when to allow different forms of
4691 "control" were broken. There should now be diagnostics for all cases when a
4692 control that does not make sense is encountered.
4694 7. $recipients is now available in the predata ACL (oversight).
4696 8. Tidy the search cache before the fork to do a delivery from a message
4697 received from the command line. Otherwise the child will trigger a lookup
4698 failure and thereby defer the delivery if it tries to use (for example) a
4699 cached ldap connection that the parent has called unbind on.
4701 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4702 of $address_data from the recipient verification was clobbered by the
4703 sender verification.
4705 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4706 was its contents. (It was OK if the option was not defined at all.)
4708 11. A "Completed" log line is now written for messages that are removed from
4709 the spool by the -Mrm option.
4711 12. $host_address is now set to the target address during the checking of
4712 ignore_target_hosts.
4714 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4715 being passed; this would have caused $sender_host_name to have been used if
4716 matching the list had actually called for a host name (not very likely,
4717 since this list is usually IP addresses). A host name is now passed as
4720 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4721 code that specifies a non-restarting handler (typically sigaction() in
4722 modern systems) in an attempt to fix a rare and obscure crash bug.
4724 15. Narrowed the window for a race in the daemon that could cause it to ignore
4725 SIGCHLD signals. This is not a major problem, because they are used only to
4726 wake it up if nothing else does.
4728 16. A malformed maildirsize file could cause Exim to calculate negative values
4729 for the mailbox size or file count. Odd effects could occur as a result.
4730 The maildirsize information is now recalculated if the size or filecount
4733 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4734 support for a long time. Removed HAVE_SYS_VFS_H.
4736 18. Updated exipick to current release from John Jetmore.
4738 19. Allow an empty sender to be matched against a lookup in an address list.
4739 Previously the only cases considered were a regular expression, or an
4742 20. Exim went into a mad DNS lookup loop when doing a callout where the
4743 host was specified on the transport, if the DNS lookup yielded more than
4746 21. The RFC2047 encoding function was originally intended for short strings
4747 such as real names; it was not keeping to the 75-character limit for
4748 encoded words that the RFC imposes. It now respects the limit, and
4749 generates multiple encoded words if necessary. To be on the safe side, I
4750 have increased the buffer size for the ${rfc2047: expansion operator from
4753 22. Failure to deliver a bounce message always caused it to be frozen, even if
4754 there was an errors_to setting on the router. The errors_to setting is now
4757 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4758 canonical form (fully expanded) before being placed in
4759 $sender_host_address.
4761 24. Updated eximstats to version 1.33
4763 25. Include certificate and key file names in error message when GnuTLS fails
4764 to set them up, because the GnuTLS error message doesn't include the name
4765 of the failing file when there is a problem reading it.
4767 26. Expand error message when OpenSSL has problems setting up cert/key files.
4770 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4771 (this can affect the format of dates).
4773 28. exim_tidydb, when checking for the continued existence of a message for
4774 which it has found a message-specific retry record, was not finding
4775 messages that were in split spool directories. Consequently, it was
4776 deleting retry records that should have stayed in existence.
4778 29. eximstats updated to version 1.35
4779 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4780 1.35 - bugfix such that pie charts by volume are generated correctly
4782 30. The SPA authentication driver was not abandoning authentication and moving
4783 on to the next authenticator when an expansion was forced to fail,
4784 contradicting the general specification for all authenticators. Instead it
4785 was generating a temporary error. It now behaves as specified.
4787 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4788 (the order specifies the preference for clients). The order is now AES256,
4789 AES128, 3DES, ARCFOUR128.
4791 31. Small patch to Sieve code - explicitly set From: when generating an
4794 32. Exim crashed if a remote delivery caused a very long error message to be
4795 recorded - for instance if somebody sent an entire SpamAssassin report back
4796 as a large number of 550 error lines. This bug was coincidentally fixed by
4797 increasing the size of one of Exim's internal buffers (big_buffer) that
4798 happened as part of the Exiscan merge. However, to be on the safe side, I
4799 have made the code more robust (and fixed the comments that describe what
4802 33. Some experimental protocols are using DNS PTR records for new purposes. The
4803 keys for these records are domain names, not reversed IP addresses. The
4804 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4805 leaves it alone. Component reversal etc. now happens only for IP addresses.
4808 34. The host_aton() function is supposed to be passed a string that is known
4809 to be a valid IP address. However, in the case of IPv6 addresses, it was
4810 not checking this. This is a hostage to fortune. Exim now panics and dies
4811 if the condition is not met. A case was found where this could be provoked
4812 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4813 components; fortuitously, this particular loophole had already been fixed
4814 by change 4.50/55 or 4.44/33 above.
4816 If there are any other similar loopholes, the new check in host_aton()
4817 itself should stop them being exploited. The report I received stated that
4818 data on the command line could provoke the exploit when Exim was running as
4819 exim, but did not say which command line option was involved. All I could
4820 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4821 running as the user.
4824 35. There was a buffer overflow vulnerability in the SPA authentication code
4825 (which came originally from the Samba project). I have added a test to the
4826 spa_base64_to_bits() function which I hope fixes it.
4829 36. The daemon start-up calls getloadavg() while still root for those OS that
4830 need the first call to be done as root, but it missed one case: when
4831 deliver_queue_load_max is set with deliver_drop_privilege. This is
4832 necessary for the benefit of the queue runner, because there is no re-exec
4833 when deliver_drop_privilege is set.
4835 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4836 was in use, was not putting the data itself into the right store pool;
4837 consequently, it could be overwritten for a subsequent message in the same
4838 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4841 38. Sometimes the final signoff response after QUIT could fail to get
4842 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4843 < 0 before doing a fflush(). This bug looks as though it goes back to the
4844 introduction of TLS in release 3.20, but "sometimes" must have been rare
4845 because the tests only now provoked it.
4851 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4852 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4853 Now it returns an integer. A number of calls were still expecting a T/F
4854 return. Fortuitously, in all cases, the tests worked in OK situations,
4855 which is the norm. However, things would have gone wrong on any write
4856 failures on the smtp file descriptor. This function is used when sending
4857 messages over SMTP and also when doing verify callouts.
4859 2. When Exim is called to do synchronous delivery of a locally submitted
4860 message (the -odf or -odi options), it no longer closes stderr before doing
4863 3. Implemented the mua_wrapper option.
4865 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4867 5. Implemented the functions header_remove(), header_testname(),
4868 header_add_at_position(), and receive_remove_recipient(), and exported them
4871 6. If an ACL "warn" statement specified the addition of headers, Exim already
4872 inserted X-ACL-Warn: at the start if there was no header name. However, it
4873 was not making this test for the second and subsequent header lines if
4874 there were newlines in the string. This meant that an invalid header could
4875 be inserted if Exim was badly configured.
4877 7. Allow an ACL "warn" statement to add header lines at the start or after all
4878 the Received: headers, as well as at the end.
4880 8. Added the rcpt_4xx retry error code.
4882 9. Added postmaster_mailfrom=xxx to callout verification option.
4884 10. Added mailfrom=xxxx to the callout verification option, for verify=
4887 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4888 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4890 12. Inserted some casts to stop certain compilers warning when using pointer
4891 differences as field lengths or precisions in printf-type calls (mostly
4892 affecting debugging statements).
4894 13. Added optional readline() support for -be (dynamically loaded).
4896 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4897 same clock tick as a message's arrival, so that its received time was the
4898 same as the "first fail" time on the retry record, and that message
4899 remained on the queue past the ultimate address timeout, every queue runner
4900 would try a delivery (because it was past the ultimate address timeout) but
4901 after another failure, the ultimate address timeout, which should have then
4902 bounced the address, did not kick in. This was a "< instead of <=" error;
4903 in most cases the first failure would have been in the next clock tick
4904 after the received time, and all would be well.
4906 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4907 being recognized when the domain list was tested by the match_domain
4908 condition in an expansion string.
4910 16. Added the ${str2b64: operator.
4912 17. Exim was always calling setrlimit() to set a large limit for the number of
4913 processes, without checking whether the existing limit was already
4914 adequate. (It did check for the limit on file descriptors.) Furthermore,
4915 errors from getrlimit() and setrlimit() were being ignored. Now they are
4916 logged to the main and panic logs, but Exim does carry on, to try to do its
4917 job under whatever limits there are.
4919 18. Imported PCRE 5.0.
4921 19. Trivial typo in log message " temporarily refused connection" (the leading
4924 20. If the log selector return_path_on_delivery was set and an address was
4925 redirected to /dev/null, the delivery process crashed because it assumed
4926 that a return path would always be set for a "successful" delivery. In this
4927 case, the whole delivery is bypassed as an optimization, and therefore no
4930 21. Internal re-arrangement: the function for sending a challenge and reading
4931 a response while authentication was assuming a zero-terminated challenge
4932 string. It's now changed to take a pointer and a length, to allow for
4933 binary data in such strings.
4935 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4937 23. Exim was not respecting finduser_retries when seeking the login of the
4938 uid under which it was called; it was always trying 10 times. (The default
4939 setting of finduser_retries is zero.) Also, it was sleeping after the final
4940 failure, which is pointless.
4942 24. Implemented tls_on_connect_ports.
4944 25. Implemented acl_smtp_predata.
4946 26. If the domain in control=submission is set empty, Exim assumes that the
4947 authenticated id is a complete email address when it generates From: or
4948 Sender: header lines.
4950 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4951 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4952 chown and chgrp in /bin and hostname in /usr/bin.
4954 28. Exim was keeping the "process log" file open after each use, just as it
4955 does for the main log. This opens the possibility of it remaining open for
4956 long periods when the USR1 signal hits a daemon. Occasional processlog
4957 errors were reported, that could have been caused by this. Anyway, it seems
4958 much more sensible not to leave this file open at all, so that is what now
4961 29. The long-running daemon process does not normally write to the log once it
4962 has entered its main loop, and it closes the log before doing so. This is
4963 so that log files can straightforwardly be renamed and moved. However,
4964 there are a couple of unusual error situations where the daemon does write
4965 log entries, and I had neglected to close the log afterwards.
4967 30. The text of an SMTP error response that was received during a remote
4968 delivery was being truncated at 512 bytes. This is too short for some of
4969 the long messages that one sometimes sees. I've increased the limit to
4972 31. It is now possible to make retry rules that apply only when a message has a
4973 specific sender, in particular, an empty sender.
4975 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4976 it possible to be selective about when SMTP synchronization is enforced.
4978 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4980 32. Implemented hosts_connection_nolog.
4982 33. Added an ACL for QUIT.
4984 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4987 35. Added mailbox_size and mailbox_filecount to appendfile.
4989 36. Added control = no_multiline_responses to ACLs.
4991 37. There was a bug in the logic of the code that waits for the clock to tick
4992 in the case where the clock went backwards by a substantial amount such
4993 that the microsecond fraction of "now" was more than the microsecond
4994 fraction of "then" (but the whole seconds number was less).
4996 38. Added support for the libradius Radius client library this is found on
4997 FreeBSD (previously only the radiusclient library was supported).
5003 1. When certain lookups returned multiple values in the form name=value, the
5004 quoting of the values was not always being done properly. Specifically:
5005 (a) If the value started with a double quote, but contained no whitespace,
5007 (b) If the value contained whitespace other than a space character (i.e.
5008 tabs or newlines or carriage returns) it was not quoted.
5009 This fix has been applied to the mysql and pgsql lookups by writing a
5010 separate quoting function and calling it from the lookup code. The fix
5011 should probably also be applied to nisplus, ibase and oracle lookups, but
5012 since I cannot test any of those, I have not disturbed their existing code.
5014 2. A hit in the callout cache for a specific address caused a log line with no
5015 reason for rejecting RCPT. Now it says "Previous (cached) callout
5016 verification failure".
5018 3. There was an off-by-one bug in the queryprogram router. An over-long
5019 return line was truncated at 256 instead of 255 characters, thereby
5020 overflowing its buffer with the terminating zero. As well as fixing this, I
5021 have increased the buffer size to 1024 (and made a note to document this).
5023 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5024 when Exim is waiting for an SMTP response from a remote server, Exim
5025 restarts its select() call on the socket, thereby resetting its timeout.
5026 This is not a problem when such interrupts are rare. Somebody set up a cron
5027 job to run exiwhat every 2 minutes, which is less than the normal select()
5028 timeout (5 or 10 minutes). This meant that the select() timeout never
5029 kicked in because it was always reset. I have fixed this by comparing the
5030 time when an interrupt arrives with the time at the start of the first call
5031 to select(). If more time than the timeout has elapsed, the interrupt is
5032 treated as a timeout.
5034 5. Some internal re-factoring in preparation for the addition of Sieve
5035 extensions (by MH). In particular, the "personal" test is moved to a
5036 separate function, and given an option for scanning Cc: and Bcc: (which is
5037 not set for Exim filters).
5039 6. When Exim created an email address using the login of the caller as the
5040 local part (e.g. when creating a From: or Sender: header line), it was not
5041 quoting the local part when it contained special characters such as @.
5043 7. Installed new OpenBSD configuration files.
5045 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5046 try to make them clearer.
5048 9. Callout options, other than the timeout value, were being ignored when
5049 verifying sender addresses in header lines. For example, when using
5051 verify = header_sender/callout=no_cache
5053 the cache was (incorrectly) being used.
5055 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5056 only the Cygwin environment.
5058 11. When return_path_on_delivery was set as a log selector, if different remote
5059 addresses in the same message used different return paths and parallel
5060 remote delivery occurred, the wrong values would sometimes be logged.
5061 (Whenever a remote delivery process finished, the return path value from
5062 the most recently started remote delivery process was logged.)
5064 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5065 lines when AUTH and/or TLS are in use. This is the "received protocol"
5066 field. Exim used to use "asmtp" for authenticated SMTP, without any
5067 indication (in the protocol name) for TLS use. Now it follows the RFC and
5068 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5069 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5070 names appear in log lines as well as in Received: header lines.
5072 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5073 extensions, and comparison tests, and to fix some bugs.
5075 14. Changes to the "personal" filter test:
5077 (1) The test was buggy in that it was just doing the equivalent of
5078 "contains" tests on header lines. For example, if a user's address was
5079 anne@some.where, the "personal" test would incorrectly be true for
5081 To: susanne@some.where
5083 This test is now done by extracting each address from the header in turn,
5084 and checking the entire address. Other tests that are part of "personal"
5085 are now done using regular expressions (for example, to check local parts
5086 of addresses in From: header lines).
5088 (2) The list of non-personal local parts in From: addresses has been
5089 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5090 taken from the Sieve specification recommendations.
5092 (3) If the message contains any header line starting with "List-" it is
5093 treated as non-personal.
5095 (4) The test for "circular" in the Subject: header line has been removed
5096 because it now seems ill-conceived.
5098 15. Minor typos in src/EDITME comments corrected.
5100 16. Installed latest exipick from John Jetmore.
5102 17. If headers_add on a router specified a text string that was too long for
5103 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5104 of string_sprintf() is now avoided.
5106 18. $message_body_size was not set (it was always zero) when running the DATA
5107 ACL and the local_scan() function.
5109 19. For the "mail" command in an Exim filter, no default was being set for
5110 the once_repeat time, causing a random time value to be used if "once" was
5111 specified. (If the value happened to be <= 0, no repeat happened.) The
5112 default is now 0s, meaning "never repeat". The "vacation" command was OK
5113 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5114 (I found it when inspecting the code).
5116 20. There is now an overall timeout for performing a callout verification. It
5117 defaults to 4 times the callout timeout, which applies to individual SMTP
5118 commands during the callout. The overall timeout applies when there is more
5119 than one host that can be tried. The timeout is checked before trying the
5120 next host. This prevents very long delays if there are a large number of
5121 hosts and all are timing out (e.g. when the network connections are timing
5122 out). The value of the overall timeout can be changed by specifying an
5123 additional sub-option for "callout", called "maxwait". For example:
5125 verify = sender/callout=5s,maxwait=20s
5127 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5128 to the end before writing, but this should make it even safer).
5130 22. Exim was forgetting that it had advertised PIPELINING for the second and
5131 subsequent messages on an SMTP connection. It was also not resetting its
5132 memory on STARTTLS and an internal HELO.
5134 23. When Exim logs an SMTP synchronization error within a session, it now
5135 records whether PIPELINING has been advertised or not.
5137 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5138 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5139 rather than long int.
5141 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5143 26. Added the never_mail option to autoreply.
5149 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5150 crash if the getsockname() call failed; this can happen if a connection is
5151 closed very soon after it is established. The problem was simply in the
5152 order in which certain operations were done, causing Exim to try to write
5153 to the SMTP stream before it had set up the file descriptor. The bug has
5154 been fixed by making things happen in the correct order.
5160 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5161 before the connection was closed, thus losing the rejection response.
5163 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5164 some early Solaris releases, but causes trouble in current releases where
5165 socklen_t is defined.
5167 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5170 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5173 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5174 that releasing the top of it at the end releases what was used for sub-
5175 expansions (unless the block got too big). However, discard this block if
5176 the first thing is a variable or header, so that we can use its block when
5177 it is dynamic (useful for very large $message_headers, for example).
5179 6. Lookups now cache *every* query, not just the most recent. A new, separate
5180 store pool is used for this. It can be recovered when all lookup caches are
5181 flushed. Lookups now release memory at the end of their result strings.
5182 This has involved some general refactoring of the lookup sources.
5184 7. Some code has been added to the store_xxx() functions to reduce the amount
5185 of flapping under certain conditions.
5187 8. log_incoming_interface used to affect only the <= reception log lines. Now
5188 it causes the local interface and port to be added to several more SMTP log
5189 lines, for example "SMTP connection from", and rejection lines.
5191 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5193 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5195 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5196 could be overwritten at the end of the current message (or the start of a
5197 new message if it was set in a HELO ACL). The value is now preserved for
5198 the duration of the SMTP connection.
5200 12. If a transport had a headers_rewrite setting, and a matching header line
5201 contained an unqualified address, that address was qualified, even if it
5202 did not match any rewriting rules. The underlying bug was that the values
5203 of the flags that permit the existence of unqualified sender and recipient
5204 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5205 non-local messages, and by -bnq for local messages) were not being
5206 preserved with the message after it was received.
5208 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5209 "next input=" as part of the text comprising the host identity instead of
5210 the correct text. The code was using the same buffer for two different
5211 strings. However, depending on which order the printing function evaluated
5212 its arguments, the bug did not always show up. Under Linux, for example, my
5213 test suite worked just fine.
5215 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5216 doesn't work with some older versions of Perl. It has been changed to "my",
5217 which in any case is probably the better facility to use.
5219 15. A really picky compiler found some instances of statements for creating
5220 error messages that either had too many or two few arguments for the format
5223 16. The size of the buffer for calls to the DNS resolver has been increased
5224 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5225 for addresses that have a lot of PTR records. This alleviates a problem; it
5226 does not fully solve it.
5228 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5229 buffer now truncates the list and logs the incident, which is the same
5230 action as happens when Exim is looking up a host name and its aliases.
5231 Previously in this situation something unpredictable would happen;
5232 sometimes it was "internal error: store_reset failed".
5234 18. If a server dropped the connection unexpectedly when an Exim client was
5235 using GnuTLS and trying to read a response, the client delivery process
5236 crashed while trying to generate an error log message.
5238 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5239 string, for example:
5241 warn message = H1: something\nH2: something
5243 the text was added as a single header line from Exim's point of view
5244 though it ended up OK in the delivered message. However, searching for the
5245 second and subsequent header lines using $h_h2: did not work. This has been
5246 fixed. Similarly, if a system filter added multiple headers in this way,
5247 the routers could not see them.
5249 20. Expanded the error message when iplsearch is called with an invalid key to
5250 suggest using net-iplsearch in a host list.
5252 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5253 ACLs are no longer actually imposed (and a message to that effect is
5256 22. If a "gecos" field in a passwd entry contained escaped characters, in
5257 particular, if it contained a \" sequence, Exim got it wrong when building
5258 a From: or a Sender: header from that name. A second bug also caused
5259 incorrect handling when an unquoted " was present following a character
5260 that needed quoting.
5262 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5263 was not being matched caselessly.
5265 24. Arranged for all hyphens in the exim.8 source to be escaped with
5268 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5269 properly. Recipient callout cache records were still being keyed to include
5270 the sender, even when use_sender was set false. This led to far more
5271 callouts that were necessary. The sender is no longer included in the key
5272 when use_sender is false.
5274 26. Added "control = submission" modifier to ACLs.
5276 27. Added the ${base62d: operator to decode base 62 numbers.
5278 28. dnsdb lookups can now access SRV records.
5280 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5281 the configuration file.
5283 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5284 (-v) mode. This makes the output for a verbose queue run more intelligible.
5286 31. Added a use_postmaster feature to recipient callouts.
5288 32. Added the $body_zerocount variable, containing the number of binary zero
5289 bytes in the message body.
5291 33. The time of last modification of the "new" subdirectory is now used as the
5292 "mailbox time last read" when there is a quota error for a maildir
5295 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5297 35. Added +ignore_unknown as a special item in host lists.
5299 36. Code for decoding IPv6 addresses in host lists is now included, even if
5300 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5301 address was recognized as an IP address, but was then not correctly decoded
5302 into binary, causing unexpected and incorrect effects when compared with
5309 1. Very minor rewording of debugging text in manualroute to say "list of
5310 hosts" instead of "hostlist".
5312 2. If verify=header_syntax was set, and a header line with an unqualified
5313 address (no domain) and a large number of spaces between the end of the
5314 name and the colon was received, the reception process suffered a buffer
5315 overflow, and (when I tested it) crashed. This was caused by some obsolete
5316 code that should have been removed. The fix is to remove it!
5318 3. When running in the test harness, delay a bit after writing a bounce
5319 message to get a bit more predictability in the log output.
5321 4. Added a call to search_tidyup() just before forking a reception process. In
5322 theory, someone could use a lookup in the expansion of smtp_accept_max_
5323 per_host which, without the tidyup, could leave open a database connection.
5325 5. Added the variables $recipient_data and $sender_data which get set from a
5326 lookup success in an ACL "recipients" or "senders" condition, or a router
5327 "senders" option, similar to $domain_data and $local_part_data.
5329 6. Moved the writing of debug_print from before to after the "senders" test
5332 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5333 problems for message scanning, either using a data ACL, or using
5334 local_scan() because the Received: header was not generated till after they
5335 were called (in order to set the time as the time of reception completion).
5336 I have revised the way this works. The header is now generated after the
5337 body is received, but before the ACL or local_scan() are called. After they
5338 are run, the timestamp in the header is updated.
5344 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5345 before starting a queue runner without re-exec. This happened only when
5346 deliver_drop_privilege was set or when the Exim user was set to root. The
5347 effect of the bug was that timeouts during subsequent deliveries caused
5348 crashes instead of being properly handled. The handler is now left at its
5349 default (and expected) setting.
5351 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5352 message, again when deliver_drop_privilege is set or Exim is run as root.
5353 The bug described in (1) was not present in this case, but the tidying up
5354 of the other signals was missing. I have made the two cases consistent.
5356 3. The ignore_target_hosts setting on a manualroute router was being ignored
5357 for hosts that were looked up using the /MX notation.
5359 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5362 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5363 operated on the sender address. After changing the $sender_address to <>
5364 for the sender address verify, Exim was re-instated it as the original
5365 (before rewriting) address, but remembering that it had rewritten it, so it
5366 wasn't rewriting it again. This bug also had the effect of breaking the
5367 sender address verification caching when the sender address was rewritten.
5369 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5370 This has been changed so that if the ip literal address matches
5371 ignore_target_hosts, the router declines.
5373 7. Added expansion conditions match_domain, match_address, and match_local_
5374 part (NOT match_host).
5376 8. The placeholder for the Received: header didn't have a length field set.
5378 9. Added code to Exim itself and to exim_lock to test for a specific race
5379 condition that could lead to file corruption when using MBX delivery. The
5380 issue is with the lockfile that is created in /tmp. If this file is removed
5381 after a process has opened it but before that process has acquired a lock,
5382 there is the potential for a second process to recreate the file and also
5383 acquire a lock. This could lead to two Exim processes writing to the file
5384 at the same time. The added code performs the same test as UW imapd; it
5385 checks after acquiring the lock that its file descriptor still refers to
5386 the same named file.
5388 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5389 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5392 11. Added the smtp_active_hostname option. If used, this will typically be made
5393 to depend on the incoming interface address. Because $interface_address is
5394 not set up until the daemon has forked a reception process, error responses
5395 that can happen earlier (such as "too many connections") no longer contain
5398 12. If an expansion in a condition on a "warn" statement fails because a lookup
5399 defers, the "warn" statement is abandoned, and the next ACL statement is
5400 processed. Previously this caused the whole ACL to be aborted.
5402 13. Added the iplsearch lookup type.
5404 14. Added ident_timeout as a log selector.
5406 15. Added tls_certificate_verified as a log selector.
5408 16. Added a global option tls_require_ciphers (compare the smtp transport
5409 option of the same name). This controls incoming TLS connections.
5411 17. I finally figured out how to make tls_require_ciphers do a similar thing
5412 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5413 before starting the TLS session.
5415 18. Tabs are now shown as \t in -bP output.
5417 19. If the log selector return_path_on_delivery was set, Exim crashed when
5418 bouncing a message because it had too many Received: header lines.
5420 20. If two routers both had headers_remove settings, and the first one included
5421 a superfluous trailing colon, the final name in the first list and the
5422 first name in the second list were incorrectly joined into one item (with a
5423 colon in the middle).
5429 1. Added -C and -D options to the exinext utility, mainly to make it easier
5430 to include in the automated testing, but these could be helpful when
5431 multiple configurations are in use.
5433 2. The exinext utility was not formatting the output nicely when there was
5434 an alternate port involved in the retry record key, nor when there was a
5435 message id as well (for retries that were specific to a specific message
5436 and a specific host). It was also confused by IPv6 addresses, because of
5437 the additional colons they contain. I have fixed the IPv4 problem, and
5438 patched it up to do a reasonable job for IPv6.
5440 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5441 delivery, the log line now contains "pipelined" if PIPELINING was used.
5443 4. An SMTP transport process used to panic and die if the bind() call to set
5444 an explicit outgoing interface failed. This has been changed; it is now
5445 treated in the same way as a connect() failure.
5447 5. A reference to $sender_host_name in the part of a conditional expansion
5448 that was being skipped was still causing a DNS lookup. This no longer
5451 6. The def: expansion condition was not recognizing references to header lines
5452 that used bh_ and bheader_.
5454 7. Added the _cache feature to named lists.
5456 8. The code for checking quota_filecount in the appendfile transport was
5457 allowing one more file than it should have been.
5459 9. For compatibility with Sendmail, the command line option
5467 and sets the incoming protocol and host name (for trusted callers). The
5468 host name and its colon can be omitted when only the protocol is to be set.
5469 Note the Exim already has two private options, -pd and -ps, that refer to
5470 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5471 "s", but I don't think that's a major issue.
5473 10. A number of refactoring changes to the code, none of which should affect
5476 (a) The number of logging options was getting close to filling up the
5477 32-bit word that was used as a bit map. I have split them into two classes:
5478 those that are passed in the argument to log_write(), and those that are
5479 only ever tested independently outside of that function. These are now in
5480 separate 32-bit words, so there is plenty of room for expansion again.
5481 There is no change in the user interface or the logging behaviour.
5483 (b) When building, for example, log lines, the code previously used a
5484 macro that called string_cat() twice, in order to add two strings. This is
5485 not really sufficiently general. Furthermore, there was one instance where
5486 it was actually wrong because one of the argument was used twice, and in
5487 one call a function was used. (As it happened, calling the function twice
5488 did not affect the overall behaviour.) The macro has been replaced by a
5489 function that can join an arbitrary number of extra strings onto a growing
5492 (c) The code for expansion conditions now uses a table and a binary chop
5493 instead of a serial search (which was left over from when there were very
5494 few conditions). Also, it now recognizes conditions like "pam" even when
5495 the relevant support is not compiled in: a suitably worded error message is
5496 given if an attempt is made to use such a condition.
5498 11. Added ${time_interval:xxxxx}.
5500 12. A bug was causing one of the ddress fields not to be passed back correctly
5501 from remote delivery subprocesses. The field in question was not being
5502 subsequently used, so this caused to problems in practice.
5504 13. Added new log selectors queue_time and deliver_time.
5506 14. Might have fixed a bug in maildirsizefile handling that threw up
5507 "unexpected character" debug warnings, and recalculated the data
5508 unnecessarily. In any case, I expanded the warning message to give more
5511 15. Added the message "Restricted characters in address" to the statements in
5512 the default ACL that block characters like @ and % in local parts.
5514 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5515 Three changes have been made:
5517 (a) There was a serious bug; a negative response to MAIL caused the whole
5518 recipient domain to be cached as invalid, thereby blocking all messages
5519 to all local parts at the same domain, from all senders. This bug has
5520 been fixed. The domain is no longer cached after a negative response to
5521 MAIL if the sender used is not empty.
5523 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5526 (c) A new callout option, "use_sender" has been added for people who want
5527 the modified behaviour.
5533 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5536 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5537 indeed breaks things for older releases.
5539 3. Added additional logging to the case where there is a problem reading data
5540 from a filter that is running in a subprocess using a pipe, in order to
5541 try to track down a specific problem.
5543 4. Testing facility fudge: when running in the test harness and attempting
5544 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5545 getting "No route to host". Convert this to a timeout.
5547 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5550 6. Some OS don't have socklen_t but use size_t instead. This affects the
5551 fifth argument of getsockopt() amongst other things. This is now
5552 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5553 can be set for individual OS. I have set it for SunOS5, OSF1, and
5554 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5555 some earlier ones do not.
5557 7. Change 4.30/15 was not doing the test caselessly.
5559 8. The standard form for an IPv6 address literal was being rejected by address
5560 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5561 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5562 this, as well as the form without the "IPv6" on the front (but only when
5563 address literals are enabled, of course).
5565 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5567 10. Exim crashed if a message with an empty sender address specified by -f
5568 encountered a router with an errors_to setting. This could be provoked only
5569 by a command such as
5573 where an empty string was supplied; "<>" did not hit this bug.
5575 11. Installed PCRE release 4.5.
5577 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5578 remained set. It is now erased.
5580 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5581 times from message ids (which are base 36 rather than the normal 62).
5583 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5584 were being counted as actual protocol errors, and logged if the log
5585 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5586 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5587 and DATA following a set of rejected RCPTs do not count as protocol errors.
5588 In other words, Exim assumes they were pipelined, though this may not
5589 actually be the case. Of course, in all cases the client gets an
5590 appropriate error code.
5592 15. If a lookup fails in an ACL condition, a message about the failure may
5593 be available; it is used if testing the ACL cannot continue, because most
5594 such messages specify what the cause of the deferral is. However, some
5595 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5596 that caused an old message to be retained and used if a later statement
5597 caused a defer, replacing the real cause of the deferral.
5599 16. If an IP address had so many PTR records that the DNS lookup buffer
5600 was not large enough to hold them, Exim could crash while trying to process
5601 the truncated data. It now detects and logs this case.
5603 17. Further to 4.21/58, another change has been made: if (and only if) the
5604 first line of a message (the first header line) ends with CRLF, a bare LF
5605 in a subsequent header line has a space inserted after it, so as not to
5606 terminate the header.
5608 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5609 unnecessarily, used atoi() instead of strtol(), and didn't check the
5610 termination when getting file sizes from file names by regex.
5612 19. Completely re-implemented the support for maildirsize files, in the light
5613 of a number of problems with the previous contributed implementation
5614 (4.30/29). In particular:
5616 . If the quota is zero, the maildirsize file is maintained, but no quota is
5619 . If the maildir directory does not exist, it is created before any attempt
5620 to write a maildirsize file.
5622 . The quota value in the file is just a cache; if the quota is changed in
5623 the transport, the new value overrides.
5625 . A regular expression is available for excluding directories from the
5628 20. The autoreply transport checks the characters in options that define the
5629 message's headers; it allows continued headers, but it was checking with
5630 isspace() after an embedded newline instead of explicitly looking for a
5633 21. If all the "regular" hosts to which an address was routed had passed their
5634 expiry times, and had not reached their retry times, the address was
5635 bounced, even if fallback hosts were defined. Now Exim should go on to try
5638 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5639 equivalent code in the SMTP transport. Some hosts send humungous responses
5640 to HELO/EHLO, more than 1024 it seems.
5642 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5643 gives compiler warnings in some environments. I've now done it "properly",
5646 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5647 (because of problems with the built-in one) was declared to return uschar *
5648 instead of char *, causing compiler failure.
5650 25. Fixed a file descriptor leak when processing alias/forward files.
5652 26. Fixed a minor format string issue in dbfn.c.
5654 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5656 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5657 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5660 29. When a sender address is verified, it is cached, to save repeating the test
5661 when there is more than one recipient in a message. However, when the
5662 verification involves a callout, it is possible for different callout
5663 options to be set for different recipients. It is too complicated to keep
5664 track of this in the cache, so now Exim always runs a verification when a
5665 callout is required, relying on the callout cache for the optimization.
5666 The overhead is duplication of the address routing, but this should not be
5669 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5670 to be verified with callout=postmaster, and the main callout worked but the
5671 postmaster check failed, the verification correctly failed. However, if a
5672 subsequent RCPT command asked for sender verification *without* the
5673 postmaster check, incorrect caching caused this verification also to fail,
5676 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5677 it was not caching the DNS options (qualify_single, search_parents) that
5678 were used when the lookup failed. A subsequent lookup with different
5679 options therefore always gave the same answer, though there were cases
5680 where it should not have. (Example: a "domains = !$mx_any" option on a
5681 dnslookup router: the "domains" option is always processed without any
5682 widening, but the router might have qualify_single set.) Now Exim uses the
5683 cached value only when the same options are set.
5685 32. Added John Jetmore's "exipick" utility to the distribution.
5687 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5688 than a timeout (e.g. a certificate is required, and is not provided), an
5689 Exim server now closes the connection immediately. Previously it waited for
5690 the client to close - but if the client is SSL, it seems that they each
5691 wait for each other, leading to a delay before one of them times out.
5693 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5694 maintained 0.8.x compatibility because I don't think many are using it, and
5695 it is clearly obsolete.
5697 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5700 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5701 client certificate was expired. A simple patch fixes this, though I don't
5702 understand the full logic of why the verify callback is called multiple
5705 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5706 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5707 which causes problems with some clients (such as the Certicom SSL Plus
5708 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5709 disables the coutermeasure allowing Eudora to connect."
5711 38. Exim was not checking that a write() to a log file succeeded. This could
5712 lead to Bad Things if a log got too big, in particular if it hit a file
5713 size limit. Exim now panics and dies if it cannot write to a log file, just
5714 as it does if it cannot open a log file.
5716 39. Modified OS/Makefile-Linux so that it now contains
5718 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5720 The two -D definitions ensure that Exim is compiled with large file
5721 support, which makes it possible to handle log files that are bigger than
5724 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5725 instance) a domain was checked against a named list that involved a lookup,
5726 causing $domain_data to be set, then another domain was checked against the
5727 same list, then the first domain was re-checked, the value of $domain_data
5728 after the final check could be wrong. In particular, if the second check
5729 failed, it could be set empty. This bug probably also applied to
5732 41. The strip_trailing_dot option was not being applied to the address given
5733 with the -f command-line option.
5735 42. The code for reading a message's header from the spool was incrementing
5736 $received_count, but never initializing it. This meant that the value was
5737 incorrect (doubled) while delivering a message in the same process in which
5738 it was received. In the most common configuration of Exim, this never
5739 happens - a fresh exec is done - but it can happen when
5740 deliver_drop_privilege is set.
5742 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5743 it now includes up to 150 characters of the unexpected data in the log
5746 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5747 and building data strings. The size of both of these buffers was 10 000
5748 bytes - far larger than anybody would *ever* want, thought I. Needless to
5749 say, somebody hit the limit. I have increased the maximum line length to
5750 20 000 and the maximum data length of concatenated lines to 100 000. I have
5751 also fixed two bugs, because there was no checking on these buffers. Tsk,
5752 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5753 buffer is too small.
5755 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5756 lsearch lookups. Now it does.
5758 46. When parsing a route_list item in a manualroute router, a fixed-length
5759 buffer was used for the list of hosts. I made this 1024 bytes long,
5760 thinking that nobody would ever have a list of hosts that long. Wrong.
5761 Somebody had a whole pile of complicated expansion conditions, and the
5762 string was silently truncated, leading to an expansion error. It turns out
5763 that it is easier to change to an unlimited length (owing to other changes
5764 that have happened since this code was originally written) than to build
5765 structure for giving a limitation error. The length of the item that
5766 expands into the list of hosts is now unlimited.
5768 47. The lsearch lookup could not handle data where the length of text line was
5769 more than 4095 characters. Such lines were truncated, leading to shortened
5770 data being returned. It should now handle lines of any length.
5772 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5773 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5776 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5777 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5778 no longer happens in comment lines. A list of replacements is now placed
5779 at the head of all of the source files, except those whose only change is
5780 to replace PERL_COMMAND in the very first #! line.
5782 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5783 messages on the queue, with a bottom-up merge sort, using code contributed
5784 by Michael Haardt. This should make operations like -bp somewhat faster on
5785 large queues. It won't affect queue runners, except when queue_run_in_order
5788 51. Installed eximstats 1.31 in the distribution.
5790 52. Added support for SRV lookups to the dnslookup router.
5792 53. If an ACL referred to $message_body or $message_body_end, the value was not
5793 reset for any messages that followed in the same SMTP session.
5795 54. The store-handling optimization for building very long strings was not
5796 differentiating between the different store pools. I don't think this
5797 actually made any difference in practice, but I've tidied it.
5799 55. While running the routers to verify a sender address, $sender_address
5800 was still set to the sender address. This is wrong, because when routing to
5801 send a bounce to the sender, it would be empty. Therefore, I have changed
5802 it so that, while verifying a sender address, $sender_address is set to <>.
5803 (There is no change to what happens when verifying a recipient address.)
5805 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5806 target A or AAAA records (if not already returned) without resetting the
5807 qualify_single or search_parents options of the DNS resolver. These are
5808 inappropriate in this case because the targets of MX and SRV records must
5809 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5810 target that, when qualified, matched something in the local domain. These
5811 two options are now turned off when doing these lookups.
5813 57. It seems that at least some releases of Reiserfs (which does not have the
5814 concept of a fixed number of inodes) returns zero and not -1 for the
5815 number of available inodes. This interacted badly with check_spool_inodes,
5816 which assumed that -1 was the "no such thing" setting. What I have done is
5817 to check that the total number of inodes is greater than zero before doing
5818 the test of how many are available.
5820 58. When a "warn" ACL statement has a log_message modifier, the message is
5821 remembered, and not repeated. This is to avoid a lot of repetition when a
5822 message has many recipients that cause the same warning to be written.
5823 However, Exim was preserving the list of already written lines for an
5824 entire SMTP session, which doesn't seem right. The memory is now reset if a
5825 new message is started.
5827 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5828 debugging output unless log_rewrite was also set.
5830 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5831 of (int)(handle) when we know that handle contains (void *)(-1).
5833 61. The Exim daemon panic-logs an error return when it closes the incoming
5834 connection. However "connection reset by peer" seems to be common, and
5835 isn't really an error worthy of noting specially, so that particular error
5838 62. When Exim is trying to find all the local interfaces, it used to panic and
5839 die if the ioctl to get the interface flags failed. However, it seems that
5840 on at least one OS (Solaris 9) it is possible to have an interface that is
5841 included in the list of interfaces, but for which you get a failure error
5842 for this call. This happens when the interface is not "plumbed" into a
5843 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5844 failure of the "get flags" call assumes that the interface is down.
5846 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5847 makes life easier for people who are doing arithmetic on fields extracted
5848 from dates, where you often get leading zeros that should not be
5849 interpreted as octal.
5851 64. Added qualify_domain to the redirect router, to override the global
5854 65. If a pathologically long header line contained very many addresses (the
5855 report of this problem mentioned 10 000) and each of them was rewritten,
5856 Exim could use up a very large amount of memory. (It kept on making new
5857 copies of the header line as it rewrote, and never released the old ones.)
5858 At the expense of a bit more processing, the header rewriting function has
5859 been changed so that it no longer eats memory in this way.
5861 66. The generation of the Received: header has been moved from the time that a
5862 message starts to be received, to the time that it finishes. The timestamp
5863 in the Received: header should now be very close to that of the <= log
5864 line. There are two side-effects of this change:
5866 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5867 logged header lines no longer include the local Received: line, because
5868 it has not yet been created. The same applies to a copy of the message
5869 that is returned to a non-SMTP sender when a message is rejected.
5871 (b) When a filter file is tested using -bf, no additional Received: header
5872 is added to the test message. After some thought, I decided that this
5875 This change does not affect the value of $received_for. It is still set
5876 after address rewriting, but before local_scan() is called.
5878 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5880 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5881 gave an unhelpful panic error message, and a defer error. I have managed to
5882 change this behaviour so that it now rejects any supplied certificate,
5883 which seems right, as the list of acceptable certificates is empty.
5885 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5886 gave an unhelpful defer error. I have not managed to make this reject any
5887 supplied certificates, but the error message it gives is "no certificate
5888 supplied", which is not helpful.
5890 70. exigrep's output now also includes lines that are not associated with any
5891 message, but which match the given pattern. Implemented by a patch from
5892 Martin Sluka, which also tidied up the Perl a bit.
5894 71. Recipient callout verification, like sender verification, was using <> in
5895 the MAIL FROM command. This isn't really the right thing, since the actual
5896 sender may affect whether the remote host accepts the recipient or not. I
5897 have changed it to use the actual sender in the callout; this means that
5898 the cache record is now keyed on a recipient/sender pair, not just the
5899 recipient address. There doesn't seem to be a real danger of callout loops,
5900 since a callout by the remote host to check the sender would use <>.
5901 [SEE ABOVE: changed after hitting problems.]
5903 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5904 temporary errors. However, in the case of such a code being given after
5905 the end of a data transmission (i.e. after ".") Exim was failing to write
5906 a retry record for the message. (Yes, there was some broken host that was
5907 actually sending 8xx at this point.)
5909 73. An unknown lookup type in a host list could cause Exim to panic-die when
5910 the list was checked. (An example that provoked this was putting <; in the
5911 middle of a list instead of at the start.) If this happened during a DATA
5912 ACL check, a -D file could be left lying around. This kind of configuration
5913 error no longer causes Exim to die; instead it causes a defer error. The
5914 incident is still logged to the main and panic logs.
5916 74. Buglet left over from Exim 3 conversion. The message "too many messages
5917 in one connection" was written to the rejectlog but not the mainlog, except
5918 when address rewriting (yes!) was being logged.
5920 75. Added write_rejectlog option.
5922 76. When a system filter was run not as root (that is, when system_filter_user
5923 was set), the values of the $n variables were not being returned to the
5924 main process; thus, they were not subsequently available in the $sn
5927 77. Added +return_path_on_delivery log selector.
5929 78. A connection timeout was being treated differently from recipients deferred
5930 when testing hosts_max_try with a message that was older than the host's
5931 retry timeout. (The host should not be counted, thus allowing all hosts to
5932 be tried at least once before bouncing.) This may have been the cause of an
5933 occasionally reported bug whereby a message would remain on the queue
5934 longer than the retry timeout, but would be bounced if a delivery was
5935 forced. I say "may" because I never totally pinned down the problem;
5936 setting up timeout/retry tests is difficult. See also the next item.
5938 79. The ultimate address timeout was not being applied to errors that involved
5939 a combination of host plus message (for example, a timeout on a MAIL
5940 command). When an address resolved to a number of possible hosts, and they
5941 were not all tried for each delivery (e.g. because of hosts_max_try), a
5942 message could remain on the queue longer than the retry timeout.
5944 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5947 81. Fixed an obscure SMTP outgoing bug which required at least the following
5948 conditions: (a) there was another message waiting for the same server;
5949 (b) the server returned 5xx to all RCPT commands in the first message so
5950 that the message was not completed; (c) the server dropped the connection
5951 or gave a negative response to the RSET that Exim sends to abort the
5952 transaction. The observed case was a dropped connection after DATA that had
5953 been sent in pipelining mode. That is, the server had advertised PIPELINING
5954 but was not implementing it correctly. The effect of the bug was incorrect
5955 behaviour, such as trying another host, and this could lead to a crash.
5961 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5962 and daemon.c were passed as pointers to ints; they should have been
5963 pointers to socklen_t variables (which are typically unsigned ints).
5965 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5968 3. Fixed a really odd bug that affected only the testing scheme; patching a
5969 certain fixed string in the binary changed the value of another string that
5970 happened to be identical to the end of the original first string.
5972 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5973 name", it returns that address as the IP address. On some operating
5974 systems (e.g. Solaris), it also passes back the IP address string as the
5975 "host name". However, on others (e.g. Linux), it passes back an empty
5976 string. Exim wasn't checking for this, and was changing the host name to an
5977 empty string, assuming it had been canonicalized.
5979 5. Although rare, it is permitted to have more than one PTR record for a given
5980 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5981 all the names associated with an address, because they do in Solaris.
5982 However, it seems that they do not in Linux for data that comes from the
5983 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5984 I found this out when I moved to a new Linux workstation and tried to run
5985 the Exim test suite.
5987 To get round this problem I have changed the code so that it now does its
5988 own call to the DNS to look up PTR records when searching for a host name.
5989 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5990 addresses that are only in /etc/hosts are still found.
5992 This behaviour is, however, controlled by an option called host_lookup_
5993 order, which defaults to "bydns:byaddr". If people want to use the other
5994 order, or indeed, just use one or the other means of lookup, they can
5995 specify it in this variable.
5997 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5998 some operating systems, this comes back from gethostbyaddr() as an empty
5999 string, and this is what Exim used to test for. However, it seems that in
6000 other systems, "." is yielded. Exim now tests for this case too.
6002 7. The values of check_spool_space and check_log_space are now held internally
6003 as a number of kilobytes instead of an absolute number of bytes. If a
6004 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6005 kilobyte. This means that much larger values can be stored.
6007 8. Exim monitor: an attempt to get the action menu when not actually pointing
6008 at a message produces an empty menu entitled "No message selected". This
6009 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6010 no entries in it ("Shell widget menu has zero width and/or height"). So I
6011 have added a single, blank menu entry in this case.
6013 9. Added ${quote_local_part.
6015 10. MIME decoding is now applied to the contents of Subject: header lines when
6018 11. Now that a reference to $sender_host_address automatically causes a reverse
6019 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6020 host lookup before query-style lookups in lists that might use this
6021 variable. This has therefore been abolished, and the "net-" prefix is no
6022 longer necessary for query-style lookups.
6024 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6025 have been a typo for LFLAGS, so it has been changed.
6027 13. The install script calls Exim with "-C /dev/null" in order to find the
6028 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6029 to be output. However, since Exim outputs its version number before the
6030 error, it didn't break the script. It just looked ugly. I fixed this by
6031 always allowing "-C /dev/null" if the caller is root.
6033 14. Ignore overlarge ACL variable number when reading spool file - insurance
6034 against a later release with more variables having written the file.
6036 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6037 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6038 this, as well as the form without the "IPv6" on the front.
6040 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6041 OS/Makefile-Darwin file.
6043 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6044 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6046 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6047 variable. However, some OS use ints for inodes. I've added cast to long int
6048 to get rid of the compiler warning.
6050 19. I had forgotten to lock out "/../" in configuration file names when
6051 ALT_CONFIG_PREFIX was set.
6053 20. Routers used for verification do not need to specify transports. However,
6054 if such a router generated a host list, and callout was configured, Exim
6055 crashed, because it could not find a port number from the (non-existent)
6056 transport. It now assumes port 25 in this circumstance.
6058 21. Added the -t option to exigrep.
6060 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6061 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6062 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6063 them is set, LOOKUP_LSEARCH is forced.
6065 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6067 24. Added sender and host information to the "rejected by local_scan()" log
6068 line; previously there was no indication of these.
6070 25. Added .include_if_exists.
6072 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6073 Linux. It turns out that not all file systems support this. Apparently some
6074 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6075 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6076 not supported on the file descriptor, is now ignored when Exim is trying to
6077 sync a directory. This applies only to Linux.
6079 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6081 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6082 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6085 29. Added support for maildirsize files from supplied patch (modified a bit).
6087 30. The use of :fail: followed by an empty string could lead Exim to respond to
6088 sender verification failures with (e.g.):
6090 550 Verification failed for <xxx>
6091 550 Sender verify failed
6093 where the first response line was missing the '-' that indicates it is not
6094 the final line of the response.
6096 31. The loop for finding the name of the user that called Exim had a hardwired
6097 limit of 10; it now uses the value of finduser_retries, which is used for
6098 all other user lookups.
6100 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6103 33. Exim was neglecting to zero errno before one call of strtol() when
6104 expanding a string and expecting an integer value. On some systems this
6105 resulted in spurious "integer overflow" errors. Also, it was casting the
6106 result into an int without checking.
6108 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6109 did not work. The code looks as if it has *never* worked, though it appears
6110 to have been documented since at least release 1.62. I have made it work.
6112 35. The "timeout_DNS" error in retry rules, also documented since at least
6113 1.62, also never worked. As it isn't clear exactly what this means, and
6114 clearly it isn't a major issue, I have abolished the feature by treating it
6115 as "timeout", and writing a warning to the main and panic logs.
6117 36. The display of retry rules for -brt wasn't always showing the error code
6120 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6121 timeout_connect_A, timeout_connect_MX.
6123 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6124 to the empty sender.
6126 39. The daemon was not analysing the content of -oX till after it had closed
6127 stderr and disconnected from the controlling terminal. This meant that any
6128 syntax errors were only noted on the panic log, and the return code from
6129 the command was 0. By re-arranging the code a little, I've made the
6130 decoding happen first, so such errors now appear on stderr, and the return
6131 code is 1. However, the actual setting up of the sockets still happens in
6132 the disconnected process, so errors there are still only recorded on the
6135 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6136 connections (as happens on some IP stacks) was logged at start up time as
6137 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6138 it from "IPv6 and IPv4", which means that two separate sockets are being
6141 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6142 says whether AF_INET or AF_INET6 was passed as an argument.
6144 42. Exiwhat output was messed up when time zones were included in log
6147 43. Exiwhat now gives more information about the daemon's listening ports,
6148 and whether -tls-on-connect was used.
6150 44. The "port" option of the smtp transport is now expanded.
6152 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6153 silently ignored. Now an error message is written to the main and panic
6156 46. There's a new ACL modifier called "logwrite" which writes to a log file
6157 as soon as it is encountered.
6159 47. Added $local_user_uid and $local_user_gid at routing time.
6161 48. Exim crashed when trying to verify a sender address that was being
6164 49. Exim was recognizing only a space character after ".include". It now also
6165 recognizes a tab character.
6167 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6168 extracting the relevant information from the specification. The man page no
6169 longer contains scrambled data for the -d option, and I've added a section
6170 at the front about calling Exim under different names.
6172 51. Added "extra_headers" argument to the "mail" command in filter files.
6174 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6177 53. Installed eximstats 1.29.
6179 54. Added transport_filter_timeout as a generic transport option.
6181 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6182 Cc: header lines. This was required by RFC 822, but it not required by RFC
6185 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6186 incoming messages that did not have them. Now it does so only if the
6187 message originates locally, that is, if there is no associated remote host
6188 address. When Resent- header lines are present, this applies to the Resent-
6189 lines rather than the non-Resent- lines.
6191 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6192 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6194 58. Messages for configuration errors now include the name of the main
6195 configuration file - useful now that there may be more than one file in a
6196 list (.included file names were always shown).
6198 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6199 for those rare installations that do not start the daemon as root or run it
6200 setuid root. I've cut out the call to initgroups() if the daemon is not
6203 60. The Exim user and group can now be bound into the binary as text strings
6204 that are looked up at the start of Exim's processing.
6206 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6208 62. Added $mailstore_basename variable.
6210 63. Installed patch to sieve.c from Michael Haardt.
6212 64. When Exim failed to open the panic log after failing to open the main log,
6213 the original message it was trying to log was written to stderr and debug
6214 output, but if they were not available (the usual case in production), it
6215 was lost. Now it is written to syslog before the two lines that record the
6216 failures to open the logs.
6218 65. Users' Exim filters run in subprocesses under the user's uid. It is
6219 possible for a "deliver" command or an alias in a "personal" command to
6220 provoke an address rewrite. If logging of address rewriting is configured,
6221 this fails because the process is not running as root or exim. There may be
6222 a better way of dealing with this, but for the moment (because 4.30 needs
6223 to be released), I have disabled address rewrite logging when running a
6224 filter in a non-root, non-exim process.
6230 1. The buildconfig auxiliary program wasn't quoting the value set for
6231 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6232 not defined. This bug was masked in 4.22 by the effect that was fixed in
6235 2. Some messages that were rejected after a message id was allocated were
6236 shown as "incomplete" by exigrep. It no longer does this for messages that
6237 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6239 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6240 have allow_domain_literals set, the ID did not get logged in the <= line.
6241 Domain literals are now always recognized in Message-ID: header lines.
6243 4. The first argument for a ${extract expansion item is the key name or field
6244 number. Leading and trailing spaces in this item were not being ignored,
6245 causing some misleading effects.
6247 5. When deliver_drop_privilege was set, single queue runner processes started
6248 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6249 same command in the process it spins off) were not dropping privilege.
6251 6. When the daemon running as "exim" started a queue runner, it always
6252 re-executed Exim in the spun-off process. This is a waste of effort when
6253 deliver_drop_privilege is set. The new process now just calls the
6254 queue-runner function directly.
6260 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6263 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6264 ignored. Though the use of -oP was forcing the writing of a pid file, it
6265 was always written to the default place.
6267 3. If the message "no IP address found for host xxxx" is generated during
6268 incoming verification, it is now followed by identification of the incoming
6269 connection (so you can more easily find what provoked it).
6271 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6273 5. Added some features to "harden" Exim a bit more against certain attacks:
6275 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6276 be put in Local/Makefile. This is like the never_users runtime option,
6277 but it cannot be overridden. The default setting is "root".
6279 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6280 prefix string with which any file named in a -C command line option
6283 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6284 is retained for -C and -D only if the caller of Exim is root. Without
6285 it, the exim user may also use -C and -D and retain privilege.
6287 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6288 command line option is disabled.
6290 6. Macro names set by the -D option must start with an upper case letter, just
6291 like macro names defined in the configuration file.
6293 7. Added "dereference=" facility to LDAP.
6295 8. Two instances of the typo "uknown" in the source files are fixed.
6297 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6298 the Configure-Makefile script screwed up while processing it.
6300 10. Incorporated PCRE 4.4.
6302 11. The SMTP synchronization check was not operating right at the start of an
6303 SMTP session. For example, it could not catch a HELO sent before the client
6304 waited for the greeting. There is now a check for outstanding input at the
6305 point when the greeting is written. Because of the duplex, asynchronous
6306 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6307 way, but not yet received, when the check is performed.
6309 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6310 on TCP/IP sockets, because this apparently causes some broken clients to
6313 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6314 unchanged) from the Cygwin maintainer.
6316 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6317 was supported instead of testing for mbx. Effectively a typo.
6319 15. The spa authenticator server code was not checking that the input it
6320 received was valid base64.
6322 16. The debug output line for the "set" modifier in ACLs was not showing the
6323 name of the variable that was being set.
6325 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6327 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6328 lookup on its own. Something else was needed to trigger the lookup. For
6329 example, a match in host_lookup or the need for a host name in a host list.
6330 Now, if $sender_host_name is referenced and the host name has not yet been
6331 looked up, a lookup is performed. If the lookup fails, the variable remains
6332 empty, and $host_lookup_failed is set to "1".
6334 19. Added "eqi" as a case-independent comparison operator.
6336 20. The saslauthd authentication condition could segfault if neither service
6337 nor realm was specified.
6339 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6340 error message that was logged was misleading, and incoming SMTP
6341 connections were dropped. The message is now more accurate, and temporary
6342 errors are given to SMTP connections.
6344 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6345 (except RSET) with a 421 error, until QUIT is received. However, it was
6346 failing to send a response to QUIT.
6348 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6349 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6350 ACL is now run after the helo_try_verify_hosts code.
6352 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6353 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6354 case-independent, so other case variants are also recognized). Apparently
6355 some systems use these upper case variants.
6357 25. If more than two messages were waiting for the same host, and a transport
6358 filter was specified for the transport, Exim sent two messages over the
6359 same TCP/IP connection, and then failed with "socket operation on non-
6360 socket" when it tried to send the third.
6362 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6364 27. The extern definition of crypt16() in expand.c was not being excluded when
6365 the OS had its own crypt16() function.
6367 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6368 as a preferred synonym for return_size_limit, both as an option and as an
6371 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6373 30. Changed the default configuration ACL to relax the local part checking rule
6374 for addresses that are not in any local domains. For these addresses,
6375 slashes and pipe symbols are allowed within local parts, but the sequence
6376 /../ is explicitly forbidden.
6378 31. SPA server authentication was not clearing the challenge buffer before
6381 32. log_message in a "warn" ACL statement was writing to the reject log as
6382 well as to the main log, which contradicts the documentation and doesn't
6383 seem right (because no rejection is happening). So I have stopped it.
6385 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6386 However, I am unable to do any testing of this.
6388 34. Fixed an infelicity in the appendfile transport. When checking directories
6389 for a mailbox, to see if any needed to be created, it was accidentally
6390 using path names with one or more superfluous leading slashes; tracing
6391 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6393 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6394 contents are added to the log line that is written for every discarded
6395 recipient. (Previously a log_message setting was ignored.)
6397 36. The ${quote: operator now quotes the string if it is empty.
6399 37. The install script runs exim in order to find its version number. If for
6400 some reason other than non-existence or emptiness, which it checks, it
6401 could not run './exim', it was installing it with an empty version number,
6402 i.e. as "exim-". This error state is now caught, and the installation is
6405 38. An argument was missing from the function that creates an error message
6406 when Exim fails to connect to the socket for saslauthd authentication.
6407 This could cause Exim to crash, or give a corrupted message.
6409 39. Added isip, isip4, and isip6 to ${if conditions.
6411 40. The ACL variables $acl_xx are now saved with the message, and can be
6412 accessed later in routers, transports, and filters.
6414 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6415 strings in the file are not string-expanded.
6417 42. If a MAIL command specified a SIZE value that was too large to fit into an
6418 int variable, the check against message_size_limit failed. Such values are
6419 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6420 day this will have to be increased, but I don't think I want to be around
6421 when emails are that large.
6428 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6429 iconv() is not standard in FreeBSD.
6431 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6432 IPv6 enabled. The observed symptom was a segmentation fault on return from
6433 the function os_common_find_running_interfaces() in src/os.c.
6435 3. In the check_special_case() function in daemon.c I had used "errno" as an
6436 argument name, which causes warnings on some systems. This was basically a
6437 typo, since it was named "eno" in the comments!
6439 4. The code that waits for the clock to tick (at a resolution of some fraction
6440 of a second) so as to ensure message-id uniqueness was always waiting for
6441 at least one whole tick, when it could have waited for less. [This is
6442 almost certainly not relevant at current processor speeds, where it is
6443 unlikely to ever wait at all. But we try to future-proof.]
6445 5. The function that sleeps for a time interval that includes fractions of a
6446 second contained a race. It did not block SIGALRM between setting the
6447 timer, and suspending (a couple of lines later). If the interval was short
6448 and the sigsuspend() was delayed until after it had expired, the suspension
6449 never ended. On busy systems this could lead to processes getting stuck for
6452 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6453 process, before it forks any delivery processes. The open lookup caching
6454 mechanism meant that the open file or database connection was passed into
6455 the delivery process. The problem was that delivery processes always tidy
6456 up cached lookup data. This could cause a problem for the next delivery
6457 process started by the queue runner, because the external queue runner
6458 process does not know about the closure. So the next delivery process
6459 still has data in the lookup cache. In the case of a file lookup, there was
6460 no problem because closing a file descriptor in a subprocess doesn't affect
6461 the parent. However, if the lookup was caching a connection to a database,
6462 the connection was closed, and the second delivery process was likely to
6463 see errors such as "PGSQL: query failed: server closed the connection
6464 unexpectedly". The problem has been fixed by closing all cached lookups
6465 in a queue runner before running a delivery process.
6467 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6468 seem to have the "const" qualifier which it has on other OS. I've
6471 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6472 *of the same type* (client or server) with the same public name that an
6473 error should be diagnosed.
6475 9. When Exim looked up a host name for an IP address, but failed to find the
6476 original IP address when looking up the host name (a safety check), it
6477 output the message "<ip address> does not match any IP for NULL", which was
6478 confusing, to say the least. The bug was that the host name should have
6479 appeared instead of "NULL".
6481 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6482 user that is built into the binary, and the -C or -D options is used, root
6483 privilege is dropped before the configuration file is read. In addition,
6484 logging is switched to stderr instead of the normal log files. If the
6485 configuration then re-defines the Exim user, the unprivileged environment
6486 is probably not what is expected, so Exim logs a panic warning message (but
6489 However, if deliver_drop_privilege is set, the unprivileged state may well
6490 be exactly what is intended, so the warning has been cut out in that case,
6491 and Exim is allowed to try to write to its normal log files.
6497 1. smtp_return_error_details was not giving details for temporary sender
6498 or receiver verification errors.
6500 2. Diagnose a configuration error if two authenticators have the same public
6503 3. Exim used not to create the message log file for a message until the first
6504 delivery attempt. This could be confusing when incoming messages were held
6505 for policy or load reasons. The message log file is now created at the time
6506 the message is received, and an initial "Received" line is written to it.
6508 4. The automatically generated man page for command line options had a minor
6509 bug that caused no ill effects; however, a more serious problem was that
6510 the procedure for building the man page automatically didn't always
6511 operate. Consequently, release 4.20 contains an out-of-date version. This
6512 shouldn't happen again.
6514 5. When building Exim with embedded Perl support, the script that builds the
6515 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6516 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6518 6. The freeze_tell option was not being used for messages that were frozen on
6519 arrival, either by an ACL or by local_scan().
6521 7. Added the smtp_incomplete_transaction log selector.
6523 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6524 was accepting AUTH without a new EHLO.
6526 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6527 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6530 10. Logging of TCP/IP connections (when configured) now happens in the main
6531 daemon process instead of the child process, so that the TCP/IP connection
6532 count is more accurate (but it can never be perfect).
6534 11. The use of "drop" in a nested ACL was not being handled correctly in the
6535 outer ACL. Now, if condition failure induced by the nested "drop" causes
6536 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6537 or "require"), the connection is dropped.
6539 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6540 that yield "discard" can now be used with an "accept" or a "discard" verb,
6541 but an error is generated for any others (because I can't see a useful way
6542 to define what should happen).
6544 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6545 are now processed in the same way as lines in the Exim configuration file.
6546 In particular, continuation lines are supported.
6548 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6550 15. Added -ti meaning -t -i.
6552 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6553 domains, and warn by logging if others are found.
6555 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6556 returned from GIFCONF when Exim is trying to find the list of interfaces on
6557 a host. The code in os.c has been modified to copy each ifreq to an aligned
6558 structure in all cases.
6560 Also, in some cases, the returned ifreq's were being copied to a 'struct
6561 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6562 means the last couple of bytes of an IPv6 address could be chopped if the
6563 ifreq contained only a normal sockaddr (14 bytes storage).
6565 18. Named domain lists were not supported in the hosts_treat_as_local option.
6566 An entry such as +xxxx was not recognized, and was treated as a literal
6569 19. Ensure that header lines added by a DATA ACL are included in the reject log
6570 if the ACL subsequently rejects the message.
6572 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6573 MD5 (which is deprecated).
6575 21. When testing a filter file using -bf, Exim was writing a message when it
6576 took the sender from a "From " line in the message, but it was not doing so
6577 when it took $return_path from a Return-Path: header line. It now does.
6579 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6580 with a valid header line field name (a series of printing characters
6581 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6583 23. Changed "disc" in the source to "disk" to conform to the documentation and
6584 the book and for uniformity.
6586 24. Ignore Sendmail's -Ooption=value command line item.
6588 25. When execve() failed while trying to run a command in a pipe transport,
6589 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6590 could be confused with a return value of 69 from the command itself. This
6591 has been changed to 127, the value the shell returns if it is asked to run
6592 a non-existent command. The wording for the related log line suggests a
6593 non-existent command as the problem.
6595 26. If received_header_text expands to an empty string, do not add a Received:
6596 header line to the message. (Well, it adds a token one on the spool, but
6597 marks it "old" so that it doesn't get used or transmitted.)
6599 27. Installed eximstats 1.28 (addition of -nt option).
6601 28. There was no check for failure on the call to getsockname() in the daemon
6602 code. This can fail if there is a shortage of resources on the system, with
6603 ENOMEM, for example. A temporary error is now given on failure.
6605 29. Contrary to the C standard, it seems that in some environments, the
6606 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6607 program. Exim now does this explicitly; it affects the formatting of
6608 timestamps using strftime().
6610 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6611 complaints. I've now initialized all the variables, to avoid this.
6613 32. Header lines added by a system filter were not being "seen" during
6614 transport-time rewrites.
6616 33. The info_callback() function passed to OpenSSL is set up with type void
6617 (*)(SSL *, int, int), as described somewhere. However, when calling the
6618 function (actually a macro) that sets it up, the type void(*)() is
6619 expected. I've put in a cast to prevent warnings from picky compilers.
6621 34. If a DNS black list lookup found a CNAME record, but there were no A
6622 records associated with the domain it pointed at, Exim crashed.
6624 35. If a DNS black list lookup returned more than one A record, Exim ignored
6625 all but the first. It now scans all returned addresses if a particular IP
6626 value is being sought. In this situation, the contents of the
6627 $dnslist_value variable are a list of all the addresses, separated by a
6630 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6631 to accept a host name and all its aliases if the forward lookup for any of
6632 them yielded the IP address of the incoming connection. Now it accepts only
6633 those names whose forward lookup yields the correct IP address. Any other
6634 names are discarded. This closes a loophole whereby a rogue DNS
6635 administrator could create reverse DNS records to break through a
6636 wildcarded host restriction in an ACL.
6638 37. If a user filter or a system filter that ran in a subprocess used any of
6639 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6640 the wrong values were passed to the pipe command ($thisaddress had the
6641 value of $0, $0 had the value of $1, etc). This bug was introduced by
6642 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6644 38. Improved the line breaking for long SMTP error messages from ACLs.
6645 Previously, if there was no break point between 40 and 75 characters, Exim
6646 left the rest of the message alone. Two changes have been made: (a) I've
6647 reduced the minimum length to 35 characters; (b) if it can't find a break
6648 point between 35 and 75 characters, it looks ahead and uses the first one
6649 that it finds. This may give the occasional overlong line, but at least the
6650 remaining text gets split now.
6652 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6653 file descriptors might be low, and that setting 1000 would always raise it.
6654 It turns out that in some environments, the limit is already over 1000 and
6655 that lowering it causes trouble. So now Exim takes care not to decrease it.
6657 40. When delivering a message, the value of $return_path is set to $sender_
6658 address at the start of routing (routers may change the value). By an
6659 oversight, this default was not being set up when an address was tested by
6660 -bt or -bv, which affected the outcome if any router or filter referred to
6663 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6664 log only when "message" or "log_message" are set. However, if one of the
6665 conditions was an address verification, or a call to a nested ACL, the
6666 messages generated by the underlying test were being passed through. This
6667 no longer happens. The underlying message is available in $acl_verify_
6668 message for both "message" and "log_message" expansions, so it can be
6669 passed through if needed.
6671 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6672 new expansion $bh_ to give the encoded byte string without charset
6673 translation. Translation happens only if iconv() is available; HAVE_ICONV
6674 indicates this at build time. HEADERS_CHARSET gives the charset to
6675 translate to; headers_charset can change it in the configuration, and
6676 "headers charset" can change it in an individual filter file.
6678 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6679 that creates RFC 2047 encoded "words" labels them as that charset instead
6680 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6681 expansion operator; (ii) when Exim creates a From: line for a local
6682 message; (iii) when a header line is rewritten to include a "phrase" part.
6684 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6685 buggy, causing it to skip the first lines of messages whose message ID
6686 ended in 'D'. This would not have bitten before Exim release 4.14, because
6687 message IDs were unlikely to end in 'D' before then. The effect was to have
6688 incorrect size information for certain domains.
6690 45. #include "config.h" was missing at the start of the crypt16.c module. This
6691 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6694 46. If there was a timeout during a "random" callout check, Exim treated it as
6695 a failure of the random address, and carried on sending RSET and the real
6696 address. If the delay was just some slowness somewhere, the response to the
6697 original RCPT would be taken as a response to RSET and so on, causing
6698 mayhem of various kinds.
6700 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6701 when I implemented it. It didn't allow for the fact that some option values
6702 may legitimately be negative (e.g. size_addition), and it didn't even do
6703 the right test for positive values.
6705 48. Domain names in DNS records are case-independent. Exim always looks them up
6706 in lower case. Some resolvers return domain names in exactly the case they
6707 appear in the zone file, that is, they may contain uppercase letters. Not
6708 all resolvers do this - some return always lower case. Exim was treating a
6709 change of case by a resolver as a change of domain, similar to a widening
6710 of a domain abbreviation. This triggered its re-routing code and so it was
6711 trying to route what was effectively the same domain again. This normally
6712 caused routing to fail (because the router wouldn't handle the domain
6713 twice). Now Exim checks for this case specially, and just changes the
6714 casing of the domain that it ultimately uses when it transmits the message
6717 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6720 50. If a filter generated a file delivery with a non-absolute name (possible if
6721 no home directory exists for the router), the forbid_file option was not
6724 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6725 the existing equality matching.
6727 52. Exim was using ints instead of ino_t variables in some places where it was
6728 dealing with inode numbers.
6730 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6731 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6732 called TMPDIR, and if it finds it is different, it changes its value.
6734 54. The smtp_printf() function is now made available to local_scan() so
6735 additional output lines can be written before returning. There is also an
6736 smtp_fflush() function to enable the detection of a dropped connection.
6737 The variables smtp_input and smtp_batched_input are exported to
6740 55. Changed the default runtime configuration: the message "Unknown user"
6741 has been removed from the ACL, and instead placed on the localuser router,
6742 using the cannot_route_message feature. This means that any verification
6743 failures that generate their own messages won't get overridden. Similarly,
6744 the "Unrouteable address" message that was in the ACL for unverifiable
6745 relay addresses has also been removed.
6747 56. Added hosts_avoid_esmtp to the smtp transport.
6749 57. The exicyclog script was not checking for the esoteric option
6750 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6751 will work only if exicyclog is run under the appropriate euid.
6753 58. Following a discussion on the list, the rules by which Exim recognises line
6754 endings on incoming messages have been changed. The -dropcr and drop_cr
6755 options are now no-ops, retained only for backwards compatibility. The
6756 following line terminators are recognized: LF CRLF CR. However, special
6757 processing applies to CR:
6759 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6760 nor a local message in the state where . is a terminator.
6762 (ii) If a bare CR is encountered in a header line, an extra space is added
6763 after the line terminator so as not to end the header. The reasoning
6764 behind this is that bare CRs in header lines are most likely either
6765 to be mistakes, or people trying to play silly games.
6767 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6768 was being incorrectly given as 18 bytes larger than it should have been.
6769 This is a VOB (very old bug).
6771 60. This may never have affected anything current, but just in case it has:
6772 When the local host is found other than at the start of a list of hosts,
6773 the local host, those with the same MX, and any that follow, are discarded.
6774 When the list in question was part of a longer list of hosts, the following
6775 hosts (not currently being processed) were also being discarded. This no
6776 longer happens. I'm not sure if this situation could ever has previously
6779 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6782 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6783 header. This is something that is recommended in a new Internet Draft, and
6784 is something that is documented as being done by Sendmail. There are two
6785 possible values. For messages generated by the autoreply transport, Exim
6788 Auto-Submitted: auto-replied
6790 whereas for all other generated messages (e.g. bounces) it adds
6792 Auto-Submitted: auto-generated
6794 63. The "personal" condition in filters now includes a test for the
6795 Auto-Submitted: header. If it contains the string "auto-" the message it
6796 not considered personal.
6798 64. Added rcpt_include_affixes as a generic transport option.
6800 65. Added queue_only_override (default true).
6802 66. Added the syslog_duplication option.
6804 67. If what should have been the first header line of a message consisted of
6805 a space followed by a colon, Exim was mis-interpreting it as a header line.
6806 It isn't of course - it is syntactically invalid and should therefore be
6807 treated as the start of the message body. The misbehaviour could have
6808 caused a number of strange effects, including loss of data in subsequent
6809 header lines, and spool format errors.
6811 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6812 client host had authenticated. This control can now be exercised by an ACL
6813 for more flexibility.
6815 69. By default, callouts do not happen when testing with -bh. There is now a
6816 variant, -bhc, which does actually run the callout code, including
6817 consulting and updating the callout cache.
6819 70. Added support for saslauthd authentication, courtesy of Alexander
6822 71. If statvfs() failed on the spool or log directories while checking their
6823 size for availability, Exim confusingly gave the error "space shortage".
6824 Furthermore, in debugging mode it crashed with a floating point exception.
6825 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6826 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6827 serious problem, Exim now writes to the main and panic logs when this
6828 happens, with details of the failure. It then refuses to accept the
6829 incoming message, giving the message "spool directory problem" or "log
6830 directory problem" with a 421 code for SMTP messages.
6832 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6833 0, 1, and 2 exist, because some OS complain for execs without them (see
6834 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6835 descriptors. However, the code omitted to check that the open succeeded,
6836 causing mysterious errors if for some reason the permissions on /dev/null
6837 got screwed. Now Exim writes a message to the main and panic logs, and
6838 bombs out if it can't open /dev/null.
6840 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6841 interact so that it is all more flexible. It is supposed to remain
6842 backwards compatible. Also added extra_local_interfaces.
6844 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6845 to bomb out with an assertion failure - to the client this appears as a
6846 connection drop. This problem occurs in the part of the code that was taken
6847 from the Samba project. Fortunately, the assertion is in a very simple
6848 function, so I have fixed this by reproducing the function inline in the
6849 one place where it is called, and arranging for authentication to fail
6850 instead of killing the process with assert().
6852 75. The SPA client code was not working when the server requested OEM rather
6853 than Unicode encoding.
6855 76. Added code to make require_files with a specific uid setting more usable in
6856 the case where statting the file as root fails - usually a non-root-mounted
6857 NFS file system. When this happens and the failure is EACCES, Exim now
6858 forks a subprocess and does the per-uid checking as the relevant uid.
6860 77. Added process_log_path.
6862 78. If log_file_path was not explicitly set, a setting of check_log_space or
6863 check_log_inodes was ignored.
6865 79. If a space check for the spool or log partitions fails, the incident is now
6866 logged. Of course, in the latter case the data may get lost...
6868 80. Added the %p formatting code to string_format() so that it can be used to
6869 print addresses in debug_print(). Adjusted all the address printing in the
6870 debugging in store.c to use %p rather than %d.
6872 81. There was a concern that a line of code in smtp_in.c could overflow a
6873 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6874 initially expressed, the concern was not well-founded, because trailing
6875 spaces are removed early. However, if the trailing spaces were followed by
6876 a NULL, they did not get removed, so the overflow was possible. Two fixes
6879 (a) I re-wrote the offending code in a cleaner fashion.
6880 (b) If an incoming SMTP command contains a NULL character, it is rejected
6883 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6884 runs initgroups(), so that if the Exim user is in any additional groups,
6885 they will be used during message reception.
6891 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob