1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
40 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
41 Disable the session-cache too, which might reduce our load. Since we
42 currrectly use a new context for every connection, both as server and
43 client, there is no benefit for these.
44 GnuTLS appears to not support tickets server-side by default (we don't
45 call gnutls_session_ticket_enable_server()) but client side is enabled
46 by default on recent versions (3.1.3 +) unless the PFS priority string
49 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
50 <https://reproducible-builds.org/specs/source-date-epoch/>.
52 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
53 the check for any unsuccessful recipients did not notice the limit, and
54 erroneously found still-pending ones.
56 JH/08 Pipeline CHUNKING command and data together, on kernels that support
57 MSG_MORE. Only in-clear (not on TLS connections).
59 JH/09 Avoid using a temporary file during transport using dkim. Unless a
60 transport-filter is involved we can buffer the headers in memory for
61 creating the signature, and read the spool data file once for the
62 signature and again for transmission.
64 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
65 4.77 as the kernel support then wasn't solid, having issues in 64bit
66 mode. Now, it's been long enough. Add support for FreeBSD also.
68 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
69 case where the routing stage had gathered several addresses to send to
70 a host before calling the transport for the first, we previously failed
71 to close down TLS in the old transport process before passing the TCP
72 connection to the new process. The new one sent a STARTTLS command
73 which naturally failed, giving a failed delivery and bloating the retry
74 database. Investigation and fix prototype from Wolfgang Breyha.
76 JH/12 Fix check on SMTP command input synchronisation. Previously there were
77 false-negatives in the check that the sender had not preempted a response
78 or prompt from Exim (running as a server), due to that code's lack of
79 awareness of the SMTP input buffering.
81 PP/04 Add commandline_checks_require_admin option.
82 Exim drops privileges sanely, various checks such as -be aren't a
83 security problem, as long as you trust local users with access to their
84 own account. When invoked by services which pass untrusted data to
85 Exim, this might be an issue. Set this option in main configuration
86 AND make fixes to the calling application, such as using `--` to stop
89 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
90 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
91 responses to those, into a single TLS record each way (this usually means
92 a single packet). As a side issue, smtp_enforce_sync now works on TLS
95 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
96 affects you only if you're dancing at the edge of the param size limits.
97 If you are, and this message makes sense to you, then: raise the
98 configured limit or use OpenSSL 1.1. Nothing we can do for older
101 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
102 cmdline element to get the documented default one. Previously it was
105 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
106 are unsuable for non-message-specific reasons - eg. connection timeout,
107 banner-time rejection.
109 JH/16 Fix logging of delivery remote port, when specified by router, under
112 PP/06 Repair manualroute's ability to take options in any order, even if one
113 is the name of a transport.
116 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
118 JH/17 Change the list-building routines interface to use the expanding-string
119 triplet model, for better allocation and copying behaviour.
121 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
122 Previously it was constructed the first time a possibly-matching string
123 was met in the configuration file input during startup; now it is done
126 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
127 compatible one, to avoid the (poorly documented) possibility of a config
128 file in the working directory redirecting the DB files, possibly correpting
131 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
132 cache-hot. Previously, although the result was properly cached, the
133 initial verify call returned a defer.
135 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
136 the main verify for receipient in uncached-mode.
138 JH/22 Retire historical build files to an "unsupported" subdir. These are
139 defined as "ones for which we have no current evidence of testing".
141 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
142 if present. Previously it was ignored.
148 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
149 than -2003 did; needs libidn2 in addition to libidn.
151 JH/02 The path option on a pipe transport is now expanded before use.
153 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
154 Patch provided by "Björn", documentation fix added too.
156 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
157 missing a wire-to-host endian conversion.
159 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
160 close after a BDAT command line could be taken as a following command,
161 giving a synch failure. Fix by only checking for synch immediately
162 before acknowledging the chunk.
164 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
165 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
166 macro. Patches provided by Josh Soref.
168 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
169 Previously we did not; the RFC seems ambiguous and VRFY is not listed
170 by IANA as a service extension. However, John Klensin suggests that we
173 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
174 the dkim code may be unix-mode line endings rather than smtp wire-format
175 CRLF, so prepend a CR to any bare LF.
177 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
178 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
180 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
181 the first were themselves being wrongly included in the feed into dkim
182 processing; with most chunk sizes in use this resulted in an incorrect
183 body hash calculated value.
185 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
186 DKIM signature block, for verification. Although advised against by
187 standards it is specifically not ruled illegal.
189 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
191 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
192 missing a body hash (the bh= tag).
194 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
195 It seems that HAProxy sends the Proxy Protocol information in clear and
196 only then does a TLS startup, so do the same.
198 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
199 TCP connections (such as for Spamd) unless the daemon successfully set
200 Fast Open mode on its listening sockets. This fixes breakage seen on
201 too-old kernels or those not configured for Fast Open, at the cost of
202 requiring both directions being enabled for TFO, and TFO never being used
203 by non-daemon-related Exim processes.
205 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
206 endings, at least on the first header line. Try to canonify any that get
207 past that check, despite the cost.
209 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
210 now limited to an arbitrary five deep, while parsing addresses with the
211 strip_excess_angle_brackets option enabled.
213 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
214 instead leave the unprompted TLS handshake in socket buffer for the
215 TLS library to consume.
217 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
219 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
221 JH/16 Drop variables when they go out of scope. Memory management drops a whole
222 region in one operation, for speed, and this leaves assigned pointers
223 dangling. Add checks run only under the testsuite which checks all
224 variables at a store-reset and panics on a dangling pointer; add code
225 explicitly nulling out all the variables discovered. Fixes one known
226 bug: a transport crash, where a dangling pointer for $sending_ip_address
227 originally assigned in a verify callout, is re-used.
229 PP/06 Drop '.' from @INC in various Perl scripts.
231 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
233 PP/08 Reduce a number of compilation warnings under clang; building with
234 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
235 should be warning-free.
237 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
239 HS/01 Fix portability problems introduced by PP/08 for platforms where
240 realloc(NULL) is not equivalent to malloc() [SunOS et al].
242 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
243 chunk. This allows us to accept broken chunked messages. We need a more
244 general solution here.
246 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
247 already-broken messages in the queue.
249 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
251 JH/19 Fix reference counting bug in routing-generated-address tracking.
257 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
258 supports it and a size is available (ie. the sending peer gave us one).
260 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
261 years of being deprecated). The replacements are the ACLs
262 acl_smtp_mime and acl_not_smtp_mime.
264 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
265 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
266 or even an in-clear connection were permitted. Now, if the host lookup
267 was dnssec and dane was requested then the host is only used if the
268 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
269 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
270 if one fails this test.
271 This means that a poorly-configured remote DNS will make it incommunicado;
272 but it protects against a DNS-interception attack on it.
274 JH/04 Bug 1810: make continued-use of an open smtp transport connection
275 non-noisy when a race steals the message being considered.
277 JH/05 If main configuration option tls_certificate is unset, generate a
278 self-signed certificate for inbound TLS connections.
280 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
281 in rewrites and routers.
283 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
284 and logged a warning sing 4.83; now they are a configuration file error.
286 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
287 (lacking @domain). Apply the same qualification processing as RCPT.
289 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
291 JH/10 Support ${sha256:} applied to a string (as well as the previous
294 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
295 a cutthrough deliver is pending, as we always want to make a connection.
296 This also avoids re-routing the message when later placing the cutthrough
297 connection after a verify cache hit.
298 Do not update it with the verify result either.
300 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
301 when routing results in more than one destination address.
303 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
304 signing (which inhibits the cutthrough capability). Previously only
305 the presence of an option was tested; now an expansion evaluating as
306 empty is permissible (obviously it should depend only on data available
307 when the cutthrough connection is made).
309 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
310 the relevant preceding SMTP command did not note the pipelining mode.
312 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
313 Previously they were not counted.
315 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
316 as one having no matching records. Previously we deferred the message
317 that needed the lookup.
319 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
320 distinguished as "(=".
322 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
323 for missing MX records. Previously it only worked for missing A records.
325 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
327 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
328 after the data-go-ahead and data-ack. Patch from Jason Betts.
330 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
331 even for a "none" policy. Patch from Tony Meyer.
333 JH/22 Fix continued use of a connection for further deliveries. If a port was
334 specified by a router, it must also match for the delivery to be
337 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
338 When one of the recipients of a message was unsuitable for the connection
339 (has no matching addresses), we lost track of needing to mark it
340 deferred. As a result mail would be lost.
342 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
344 JH/25 Decoding ACL controls is now done using a binary search; the source code
345 takes up less space and should be simpler to maintain. Merge the ACL
346 condition decode tables also, with similar effect.
348 JH/26 Fix problem with one_time used on a redirect router which returned the
349 parent address unchanged. A retry would see the parent address marked as
350 delivered, so not attempt the (identical) child. As a result mail would
353 JH/27 Fix a possible security hole, wherein a process operating with the Exim
354 UID can gain a root shell. Credit to http://www.halfdog.net/ for
355 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
358 JH/28 Enable {spool,log} filesystem space and inode checks as default.
359 Main config options check_{log,spool}_{inodes,space} are now
360 100 inodes, 10MB unless set otherwise in the configuration.
362 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
363 Previously it only applied to the main-section connection policy
366 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
368 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
369 by me. Added RFC7919 DH primes as an alternative.
371 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
372 are not in the system include path.
374 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
375 GnuTLS, when a session startup failed (eg because the client disconnected)
376 Exim did stdio operations after fclose. This was exposed by a recent
377 change which nulled out the file handle after the fclose.
379 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
380 signed directly by the cert-signing cert, rather than an intermediate
381 OCSP-signing cert. This is the model used by LetsEncrypt.
383 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
385 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
386 an incoming connection.
388 HS/02 Bug 1802: Do not half-close the connection after sending a request
391 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
392 fallback to "prime256v1".
394 JH/34 SECURITY: Use proper copy of DATA command in error message.
395 Could leak key material. Remotely exploitable. CVE-2016-9963.
401 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
402 and 3.4.4 - once the server is enabled to respond to an OCSP request
403 it does even when not requested, resulting in a stapling non-aware
404 client dropping the TLS connection.
406 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
407 support variable-length bit vectors. No functional change.
409 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
410 The I= interface field on outgoing lines is now after the H= remote
411 host field, same as incoming lines. There is a separate
412 outgoing_interface log selector which allows you to disable the
415 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
416 If not running log_selector +smtp_connection the mainlog would be held
417 open indefinitely after a "too many connections" event, including to a
418 deleted file after a log rotate. Leave the per net connection logging
419 leaving it open for efficiency as that will be quickly detected by the
420 check on the next write.
422 HS/01 Bug 1671: Fix post transport crash.
423 Processing the wait-<transport> messages could crash the delivery
424 process if the message IDs didn't exist for some reason. When
425 using 'split_spool_directory=yes' the construction of the spool
426 file name failed already, exposing the same netto behaviour.
428 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
429 mime_regex ACL conditions.
431 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
432 to DSN fail messages (bounces): remote IP, remote greeting, remote response
433 to HELO, local diagnostic string.
435 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
436 log line to debug. Even when configured with a tls authenticator many
437 client connections are expected to not authenticate in this way, so
438 an authenticate fail is not an error.
440 HS/02 Add the Exim version string to the process info. This way exiwhat
441 gives some more detail about the running daemon.
443 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
444 matter for fast-change records such as DNSBLs.
446 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
447 retry record, even if constant. There may be multiple transports with
448 different interface settings and the retry behaviour needs to be kept
451 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
453 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
455 JH/10 Bug 840: fix log_defer_output option of pipe transport
457 JH/11 Bug 830: use same host for all RCPTS of a message, even under
458 hosts_randomize. This matters a lot when combined with mua_wrapper.
460 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
461 ${quote_pgsql:<string>} operator.
463 JH/13 Bug 1708: avoid misaligned access in cached lookup.
465 JH/14 Change header file name for freeradius-client. Relevant if compiling
466 with Radius support; from the Gentoo tree and checked under Fedora.
468 JH/15 Bug 1712: Introduce $prdr_requested flag variable
470 JH/16 Bug 1714: Permit an empty string as expansion result for transport
471 option transport_filter, meaning no filtering.
473 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
475 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
476 defaults to "*" (all hosts). The variable is now available when not built
477 with TLS, default unset, mainly to enable keeping the testsuite sane.
478 If a server certificate is not supplied (via tls_certificate) an error is
479 logged, and clients will find TLS connections fail on startup. Presumably
480 they will retry in-clear.
481 Packagers of Exim are strongly encouraged to create a server certificate
482 at installation time.
484 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
485 with the $config_file variable.
487 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
488 in transport context, after the attempt, and per-recipient. The latter type
489 is per host attempted. The event data is the error message, and the errno
490 information encodes the lookup type (A vs. MX) used for the (first) host,
491 and the trailing two digits of the smtp 4xx response.
493 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
494 to write to mainlog (or rejectlog, paniclog) in the window between file
495 creation and permissions/ownership being changed. Particularly affects
496 installations where exicyclog is run as root, rather than exim user;
497 result is that the running daemon panics and dies.
499 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
501 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
502 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
503 "pri" and "weight". Note that the previous implicit priority given by the
504 list order is no longer honoured.
506 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
509 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
510 by defining SUPPORT_SOCKS.
512 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
513 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
514 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
515 variables are renamed to proxy_{local,external}_{address,port}.
517 JH/27 Move Internationalisation support from Experimental to mainline, enabled
518 for a build by defining SUPPORT_I18N
520 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
521 of the query string, and make ${quote_redis:} do that quoting.
523 JH/29 Move Events support from Experimental to mainline, enabled by default
524 and removable for a build by defining DISABLE_EVENT.
526 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
528 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
529 cached by the daemon.
531 JH/32 Move Redis support from Experimental to mainline, enabled for a build
532 by defining LOOKUP_REDIS. The libhiredis library is required.
534 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
535 keys are given for lookup.
537 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
538 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
539 only supported when built with TLS support. The PolarSSL SHA routines
540 are still used when the TLS library is too old for convenient support.
542 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
543 openssl_options), for security. OpenSSL forces this from version 1.1.0
544 server-side so match that on older versions.
546 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
547 allocation for $value could be released as the expansion processing
548 concluded, but leaving the global pointer active for it.
550 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
551 and to use the domains and local_parts ACL conditions.
553 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
554 incorrectly not doubled on cutthrough transmission, hence seen as a
555 body-termination at the receiving system - resulting in truncated mails.
556 Commonly the sender saw a TCP-level error, and retransmitted the message
557 via the normal store-and-forward channel. This could result in duplicates
558 received - but deduplicating mailstores were liable to retain only the
559 initial truncated version.
561 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
563 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
565 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
566 we're in there, support oversigning also; bug 1309.
568 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
570 HS/04 Add support for keep_environment and add_environment options.
572 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
573 either intentional arithmetic overflow during PRNG, or testing config-
576 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
577 delivery resulted in actual delivery. Cancel cutthrough before DATA
580 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
581 rejects a recipient: pass the reject to the originator.
583 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
584 Many were false-positives and ignorable, but it's worth fixing the
587 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
588 for the new environment-manipulation done at startup. Move the routines
589 from being local to tls.c to being global via the os.c file.
591 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
592 an extract embedded as result-arg for a map, the first arg for extract
593 is unavailable so we cannot tell if this is a numbered or keyed
594 extraction. Accept either.
600 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
603 JH/02 The smtp transport option "multi_domain" is now expanded.
605 JH/03 The smtp transport now requests PRDR by default, if the server offers
608 JH/04 Certificate name checking on server certificates, when exim is a client,
609 is now done by default. The transport option tls_verify_cert_hostnames
610 can be used to disable this per-host. The build option
611 EXPERIMENTAL_CERTNAMES is withdrawn.
613 JH/05 The value of the tls_verify_certificates smtp transport and main options
614 default to the word "system" to access the system default CA bundle.
615 For GnuTLS, only version 3.0.20 or later.
617 JH/06 Verification of the server certificate for a TLS connection is now tried
618 (but not required) by default. The verification status is now logged by
619 default, for both outbound TLS and client-certificate supplying inbound
622 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
625 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
626 Status Notification (bounce) messages are now MIME format per RFC 3464.
627 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
628 under the control of the dsn_advertise_hosts option, and routers may
629 have a dsn_lasthop option.
631 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
632 default, modifiable by a malware= option. The list separator for
633 the options can now be changed in the usual way. Bug 68.
635 JH/10 The smtp_receive_timeout main option is now expanded before use.
637 JH/11 The incoming_interface log option now also enables logging of the
638 local interface on delivery outgoing connections.
640 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
641 if the interface and destination host and port all match.
643 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
646 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
647 Patch from Andrew Lewis.
649 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
650 now supports optional time-restrictions, weighting, and priority
651 modifiers per server. Patch originally by <rommer@active.by>.
653 JH/16 The spamd_address main option now supports a mixed list of local
654 and remote servers. Remote servers can be IPv6 addresses, and
655 specify a port-range.
657 JH/17 Bug 68: The spamd_address main option now supports an optional
658 timeout value per server.
660 JH/18 Bug 1581: Router and transport options headers_add/remove can
661 now have the list separator specified.
663 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
666 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
669 JH/21 Support for the A6 type of dns record is withdrawn.
671 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
672 rather than the verbs used.
674 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
675 from 255 to 1024 chars.
677 JH/24 Verification callouts now attempt to use TLS by default.
679 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
680 are generic router options now. The defaults didn't change.
682 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
683 Original patch from Alexander Shikoff, worked over by JH.
685 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
686 files if ZCAT_COMMAND is not executable.
688 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
690 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
692 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
693 Normally benign, it bites when the pair was led to by a CNAME;
694 modern usage is to not canonicalize the domain to a CNAME target
695 (and we were inconsistent anyway for A-only vs AAAA+A).
697 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
699 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
700 when evaluating $sender_host_dnssec.
702 JH/31 Check the HELO verification lookup for DNSSEC, adding new
703 $sender_helo_dnssec variable.
705 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
707 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
709 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
711 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
712 documented as working, but never had. Support all but $spam_report.
714 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
715 added for tls authenticator.
717 HS/03 Add perl_taintmode main config option
723 TL/01 When running the test suite, the README says that variables such as
724 no_msglog_check are global and can be placed anywhere in a specific
725 test's script, however it was observed that placement needed to be near
726 the beginning for it to behave that way. Changed the runtest perl
727 script to read through the entire script once to detect and set these
728 variables, reset to the beginning of the script, and then run through
729 the script parsing/test process like normal.
731 TL/02 The BSD's have an arc4random API. One of the functions to induce
732 adding randomness was arc4random_stir(), but it has been removed in
733 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
734 function when detected.
736 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
737 cause callback expansion.
739 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
740 syntax errors in an expansion can be treated as a string instead of
741 logging or causing an error, due to the internal use of bool_lax
742 instead of bool when processing it.
744 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
745 server certificates when making smtp deliveries.
747 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
749 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
751 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
753 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
754 Merged patch from Sebastian Wiedenroth.
756 JH/05 Fix results-pipe from transport process. Several recipients, combined
757 with certificate use, exposed issues where response data items split
758 over buffer boundaries were not parsed properly. This eventually
759 resulted in duplicates being sent. This issue only became common enough
760 to notice due to the introduction of connection certificate information,
761 the item size being so much larger. Found and fixed by Wolfgang Breyha.
763 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
764 size buffer was used, resulting in syntax errors when an expansion
767 JH/07 Add support for directories of certificates when compiled with a GnuTLS
768 version 3.3.6 or later.
770 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
771 is EXPERIMENTAL_EVENT, the main-configuration and transport options
772 both become "event_action", the variables become $event_name, $event_data
773 and $event_defer_errno. There is a new variable $verify_mode, usable in
774 routers, transports and related events. The tls:cert event is now also
775 raised for inbound connections, if the main configuration event_action
778 TL/06 In test suite, disable OCSP for old versions of openssl which contained
779 early OCSP support, but no stapling (appears to be less than 1.0.0).
781 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
782 server certificate names available under the smtp transport option
783 "tls_verify_cert_hostname" now do not permit multi-component wildcard
786 JH/10 Time-related extraction expansions from certificates now use the main
787 option "timezone" setting for output formatting, and are consistent
788 between OpenSSL and GnuTLS compilations. Bug 1541.
790 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
791 encoded parameter in the incoming message. Bug 1558.
793 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
794 include certificate info, eximon was claiming there were spoolfile
797 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
799 JH/14 Log delivery-related information more consistently, using the sequence
800 "H=<name> [<ip>]" wherever possible.
802 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
803 are problematic for Debian distribution, omit them from the release
806 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
808 JH/16 Fix string representation of time values on 64bit time_t architectures.
811 JH/17 Fix a null-indirection in certextract expansions when a nondefault
812 output list separator was used.
817 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
818 checkers that were complaining about end of non-void function with no
821 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
822 This was a regression introduced in 4.83 by another bugfix.
824 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
826 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
827 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
833 TF/01 Correctly close the server side of TLS when forking for delivery.
835 When a message was received over SMTP with TLS, Exim failed to clear up
836 the incoming connection properly after forking off the child process to
837 deliver the message. In some situations the subsequent outgoing
838 delivery connection happened to have the same fd number as the incoming
839 connection previously had. Exim would try to use TLS and fail, logging
840 a "Bad file descriptor" error.
842 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
843 utilities have not been installed.
845 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
846 temporary space as the ACL may create new global variables.
848 TL/01 LDAP support uses per connection or global context settings, depending
849 upon the detected version of the libraries at build time.
851 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
852 to extract and use the src ip:port in logging and expansions as if it
853 were a direct connection from the outside internet. PPv2 support was
854 updated based on HAProxy spec change in May 2014.
856 JH/02 Add ${listextract {number}{list}{success}{fail}}.
858 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
859 Properly escape header and check for NULL return.
861 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
864 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
866 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
867 characters in header names, implemented as a verify condition.
868 Contributed by Michael Fischer v. Mollard.
870 TL/05 Rename SPF condition results err_perm and err_temp to standardized
871 results permerror and temperror. Previous values are deprecated but
872 still accepted. In a future release, err_perm and err_temp will be
873 completely removed, which will be a backward incompatibility if the
874 ACL tests for either of these two old results. Patch contributed by
875 user bes-internal on the mailing list.
877 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
879 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
880 selectors, in both main and reject logs.
882 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
885 JH/07 Add malware type "sock" for talking to simple daemon.
887 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
889 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
890 routers/transports under cutthrough routing.
892 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
893 numbers. Touch up "bool" conditional to keep the same definition.
895 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
897 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
899 JH/12 Expand items in router/transport headers_add or headers_remove lists
900 individually rather than the list as a whole. Bug 1452.
902 Required for reasonable handling of multiple headers_ options when
903 they may be empty; requires that headers_remove items with embedded
904 colons must have them doubled (or the list-separator changed).
906 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
907 view the policy declared in the DMARC record. Currently, $dmarc_status
908 is a combined value of both the record presence and the result of the
911 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
913 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
914 dnslookup router and the smtp transport (applying to the forward
917 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
918 of ldap servers used for a specific lookup. Patch provided by Heiko
921 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
922 New variable $lookup_dnssec_authenticated for observability.
924 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
925 Patch submitted by Lars Timman.
927 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
929 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
930 Requires trusted mode and valid format message id, aborts otherwise.
931 Patch contributed by Heiko Schlichting.
933 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
934 certextract with support for various fields. Bug 1358.
936 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
937 is requested by default, modifiable by smtp transport option
940 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
941 operate on certificate variables to give certificate fingerprints
942 Also new ${sha256:cert_variable}.
944 JH/23 The PRDR feature is moved from being Experimental into the mainline.
946 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
947 Christian Aistleitner.
949 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
951 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
952 file. Patch from Wolfgang Breyha.
954 JH/25 Expand the coverage of the delivery $host and $host_address to
955 client authenticators run in verify callout. Bug 1476.
957 JH/26 Port service names are now accepted for tls_on_connect_ports, to
958 align with daemon_smtp_ports. Bug 72.
960 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
961 support and error reporting did not work properly.
963 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
964 and is readable. Patch from Andrew Colin Kissa.
966 TL/14 Enhance documentation of ${run expansion and how it parses the
967 commandline after expansion, particularly in the case when an
968 unquoted variable expansion results in an empty value.
970 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
972 PP/02 Fix internal collision of T_APL on systems which support RFC3123
973 by renaming away from it. Addresses GH issue 15, reported by
976 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
978 TL/15 SECURITY: prevent double expansion in math comparison functions
979 (can expand unsanitized data). Not remotely exploitable.
986 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
988 PP/02 Make -n do something, by making it not do something.
989 When combined with -bP, the name of an option is not output.
991 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
994 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
995 $sender_host_name and config options to manage this, and basic check
998 PP/05 DSCP support for outbound connections and control modifier for inbound.
1000 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1001 (Only plugin which currently uses this is kerberos4, which nobody should
1002 be using, but we should make it available and other future plugins might
1003 conceivably use it, even though it would break NAT; stuff *should* be
1004 using channel bindings instead).
1006 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1007 name; added for Sendmail compatibility; requires admin caller.
1008 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1009 just ignore it); requires trusted caller.
1010 Also parse but ignore: -Ac -Am -X<logfile>
1013 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1015 TL/02 Add +smtp_confirmation as a default logging option.
1017 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1018 Patch by Magnus Holmgren from 2007-02-20.
1020 TL/04 Bugzilla 1281 - Spec typo.
1021 Bugzilla 1283 - Spec typo.
1022 Bugzilla 1290 - Spec grammar fixes.
1024 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1026 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1028 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1029 mailing list by by Dmitry Isaikin.
1031 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1033 JH/02 Support "G" suffix to numbers in ${if comparisons.
1035 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1037 NM/01 Bugzilla 1197 - Spec typo
1038 Bugzilla 1196 - Spec examples corrections
1040 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1042 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1043 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1046 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1047 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1049 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1050 "acl {{name}{arg}...}", and optional args on acl condition
1053 JH/05 Permit multiple router/transport headers_add/remove lines.
1055 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1057 JH/07 Avoid using a waiting database for a single-message-only transport.
1058 Performance patch from Paul Fisher. Bugzilla 1262.
1060 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1063 JH/09 Add $headers_added variable, with content from use of ACL modifier
1064 add_header (but not yet added to the message). Bugzilla 199.
1066 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1067 Pulled from Bugzilla 817 by Wolfgang Breyha.
1069 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1071 (nb: this is the same fix as in Exim 4.80.1)
1073 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1076 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1079 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1081 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1082 advertise SMTP AUTH mechanism to us, instead of a generic
1083 protocol violation error. Also, make Exim more robust to bad
1084 data from the Dovecot auth socket.
1086 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1088 When a queue runner is handling a message, Exim first routes the
1089 recipient addresses, during which it prunes them based on the retry
1090 hints database. After that it attempts to deliver the message to
1091 any remaining recipients. It then updates the hints database using
1094 So if a recipient address works intermittently, it can get repeatedly
1095 deferred at routing time. The retry hints record remains fresh so the
1096 address never reaches the final cutoff time.
1098 This is a fairly common occurrence when a user is bumping up against
1099 their storage quota. Exim had some logic in its local delivery code
1100 to deal with this. However it did not apply to per-recipient defers
1101 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1103 This change adds a proper retry rule check during routing so that the
1104 final cutoff time is checked against the message's age. We only do
1105 this check if there is an address retry record and there is not a
1106 domain retry record; this implies that previous attempts to handle
1107 the address had the retry_use_local_parts option turned on. We use
1108 this as an approximation for the destination being like a local
1109 delivery, as in LMTP.
1111 I suspect this new check makes the old local delivery cutoff check
1112 redundant, but I have not verified this so I left the code in place.
1114 TF/02 Correct gecos expansion when From: is a prefix of the username.
1116 Test 0254 submits a message to Exim with the header
1120 When I ran the test suite under the user fanf2, Exim expanded
1121 the header to contain my full name, whereas it should have added
1122 a Resent-Sender: header. It erroneously treats any prefix of the
1123 username as equal to the username.
1125 This change corrects that bug.
1127 GF/01 DCC debug and logging tidyup
1128 Error conditions log to paniclog rather than rejectlog.
1129 Debug lines prefixed by "DCC: " to remove any ambiguity.
1131 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1133 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1134 Bug spotted by Jeremy Harris; was flawed since initial commit.
1135 Would have resulted in OCSP responses post-SNI triggering an Exim
1136 NULL dereference and crash.
1138 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1140 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1141 Bug detection, analysis and fix by Samuel Thibault.
1142 Bugzilla 1331, Debian bug #698092.
1144 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1146 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1147 Server implementation by Todd Lyons, client by JH.
1148 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1149 config variable "prdr_enable" controls whether the server
1150 advertises the facility. If the client requests PRDR a new
1151 acl_data_smtp_prdr ACL is called once for each recipient, after
1152 the body content is received and before the acl_smtp_data ACL.
1153 The client is controlled by both of: a hosts_try_prdr option
1154 on the smtp transport, and the server advertisement.
1155 Default client logging of deliveries and rejections involving
1156 PRDR are flagged with the string "PRDR".
1158 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1159 fclose(). Diagnosis by Todd Lyons.
1161 PP/17 Update configure.default to handle IPv6 localhost better.
1162 Patch by Alain Williams (plus minor tweaks).
1165 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1166 This is now consistent with GnuTLS, and is now documented: the
1167 previous undocumented portable approach to treating the option as
1168 unset was to force an expansion failure. That still works, and
1169 an empty string is now equivalent.
1171 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1172 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1173 not performing validation itself.
1175 PP/20 Added force_command boolean option to pipe transport.
1176 Patch from Nick Koston, of cPanel Inc.
1178 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1181 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1183 PP/21 Fix eximon continuous updating with timestamped log-files.
1184 Broken in a format-string cleanup in 4.80, missed when I repaired the
1185 other false fix of the same issue.
1186 Report and fix from Heiko Schlichting.
1189 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1190 Report from Prashanth Katuri.
1192 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1193 It's SecureTransport, so affects any MacOS clients which use the
1194 system-integrated TLS libraries, including email clients.
1196 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1197 using a MIME ACL for non-SMTP local injection.
1198 Report and assistance in diagnosis by Warren Baker.
1200 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1202 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1204 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1205 last id that failed so it may be referenced in subsequent ACL's.
1207 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1210 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1211 ldap library initialization, allowing self-signed CA's to be
1212 used. Also properly sets require_cert option later in code by
1213 using NULL (global ldap config) instead of ldap handle (per
1214 session). Bug diagnosis and testing by alxgomz.
1216 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1217 the src/util/ subdirectory.
1219 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1220 renamed to Transport Post Delivery Action by Jeremy Harris, as
1223 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1224 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1225 redis_servers = needs to be configured which will be used by the redis
1226 lookup. Patch from Warren Baker, of The Packet Hub.
1228 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1230 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1231 hostname or reverse DNS when processing a host list. Used suggestions
1232 from multiple comments on this bug.
1234 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1236 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1237 Missed a few lines, added it to make the runtest require no keyboard
1240 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1241 contains upper case chars. Make router use caseful_local_part.
1243 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1244 support when GnuTLS has been built with p11-kit.
1250 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1252 This, or similar/improved, will also be change PP/11 of 4.82.
1258 PP/01 Handle short writes when writing local log-files.
1259 In practice, only affects FreeBSD (8 onwards).
1260 Bugzilla 1053, with thanks to Dmitry Isaikin.
1262 NM/01 Bugzilla 949 - Documentation tweak
1264 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1267 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1269 PP/02 Implemented gsasl authenticator.
1271 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1273 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1274 `pkg-config foo` for cflags/libs.
1276 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1277 with rest of GSASL and with heimdal_gssapi.
1279 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1280 `pkg-config foo` for cflags/libs for the TLS implementation.
1282 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1283 properties get this fed in as external SSF. A number of robustness
1284 and debugging improvements to the cyrus_sasl authenticator.
1286 PP/08 cyrus_sasl server now expands the server_realm option.
1288 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1289 Patch by Jeremy Harris.
1291 PP/10 Added dbmjz lookup type.
1293 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1295 PP/12 MAIL args handles TAB as well as SP, for better interop with
1296 non-compliant senders.
1297 Analysis and variant patch by Todd Lyons.
1299 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1300 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1301 Patch from Dirk Mueller <dmueller@suse.com>
1303 PP/13 tls_peerdn now print-escaped for spool files.
1304 Observed some $tls_peerdn in wild which contained \n, which resulted
1305 in spool file corruption.
1307 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1308 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1309 or write after TLS renegotiation, which otherwise led to messages
1312 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1313 as a tracking header (ie: a signed header comes before the signature).
1314 Patch from Wolfgang Breyha.
1316 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1317 comma-sep list; embedded commas doubled.
1319 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1321 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1323 Report and patch from Dmitry Banschikov.
1325 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1326 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1327 failures. We appear to now support TLS1.1+ with Exim.
1329 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1330 lets Exim select keys and certificates based upon TLS SNI from client.
1331 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1332 before an outbound SMTP session. New log_selector, +tls_sni.
1334 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1335 NULL dereference. Report and patch from Alun Jones.
1337 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1338 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1339 Not seeing resolver debug output on NetBSD, but suspect this is a
1340 resolver implementation change.
1342 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1343 Left warnings. Added "eximon gdb" invocation mode.
1345 PP/21 Defaulting "accept_8bitmime" to true, not false.
1347 PP/22 Added -bw for inetd wait mode support.
1349 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1350 locate the relevant includes and libraries. Made this the default.
1352 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1353 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1355 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1356 This may cause build issues on older platforms.
1358 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1359 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1360 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1361 Added SNI support via GnuTLS too.
1362 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1364 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1366 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1367 Applied second patch from Janne, implementing suggestion to default
1368 multiple-strings-in-record handling to match SPF spec.
1370 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1372 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1373 read-only, out of scope).
1374 Patch from Wolfgang Breyha, report from Stuart Northfield.
1376 PP/29 Fix three issues highlighted by clang analyser static analysis.
1377 Only crash-plausible issue would require the Cambridge-specific
1378 iplookup router and a misconfiguration.
1379 Report from Marcin Mirosław.
1381 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1383 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1384 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1385 As part of this, removing so much warning spew let me fix some minor
1386 real issues in debug logging.
1388 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1389 assignment on my part. Fixed.
1391 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1392 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1393 Janne Snabb (who went above and beyond: thank you).
1395 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1396 string otherwise requires a connection and a bunch more work and it's
1397 relatively easy to get wrong. Should also expose TLS library linkage
1400 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1401 64-bit ${eval} (JH/03).
1403 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1404 GNU libc to support some of the 64-bit stuff, should not lead to
1405 conflicts. Defined before os.h is pulled in, so if a given platform
1406 needs to override this, it can.
1408 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1409 protection layer was required, which is not implemented.
1410 Bugzilla 1254, patch from Wolfgang Breyha.
1412 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1413 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1414 tls_dhparam take prime identifiers. Also unbreak combination of
1415 OpenSSL+DH_params+TLSSNI.
1417 PP/39 Disable SSLv2 by default in OpenSSL support.
1423 PP/01 Solaris build fix for Oracle's LDAP libraries.
1424 Bugzilla 1109, patch from Stephen Usher.
1426 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1428 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1431 TF/02 Fix a couple more cases where we did not log the error message
1432 when unlink() failed. See also change 4.74-TF/03.
1434 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1435 lock up or crash if it happened to be inside a call to libc when it
1436 got a SIGUSR1 from exiwhat.
1438 The SIGUSR1 handler appends the current process status to the process
1439 log which is later printed by exiwhat. It used to use the general
1440 purpose logging code to do this, but several functions it calls are
1441 not safe for signals.
1443 The new output code in the SIGUSR1 handler is specific to the process
1444 log, and simple enough that it's easy to inspect for signal safety.
1445 Removing some special cases also simplifies the general logging code.
1446 Removing the spurious timestamps from the process log simplifies
1449 TF/04 Improved ratelimit ACL condition.
1451 The /noupdate option has been deprecated in favour of /readonly which
1452 has clearer semantics. The /leaky, /strict, and /readonly update modes
1453 are mutually exclusive. The update mode is no longer included in the
1454 database key; it just determines when the database is updated. (This
1455 means that when you upgrade Exim will forget old rate measurements.)
1457 Exim now checks that the per_* options are used with an update mode that
1458 makes sense for the current ACL. For example, when Exim is processing a
1459 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1460 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1461 must specify per_mail/readonly. If you omit the update mode it defaults to
1462 /leaky where that makes sense (as before) or /readonly where required.
1464 The /noupdate option is now undocumented but still supported for
1465 backwards compatibility. It is equivalent to /readonly except that in
1466 ACLs where /readonly is required you may specify /leaky/noupdate or
1467 /strict/noupdate which are treated the same as /readonly.
1469 A useful new feature is the /count= option. This is a generalization
1470 of the per_byte option, so that you can measure the throughput of other
1471 aggregate values. For example, the per_byte option is now equivalent
1472 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1474 The per_rcpt option has been generalized using the /count= mechanism
1475 (though it's more complicated than the per_byte equivalence). When it is
1476 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1477 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1478 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1479 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1480 Note that using per_rcpt with a non-readonly update mode in more than
1481 one ACL will cause the recipients to be double-counted. (The per_mail
1482 and per_byte options don't have this problem.)
1484 The handling of very low rates has changed slightly. If the computed rate
1485 is less than the event's count (usually one) then this event is the first
1486 after a long gap. In this case the rate is set to the same as this event's
1487 count, so that the first message of a spam run is counted properly.
1489 The major new feature is a mechanism for counting the rate of unique
1490 events. The new per_addr option counts the number of different
1491 recipients that someone has sent messages to in the last time period. It
1492 behaves like per_rcpt if all the recipient addresses are different, but
1493 duplicate recipient addresses do not increase the measured rate. Like
1494 the /count= option this is a general mechanism, so the per_addr option
1495 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1496 example, measure the rate that a client uses different sender addresses
1497 with the options per_mail/unique=$sender_address. There are further
1498 details in the main documentation.
1500 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1502 TF/06 Removed a few PCRE remnants.
1504 TF/07 Automatically extract Exim's version number from tags in the git
1505 repository when doing development or release builds.
1507 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1508 Bugzilla 879. Patch from Paul Fisher.
1510 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1511 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1514 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1516 PP/05 Variable $av_failed, true if the AV scanner deferred.
1517 Bugzilla 1078. Patch from John Horne.
1519 PP/06 Stop make process more reliably on build failure.
1520 Bugzilla 1087. Patch from Heiko Schlittermann.
1522 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1523 Bugzilla 1089. Patch from Heiko Schlittermann.
1525 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1526 Bugzilla 1131. Patch from Holger Weiß.
1528 PP/09 Handle IPv6 addresses with SPF.
1529 Bugzilla 860. Patch from Wolfgang Breyha.
1531 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1533 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1536 PP/11 match_* no longer expand right-hand-side by default.
1537 New compile-time build option, EXPAND_LISTMATCH_RHS.
1538 New expansion conditions, "inlist", "inlisti".
1540 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1542 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1544 PP/14 fix log_write() format string regression from TF/03.
1545 Bugzilla 1152. Patch from Dmitry Isaikin.
1551 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1553 PP/02 Harmonised TLS library version reporting; only show if debugging.
1554 Layout now matches that introduced for other libraries in 4.74 PP/03.
1556 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1558 PP/04 New "dns_use_edns0" global option.
1560 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1563 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1564 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1566 TK/01 Updated PolarSSL code to 0.14.2.
1567 Bugzilla 1097. Patch from Andreas Metzler.
1569 PP/07 Catch divide-by-zero in ${eval:...}.
1570 Fixes bugzilla 1102.
1572 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1575 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1576 format-string attack -- SECURITY: remote arbitrary code execution.
1578 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1579 time unintentionally subject to list matching rules, letting the header
1580 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1581 arbitrary string expansion). This allowed for information disclosure.
1583 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1584 INT_MIN/-1 -- value coerced to INT_MAX.
1590 NM/01 Workaround for PCRE version dependency in version reporting
1593 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1594 This fixes portability to compilers other than gcc, notably
1595 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1597 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1598 makefiles for portability to HP-UX and POSIX correctness.
1600 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1601 Also via indented variable definition in the Makefile.
1602 (Debugging by Oliver Heesakkers).
1604 PP/02 Restore caching of spamd results with expanded spamd_address.
1605 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1607 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1608 Improves build reliability. Fix from: Frank Elsner
1610 NM/02 Fix wide character breakage in the rfc2047 coding
1611 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1613 NM/03 Allow underscore in dnslist lookups
1614 Fixes bug 1026. Patch from Graeme Fowler
1616 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1617 Code patches from Adam Ciarcinski of NetBSD.
1619 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1622 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1623 is logged, to avoid truncation. Patch from John Horne.
1625 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1626 Patch from Jakob Hirsch.
1628 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1629 SQL string expansion failure details.
1630 Patch from Andrey Oktyabrski.
1632 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1633 Patch from Simon Arlott.
1635 PP/09 New lookups functionality failed to compile on old gcc which rejects
1636 extern declarations in function scope.
1637 Patch from Oliver Fleischmann
1639 PP/10 Use sig_atomic_t for flags set from signal handlers.
1640 Check getgroups() return and improve debugging.
1641 Fixed developed for diagnosis in bug 927 (which turned out to be
1644 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1645 Patch from Mark Zealey.
1647 PP/12 Bugzilla 1056: Improved spamd server selection.
1648 Patch from Mark Zealey.
1650 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1651 Based on patch from Heiko Schlittermann.
1653 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1654 Patch from Uwe Doering, sign-off by Michael Haardt.
1656 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1657 variable declaration deep within a block. Bug and patch from
1660 PP/15 lookups-Makefile IRIX compatibility coercion.
1662 PP/16 Make DISABLE_DKIM build knob functional.
1664 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1665 Patch by Simon Arlott
1667 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1668 variable argument macros. Our copy now differs from upstream.
1674 TF/01 Failure to get a lock on a hints database can have serious
1675 consequences so log it to the panic log.
1677 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1678 controlled using the smtp_confirmation log selector.
1680 TF/03 Include the error message when we fail to unlink a spool file.
1682 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1683 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1684 for maintaining out-of-tree patches for some time.
1686 PP/01 Bugzilla 139: Documentation and portability issues.
1687 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1688 Handle per-OS dynamic-module compilation flags.
1690 PP/02 Let /dev/null have normal permissions.
1691 The 4.73 fixes were a little too stringent and complained about the
1692 permissions on /dev/null. Exempt it from some checks.
1693 Reported by Andreas M. Kirchwitz.
1695 PP/03 Report version information for many libraries, including
1696 Exim version information for dynamically loaded libraries. Created
1697 version.h, now support a version extension string for distributors
1698 who patch heavily. Dynamic module ABI change.
1700 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1701 privilege escalation vulnerability whereby the Exim run-time user
1702 can cause root to append content of the attacker's choosing to
1705 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1708 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1709 If dropping privileges for untrusted macros, we disabled normal logging
1710 on the basis that it would fail; for the Exim run-time user, this is not
1711 the case, and it resulted in successful deliveries going unlogged.
1712 Fixed. Reported by Andreas Metzler.
1718 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1719 only prepend for the Resent-* case. Fixes regression introduced in
1720 Exim 4.70 by NM/22 for Bugzilla 607.
1722 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1723 increasing numbers of administrators be bitten by this.
1725 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1727 PP/03 Bugzilla 994: added openssl_options main configuration option.
1729 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1731 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1733 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1735 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1736 without a peer certificate, leading to a segfault because of an
1737 assumption that peers always have certificates. Be a little more
1738 paranoid. Problem reported by Martin Tscholak.
1740 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1741 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1742 NB: ClamAV planning to remove STREAM in "middle of 2010".
1743 CL also introduces -bmalware, various -d+acl logging additions and
1744 more caution in buffer sizes.
1746 PP/09 Implemented reverse_ip expansion operator.
1748 PP/10 Bugzilla 937: provide a "debug" ACL control.
1750 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1752 PP/12 Bugzilla 973: Implement --version.
1754 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1756 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1758 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1760 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1761 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1762 ignore trailing whitespace.
1764 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1766 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1769 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1770 Notification from Dr Andrew Aitchison.
1772 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1773 ExtendedDetectionInfo response format.
1774 Notification from John Horne.
1776 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1779 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1780 XSL and documented dependency on system catalogs, with examples of how
1783 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1786 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1787 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1788 configuration file which is writeable by the Exim user or group.
1790 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1791 of configuration files to cover files specified with the -C option if
1792 they are going to be used with root privileges, not just the default
1795 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1796 option (effectively making it always true).
1798 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1799 files to be used while preserving root privileges.
1801 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1802 that rogue child processes cannot use them.
1804 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1805 run-time user, instead of root.
1807 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1808 Exim run-time user without dropping privileges.
1810 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1811 result string, instead of calling string_vformat() twice with the same
1814 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1815 for other users. Others should always drop root privileges if they use
1816 -C on the command line, even for a whitelisted configure file.
1818 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1820 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1826 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1827 $data_path, and $header_path variables; fixed documentation bugs and
1830 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1831 exipick to access non-standard spools, including the "frozen" queue
1834 NM/01 Bugzilla 965: Support mysql stored procedures.
1835 Patch from Alain Williams
1837 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1839 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1840 Patch from Andreas Metzler
1842 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1843 Patch from Kirill Miazine
1845 NM/05 Bugzilla 671: Added umask to procmail example.
1847 JJ/03 installed exipick 20100323.0, fixing doc bug
1849 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1850 directory. Notification and patch from Dan Rosenberg.
1852 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1854 TK/02 Improve log output when DKIM signing operation fails.
1856 MH/01 Treat the transport option dkim_domain as a colon separated
1857 list, not as a single string, and sign the message with each element,
1858 omitting multiple occurences of the same signer.
1860 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1861 Bugzilla 985, 986. Patch by Simon Arlott
1863 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1864 Patch by Simon Arlott
1866 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1867 MBX locking. Notification from Dan Rosenberg.
1873 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1875 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1877 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1879 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1881 NM/04 Bugzilla 915: Flag broken perl installation during build.
1887 TK/01 Added patch by Johannes Berg that expands the main option
1888 "spamd_address" if it starts with a dollar sign.
1890 TK/02 Write list of recipients to X-Envelope-Sender header when building
1891 the mbox-format spool file for content scanning (suggested by Jakob
1894 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1895 (http://www.dcc-servers.net/) support via dccifd. Activated by
1896 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1898 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1899 by Mark Daniel Reidel <mr@df.eu>.
1901 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1902 When building exim an external PCRE library is now needed -
1903 PCRE is a system library on the majority of modern systems.
1904 See entry on PCRE_LIBS in EDITME file.
1906 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1907 conversation. Added nologin parameter to request.
1908 Patch contributed by Kirill Miazine.
1910 TF/01 Do not log submission mode rewrites if they do not change the address.
1912 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1914 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1915 log files in place. Contributed by Roberto Lima.
1917 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1919 TF/03 Bugzilla 615: When checking the local_parts router precondition
1920 after a local_part_suffix or local_part_prefix option, Exim now
1921 does not use the address's named list lookup cache, since this
1922 contains cached lookups for the whole local part.
1924 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1925 Robert Millan. Documentation is in experimental-spec.txt.
1927 TF/04 Bugzilla 668: Fix parallel build (make -j).
1929 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1931 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1932 Patch provided by Jan Srzednicki.
1934 TF/05 Leading white space used to be stripped from $spam_report which
1935 wrecked the formatting. Now it is preserved.
1937 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1938 that they are available at delivery time.
1940 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1942 TF/08 TLS error reporting now respects the incoming_interface and
1943 incoming_port log selectors.
1945 TF/09 Produce a more useful error message if an SMTP transport's hosts
1946 setting expands to an empty string.
1948 NM/06 Bugzilla 744: EXPN did not work under TLS.
1949 Patch provided by Phil Pennock.
1951 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1952 Patch provided by Richard Godbee.
1954 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1955 acl_smtp_notquit, added index entry.
1957 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1958 Patch provided by Eugene Bujak.
1960 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1961 accept(). Patch provided by Maxim Dounin.
1963 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1964 Patch provided by Phil Pennock.
1966 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1968 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1969 Patch provided by Brad "anomie" Jorsch.
1971 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1972 Patch provided by Dean Brooks.
1974 TK/05 Add native DKIM support (does not depend on external libraries).
1976 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1977 Patch provided by Graeme Fowler.
1979 NM/16 Bugzilla 851: Documentation example syntax fix.
1981 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1983 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1986 NM/19 Bugzilla 745: TLS version reporting.
1987 Patch provided by Phil Pennock.
1989 NM/20 Bugzilla 167: bool: condition support.
1990 Patch provided by Phil Pennock.
1992 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1993 clients. Patch provided by Phil Pennock.
1995 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1996 Patch provided by Brad "anomie" Jorsch.
1998 NM/23 Bugzilla 687: Fix misparses in eximstats.
1999 Patch provided by Heiko Schlittermann.
2001 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2002 Patch provided by Heiko Schlittermann.
2004 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2005 plus update to original patch.
2007 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2009 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2010 Patch provided by David Brownlee.
2012 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2014 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2016 NM/30 Bugzilla 888: TLS documentation bugfixes.
2018 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2020 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2021 Unlike the original bugzilla I have changed all shell scripts in src tree.
2023 NM/33 Bugzilla 898: Transport filter timeout fix.
2024 Patch by Todd Rinaldo.
2026 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2027 Patch by Serge Demonchaux.
2029 NM/35 Bugzilla 39: Base64 decode bug fixes.
2030 Patch by Jakob Hirsch.
2032 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2034 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2036 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2038 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2044 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2045 ALT-N's libdkim that I have put here:
2046 http://duncanthrax.net/exim-experimental/
2048 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2049 were called 'true' and it seems that C99 defines that as a reserved
2050 keyword to be used with 'bool' variable types. That means you could
2051 not include C99-style headers which use bools without triggering
2052 build errors in sieve.c.
2054 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2055 as mailq or other aliases. Changed the --help handling significantly
2056 to do whats expected. exim_usage() emits usage/help information.
2058 SC/01 Added the -bylocaldomain option to eximstats.
2060 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2062 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2064 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2070 PH/01 Another patch from the Sieve maintainer.
2072 PH/02 When an IPv6 address is converted to a string for single-key lookup
2073 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2074 dots are used instead of colons so that keys in lsearch files need not
2075 contain colons. This was done some time before quoting was made available
2076 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2077 (notated using the quote facility) so as to distinguish them from IPv4
2078 keys. This meant that lookups for IP addresses in host lists did not work
2079 for iplsearch lookups.
2081 This has been fixed by arranging for IPv6 addresses to be expressed with
2082 colons if the lookup type is iplsearch. This is not incompatible, because
2083 previously such lookups could never work.
2085 The situation is now rather anomalous, since one *can* have colons in
2086 ordinary lsearch keys. However, making the change in all cases is
2087 incompatible and would probably break a number of configurations.
2089 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2092 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2093 conversion specification without a maximum field width, thereby enabling
2094 a rogue spamd server to cause a buffer overflow. While nobody in their
2095 right mind would setup Exim to query an untrusted spamd server, an
2096 attacker that gains access to a server running spamd could potentially
2097 exploit this vulnerability to run arbitrary code as the Exim user.
2099 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2100 $primary_hostname instead of what libspf2 thinks the hosts name is.
2102 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2103 a directory entry by the name of the lookup key. Previously, if a
2104 symlink pointed to a non-existing file or a file in a directory that
2105 Exim lacked permissions to read, a lookup for a key matching that
2106 symlink would fail. Now it is enough that a matching directory entry
2107 exists, symlink or not. (Bugzilla 503.)
2109 PH/03 The body_linecount and body_zerocount variables are now exported in the
2112 PH/04 Added the $dnslist_matched variable.
2114 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2115 This means they are set thereafter only if the connection becomes
2118 PH/06 Added the client_condition to authenticators so that some can be skipped
2119 by clients under certain conditions.
2121 PH/07 The error message for a badly-placed control=no_multiline_responses left
2122 "_responses" off the end of the name.
2124 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2126 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2127 (without spaces) instead of just copying the configuration text.
2129 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2131 PH/11 Added $max_received_linelength.
2133 PH/12 Added +ignore_defer and +include_defer to host lists.
2135 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2136 way in which PCRE > 7.0 is built.
2138 PH/14 Implemented queue_only_load_latch.
2140 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2141 MAIL command. The effect was to mangle the value on 64-bit systems.
2143 PH/16 Another patch from the Sieve maintainer.
2145 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2147 PH/18 If a system quota error occurred while trying to create the file for
2148 a maildir delivery, the message "Mailbox is full" was not appended to the
2149 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2150 only to a quota excession during the actual writing of the file.
2152 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2153 characters?) which causes problems in log lines. The DN values are now
2154 passed through string_printing() before being added to log lines.
2156 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2157 and InterBase are left for another time.)
2159 PH/21 Added message_body_newlines option.
2161 PH/22 Guard against possible overflow in moan_check_errorcopy().
2163 PH/23 POSIX allows open() to be a macro; guard against that.
2165 PH/24 If the recipient of an error message contained an @ in the local part
2166 (suitably quoted, of course), incorrect values were put in $domain and
2167 $local_part during the evaluation of errors_copy.
2173 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2174 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2177 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2178 issue a MAIL command.
2180 PH/02 In an ACL statement such as
2182 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2184 if a client was not listed at all, or was listed with a value other than
2185 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2186 the condition was not true (as it should be), so access was not denied.
2187 The bug was that the ! inversion was incorrectly passed on to the second
2188 item. This has been fixed.
2190 PH/03 Added additional dnslists conditions == and =& which are different from
2191 = and & when the dns lookup returns more than one IP address.
2193 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2194 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2196 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2197 FSYNC, which compiles an option called disable_fsync that allows for
2198 bypassing fsync(). The documentation is heavily laced with warnings.
2200 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2202 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2203 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2204 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2205 including adding "make clean"; (3) Added -fPIC when compiling the test
2206 dynamically loaded module, to get rid of a warning.
2208 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2209 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2210 = 0s. The bug is otherwise harmless.
2212 PH/07 There was a bug in the dovecot authenticator such that the value of
2213 $auth1 could be overwritten, and so not correctly preserved, after a
2214 successful authentication. This usually meant that the value preserved by
2215 the server_setid option was incorrect.
2217 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2219 PH/09 Installed PCRE release 7.0.
2221 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2222 run for batched SMTP input. It is now run at the start of every message
2223 in the batch. While fixing this I discovered that the process information
2224 (output by running exiwhat) was not always getting set for -bs and -bS
2225 input. This is fixed, and it now also says "batched" for BSMTP.
2227 PH/11 Added control=no_pipelining.
2229 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2230 patch, slightly modified), and move the expansion of helo_data till after
2231 the connection is made in the smtp transport (so it can use these
2234 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2236 PH/14 Added log_selector = +pid.
2238 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2240 PH/16 Add ${if forany and ${if forall.
2242 PH/17 Added dsn_from option to vary the From: line in DSNs.
2244 PH/18 Flush SMTP output before performing a callout, unless control =
2245 no_callout_flush is set.
2247 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2248 was true (the default) a successful delivery failed to delete the retry
2249 item, thus causing premature timeout of the address. The bug is now
2252 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2254 PH/21 Long custom messages for fakedefer and fakereject are now split up
2255 into multiline responses in the same way that messages for "deny" and
2256 other ACL rejections are.
2258 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2259 with slight modification.
2261 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2262 draft, changing the syntax and factoring some duplicate code".
2264 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2265 for deliveries of the second and subsequent messages over the same SMTP
2268 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2269 ${reduce, with only minor "tidies".
2271 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2273 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2274 expansion side effects.
2276 PH/27 When a message times out after an over-quota error from an Exim-imposed
2277 quota, the bounce message says "mailbox is full". This message was not
2278 being given when it was a system quota that was exceeded. It now should
2281 MH/03 Made $recipients available in local_scan(). local_scan() already has
2282 better access to the recipient list through recipients_list[], but
2283 $recipients can be useful in postmaster-provided expansion strings.
2285 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2286 in the case of a MAIL command with additional options following the
2287 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2288 were accidentally chopped off.
2290 PH/29 SMTP synchronization checks are implemented when a command is read -
2291 there is a check that no more input is waiting when there shouldn't be
2292 any. However, for some commands, a delay in an ACL can mean that it is
2293 some time before the response is written. In this time, more input might
2294 arrive, invalidly. So now there are extra checks after an ACL has run for
2295 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2296 pipelining has not been advertised.
2298 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2300 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2301 returned in the SMTP response when smtp_return_error_details was false.
2302 This has been fixed.
2304 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2305 instead of the C I/O that was originally supplied, because problems were
2306 reported on Solaris.
2308 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2309 Exim which did not show up earlier: it was assuming that a call to
2310 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2311 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2312 was a macro that became an assignment, so it seemed to work. This has
2313 changed to a proper function call with a void return, hence the compile
2314 error. Exim's code has been fixed.
2316 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2319 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2321 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2323 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2324 because it turns out to be too big or there is a timeout) there is no
2325 "Completed" line in the log. When some messages of this type were
2326 selected by exigrep, they were listed as "not completed". Others were
2327 picked up by some special patterns. I have improved the selection
2328 criteria to be more general.
2330 PH/38 The host_find_failed option in the manualroute router can now be set
2331 to "ignore", to completely ignore a host whose IP address cannot be
2332 found. If all hosts are ignored, the behaviour is controlled by the new
2333 host_all_ignored option.
2335 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2336 homing or because of multiple MX records with /mx) generated more than
2337 one IP address, and the following item turned out to be the local host,
2338 all the secondary addresses of the first item were incorrectly removed
2339 from the list, along with the local host and any following hosts (which
2340 is what is supposed to happen).
2342 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2343 whoever called Exim into the -H file. In the case of the daemon it was
2344 behaving confusingly. When first started, it used values for whoever
2345 started the daemon, but after a SIGHUP it used the Exim user (because it
2346 calls itself on a restart). I have changed the code so that it now always
2349 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2350 message are rejected with the same error (e.g. no authentication or bad
2351 sender address), and a DATA command is nevertheless sent (as can happen
2352 with PIPELINING or a stupid MUA), the error message that was given to the
2353 RCPT commands is included in the rejection of the DATA command. This is
2354 intended to be helpful for MUAs that show only the final error to their
2357 PH/42 Another patch from the Sieve maintainer.
2359 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2360 Eximstats - Fixed some broken HTML links and added missing column headers
2362 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2363 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2365 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2367 PH/43 Yet another patch from the Sieve maintainer.
2369 PH/44 I found a way to check for a TCP/IP connection going away before sending
2370 the response to the final '.' that terminates a message, but only in the
2371 case where the client has not sent further data following the '.'
2372 (unfortunately, this is allowed). However, in many cases there won't be
2373 any further data because there won't be any more messages to send. A call
2374 to select() can be used: if it shows that the input is "ready", there is
2375 either input waiting, or the socket has been closed. An attempt to read
2376 the next input character can distinguish the two cases. Previously, Exim
2377 would have sent an OK response which the client would never have see.
2378 This could lead to message repetition. This fix should cure that, at
2379 least in a lot of common cases.
2381 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2382 advertised in response to EHLO.
2388 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2389 fixed by 4.65/MH/01 (is this a record?) are fixed:
2391 (i) An empty string was always treated as zero by the numeric comparison
2392 operators. This behaviour has been restored.
2394 (ii) It is documented that the numeric comparison operators always treat
2395 their arguments as decimal numbers. This was broken in that numbers
2396 starting with 0 were being interpreted as octal.
2398 While fixing these problems I realized that there was another issue that
2399 hadn't been noticed. Values of message_size_limit (both the global option
2400 and the transport option) were treated as octal if they started with 0.
2401 The documentation was vague. These values are now always treated as
2402 decimal, and I will make that clear in the documentation.
2408 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2409 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2412 MH/01 Don't check that the operands of numeric comparison operators are
2413 integers when their expansion is in "skipping" mode (fixes bug
2414 introduced by 4.64-PH/07).
2416 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2417 child addresses, Exim now panics and dies. Previously, because the count
2418 is held in a short int, deliveries were likely to be lost. As such a
2419 large number of recipients for a single message is ridiculous
2420 (performance will be very, very poor), I have chosen to impose a limit
2421 rather than extend the field.
2427 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2428 leftover -K file (the existence of which was triggered by #402).
2429 While we were at it, introduced process PID as part of the -K
2430 filename. This should rule out race conditions when creating
2433 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2434 processing considerably. Previous code took too long for large mails,
2435 triggering a timeout which in turn triggers #401.
2437 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2438 in the DK code in transports.c. sendfile() is not really portable,
2439 hence the _LINUX specificness.
2441 TF/01 In the add_headers option to the mail command in an Exim filter,
2442 there was a bug that Exim would claim a syntax error in any
2443 header after the first one which had an odd number of characters
2446 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2447 callout verification, Exim cached a "reject" for the entire domain. This
2448 is correct for most verifications, but it is not correct for a recipient
2449 verification with use_sender or use_postmaster set, because in that case
2450 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2451 case of MAIL FROM:<> rejection from other early rejections (e.g.
2452 rejection of HELO). When verifying a recipient using a non-null MAIL
2453 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2454 Whatever the result of the callout, the value of the domain cache is
2455 left unchanged (for any other kind of callout, getting as far as trying
2456 RCPT means that the domain itself is ok).
2458 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2461 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2462 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2463 session, the daemon ignores ECONNECT errors and logs others; it now
2464 ignores EPIPE as well.
2466 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2467 (quoted-printable decoding).
2469 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2470 later the small subsequent patch to fix an introduced bug.
2472 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2474 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2476 PH/08 An error is now given if message_size_limit is specified negative.
2478 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2479 to be given (somewhat) arbitrary names.
2481 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2484 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2485 miscellaneous code fixes
2487 PH/10 Added the log_reject_target ACL modifier to specify where to log
2490 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2491 hostname. This is wrong, because it relates to the incoming message (and
2492 probably the interface on which it is arriving) and not to the outgoing
2493 callout (which could be using a different interface). This has been
2494 changed to use the value of the helo_data option from the smtp transport
2495 instead - this is what is used when a message is actually being sent. If
2496 there is no remote transport (possible with a router that sets up host
2497 addresses), $smtp_active_hostname is used.
2499 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2500 tweaks were necessary in order to get it to work (see also 21 below):
2501 (a) The code assumed that strncpy() returns a negative number on buffer
2502 overflow, which isn't the case. Replaced with Exim's string_format()
2504 (b) There were several signed/unsigned issues. I just did the minimum
2505 hacking in of casts. There is scope for a larger refactoring.
2506 (c) The code used strcasecmp() which is not a standard C function.
2507 Replaced with Exim's strcmpic() function.
2508 (d) The code set only $1; it now sets $auth1 as well.
2509 (e) A simple test gave the error "authentication client didn't specify
2510 service in request". It would seem that Dovecot has changed its
2511 interface. Fortunately there's a specification; I followed it and
2512 changed what the client sends and it appears to be working now.
2514 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2517 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2518 address A is aliased to B and C, where B exists and C does not. Without
2519 -v the output is "A verified" because verification stops after a
2520 successful redirection if more than one address is generated. However,
2521 with -v the child addresses are also verified. Exim was outputting "A
2522 failed to verify" and then showing the successful verification for C,
2523 with its parentage. It now outputs "B failed to verify", showing B's
2524 parentage before showing the successful verification of C.
2526 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2527 look up a TXT record in a specific list after matching in a combined
2530 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2531 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2532 they consult the DNS. I had assumed they would set it the way they
2533 wanted; and indeed my experiments on Linux seem to show that in some
2534 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2535 To be on the safe side, however, I have now made the interface to
2536 host_find_byname() similar to host_find_bydns(), with an argument
2537 containing the DNS resolver options. The host_find_byname() function now
2538 sets these options at its start, just as host_find_bydns() does. The smtp
2539 transport options dns_qualify_single and dns_search_parents are passed to
2540 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2541 of host_find_byname() use the default settings of RES_DEFNAMES
2542 (qualify_single) but not RES_DNSRCH (search_parents).
2544 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2545 spool_read_header() do less string testing, by means of a preliminary
2546 switch on the second character of optional "-foo" lines. (This is
2547 overdue, caused by the large number of possibilities that now exist.
2548 Originally there were few.) While I was there, I also converted the
2549 str(n)cmp tests so they don't re-test the leading "-" and the first
2550 character, in the hope this might squeeze out yet more improvement.
2552 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2553 flag allowing group syntax was set by the header_syntax check but not
2554 turned off, possible causing trouble later; (2) The flag was not being
2555 set at all for the header_verify test, causing "group"-style headers to
2556 be rejected. I have now set it in this case, and also caused header_
2557 verify to ignore an empty address taken from a group. While doing this, I
2558 came across some other cases where the code for allowing group syntax
2559 while scanning a header line wasn't quite right (mostly, not resetting
2560 the flag correctly in the right place). These bugs could have caused
2561 trouble for malformed header lines. I hope it is now all correct.
2563 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2564 with the "reply" argument non-NULL. The code, however (which originally
2565 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2566 but it didn't always do it. This confused somebody who was copying the
2567 code for some other use. I have removed all the tests.
2569 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2570 feature that was used to support insecure browsers during the U.S. crypto
2571 embargo. It requires special client support, and Exim is probably the
2572 only MTA that supported it -- and would never use it because real RSA is
2573 always available. This code has been removed, because it had the bad
2574 effect of slowing Exim down by computing (never used) parameters for the
2575 RSA_EXPORT functionality.
2577 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2578 authenticator to fail if there's a tab character in the incoming data
2579 (there should never be unless someone is messing about, as it's supposed
2580 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2581 if the connection is using TLS or if the remote IP is the same as the
2582 local IP, and the "valid-client-cert option" if a client certificate has
2585 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2586 authenticators. This can be used for authorization after authentication
2587 succeeds. (In the case of plaintext, it servers for both authentication
2590 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2591 if any retry times were supplied.
2593 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2594 connection, where there is no client IP address to check. In this
2595 situation, the verify now always succeeds.
2597 PH/25 Applied John Jetmore's -Mset patch.
2599 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2601 PH/27 In a string expansion for a processed (not raw) header when multiple
2602 headers of the same name were present, leading whitespace was being
2603 removed from all of them, but trailing whitespace was being removed only
2604 from the last one. Now trailing whitespace is removed from each header
2605 before concatenation. Completely empty headers in a concatenation (as
2606 before) are ignored.
2608 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2609 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2611 PH/29 [Removed. This was a change that I later backed out, and forgot to
2612 correct the ChangeLog entry (that I had efficiently created) before
2613 committing the later change.]
2615 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2616 address errors (4xx response to RCPT) over the same connection as other
2617 messages routed to the same hosts. Such deliveries are always "forced",
2618 so retry times are not inspected. This resulted in far too many retries
2619 for the affected addresses. The effect occurred only when there were more
2620 hosts than the hosts_max_try setting in the smtp transport when it had
2621 the 4xx errors. Those hosts that it had tried were not added to the list
2622 of hosts for which the message was waiting, so if all were tried, there
2623 was no problem. Two fixes have been applied:
2625 (i) If there are any address or message errors in an SMTP delivery, none
2626 of the hosts (tried or untried) are now added to the list of hosts
2627 for which the message is waiting, so the message should not be a
2628 candidate for sending over the same connection that was used for a
2629 successful delivery of some other message. This seems entirely
2630 reasonable: after all the message is NOT "waiting for some host".
2631 This is so "obvious" that I'm not sure why it wasn't done
2632 previously. Hope I haven't missed anything, but it can't do any
2633 harm, as the worst effect is to miss an optimization.
2635 (ii) If, despite (i), such a delivery is accidentally attempted, the
2636 routing retry time is respected, so at least it doesn't keep
2637 hammering the server.
2639 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2640 in ${readsocket because some servers need this prod.
2642 PH/32 Added some extra debug output when updating a wait-xxx database.
2644 PH/33 The hint "could be header name not terminated by colon", which has been
2645 given for certain expansion errors for a long time, was not being given
2646 for the ${if def:h_colon_omitted{... case.
2648 PH/34 The spec says: "With one important exception, whenever a domain list is
2649 being scanned, $domain contains the subject domain." There was at least
2650 one case where this was not true.
2652 PH/35 The error "getsockname() failed: connection reset by peer" was being
2653 written to the panic log as well as the main log, but it isn't really
2654 panic-worthy as it just means the connection died rather early on. I have
2655 removed the panic log writing for the ECONNRESET error when getsockname()
2658 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2659 runs only) independently of the message's sender address. This meant
2660 that, if the 4xx error was in fact related to the sender, a different
2661 message to the same recipient with a different sender could confuse
2662 things. In particular, this can happen when sending to a greylisting
2663 server, but other circumstances could also provoke similar problems.
2664 I have changed the default so that the retry time for these errors is now
2665 based a combination of the sender and recipient addresses. This change
2666 can be overridden by setting address_retry_include_sender=false in the
2669 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2670 remote server are returned as part of bounce messages. This was not
2671 happening for LMTP over a pipe (the lmtp transport), but now it is the
2672 same for both kinds of LMTP.
2674 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2675 in header lines that were in fact CNAMEs. This is no longer the case.
2677 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2678 and queue runs started by the daemon processed all messages. This has
2679 been fixed so that -R and -S can now usefully be given with -q<time>.
2681 PH/40 Import PCRE release 6.7 (fixes some bugs).
2683 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2685 PH/42 Give an error if -q is specified more than once.
2687 PH/43 Renamed the variables $interface_address and $interface_port as
2688 $received_ip_address and $received_port, to make it clear that these
2689 values apply to message reception, and not to the outgoing interface when
2690 a message is delivered. (The old names remain recognized, of course.)
2692 PH/44 There was no timeout on the connect() call when using a Unix domain
2693 socket in the ${readsocket expansion. There now is.
2695 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2696 be meaningful with "accept".
2698 SC/01 Eximstats V1.43
2699 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2701 SC/02 Eximstats V1.44
2702 Use a glob alias rather than an array ref in the generated
2703 parser. This improves both readability and performance.
2705 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2706 Collect SpamAssassin and rejection statistics.
2707 Don't display local sender or destination tables unless
2708 there is data to show.
2709 Added average volumes into the top table text output.
2711 SC/04 Eximstats V1.46
2712 Collect data on the number of addresses (recipients)
2713 as well as the number of messages.
2715 SC/05 Eximstats V1.47
2716 Added 'Message too big' to the list of mail rejection
2717 reasons (thanks to Marco Gaiarin).
2719 SC/06 Eximstats V1.48
2720 Mainlog lines which have GMT offsets and are too short to
2721 have a flag are now skipped.
2723 SC/07 Eximstats V1.49 (Alain Williams)
2724 Added the -emptyok flag.
2726 SC/08 Eximstats V1.50
2727 Fixes for obtaining the IP address from reject messages.
2729 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2730 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2731 whitespace changes from 4.64-PH/27
2733 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2736 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2737 are found, allow negative numbers in numeric criteria)
2739 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2741 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2742 to match changes made in 4.64-PH/43
2744 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2746 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2747 "domain" is actually an IP address, and force a failure. This locks out
2748 those revolvers/nameservers that support "A-for-A" lookups, in
2749 contravention of the specifications.
2751 PH/48 When a host name was looked up from an IP address, and the subsequent
2752 forward lookup of the name timed out, the host name was left in
2753 $sender_host_name, contrary to the specification.
2755 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2756 restricted to single-key lookups, Exim was not diagnosing an error if
2757 * or *@ was used with a query-style lookup.
2759 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2761 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2762 long ago, but noone interested enough thought of it. Let's just say that
2763 the "1.1" means that there are some new functions that weren't there at
2764 some point in the past.
2766 PH/51 Error processing for expansion failure of helo_data from an smtp
2767 transport during callout processing was broken.
2769 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2770 tested/used via the -bh/-bhc/-bs options.
2772 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2773 bug, fixed in subsequent PCRE releases).
2775 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2776 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2778 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2784 SC/01 Use a glob alias rather than an array ref in eximstats generated
2785 parser. This improves both readability and performance.
2787 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2788 Don't display local sender or destination tables in eximstats unless
2789 there is data to show.
2790 Added average volumes into the eximstats top table text output.
2792 SC/03 Collect data on the number of addresses (recipients) as well
2793 as the number of messages in eximstats.
2795 TF/01 Correct an error in the documentation for the redirect router. Exim
2796 does (usually) call initgroups() when daemonizing.
2798 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2799 with consistent privilege compared to when running as a daemon.
2801 TF/03 Note in the spec that $authenticated_id is not set for local
2802 submissions from trusted users.
2804 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2805 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2807 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2808 by adding some example configuration directives to the default
2809 configuration file. A little bit of work is required to uncomment the
2810 directives and define how usernames and passwords are checked, but
2811 there is now a framework to start from.
2813 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2814 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2815 without this. I don't know how relevant this is to other LDAP libraries.
2817 PH/02 Add the verb name to the "unknown ACL verb" error.
2819 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2821 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2823 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2824 directory not expanded when it should be if an expanded home directory
2825 was set for the address (which is overridden by the transport).
2827 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2830 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2831 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2832 because it is too late at that time, and has no effect.
2834 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2835 security issue with \' (bugzilla #107). I could not use the
2836 PQescapeStringConn() function, because it needs a PGconn value as one of
2839 PH/08 When testing addresses using -bt, indicate those final addresses that
2840 are duplicates that would not cause an additional delivery. At least one
2841 person was confused, thinking that -bt output corresponded to deliveries.
2842 (Suppressing duplicates isn't a good idea as you lose the information
2843 about possibly different redirections that led to the duplicates.)
2845 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2846 systems where poll() doesn't work, in particular OS X.
2848 PH/10 Added more information to debugging output for retry time not reached.
2850 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2851 operations in malware.c.
2853 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2856 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2857 syslog_duplication set false, log lines that would normally be written
2858 both the the main log and to the reject log were not written to syslog at
2861 PH/14 In the default configuration, change the use of "message" in ACL warn
2862 statements to "add_header".
2864 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2865 not followed by a command (e.g. "seen endif").
2867 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2868 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2871 PH/17 Added extra conditions to the default value of delay_warning_condition
2875 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2876 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2877 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2880 The Auto-Submitted: and various List- headers are standardised, whereas I
2881 don't think Precedence: ever was.
2883 PH/18 Refactored debugging code in route_finduser() to show more information,
2884 in particular, the error code if getpwnam() issues one.
2886 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2887 This is apparently needed in addition to the PH/07 change above to avoid
2888 any possible encoding problems.
2890 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2891 but not after initializing Perl.
2893 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2894 output them only if debugging. By default they are written stderr,
2895 apparently, which is not desirable.
2897 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2900 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2903 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2905 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2906 authenticated or an ident call has been made. Suppress the default
2907 values for $authenticated_id and $authenticated_sender (but permit -oMai
2908 and -oMas) when testing with -bh.
2910 PH/24 Re-jigged the order of the tests in the default configuration so that the
2911 tests for valid domains and recipients precede the DNS black list and CSA
2912 tests, on the grounds that those ones are more expensive.
2914 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2915 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2916 This bug exists in every version of Exim that I still have, right back to
2919 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2920 However, an attempt to turn on case-sensitivity in a regex key by
2921 including (?-i) didn't work because the subject string was already
2922 lowercased, and the effects were non-intuitive. It turns out that a
2923 one-line patch can be used to allow (?-i) to work as expected.
2929 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2930 other effects) broke the use of negated acl sub-conditions.
2932 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2935 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2936 "Deny" causes Exim to reject the incoming connection with a 554 error.
2937 Unfortunately, if there is a major crisis, such as a disk failure,
2938 tcp-wrappers gives "deny", whereas what one would like would be some
2939 kind of temporary error. A kludge has been added to help with this.
2940 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2941 554 error is used if errno is still zero or contains ENOENT (which occurs
2942 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2945 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2947 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2948 errors. Otherwise a message that provokes a temporary error (when other
2949 messages do not) can cause a whole host to time out.
2951 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2952 addresses were routed directly to files or pipes from a redirect router.
2953 File deliveries just didn't batch; pipe deliveries might have suffered
2956 PH/06 A failure to get a lock for a hints database would erroneously always say
2957 "Failed to get write lock", even when it was really a read lock.
2959 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2960 of 0600. This has been changed to use the value of the lockfile_mode
2961 option (which defaults to 0600).
2963 PH/08 Applied small patch from the Sieve maintainer.
2965 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2966 folder from quota calculations, a direct delivery into this folder messed
2967 up the contents of the maildirsize file. This was because the regex was
2968 used only to exclude .Trash (or whatever) when the size of the mailbox
2969 was calculated. There was no check that a delivery was happening into an
2970 excluded directory. This bug has been fixed by ignoring all quota
2971 processing for deliveries into excluded directories.
2973 PH/10 Added the maildirfolder_create_regex option to appendfile.
2979 PH/01 The code for finding all the local interface addresses on a FreeBSD
2980 system running IPv6 was broken. This may well have applied to all BSD
2981 systems, as well as to others that have similar system calls. The broken
2982 code found IPv4 interfaces correctly, but gave incorrect values for the
2983 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2984 that it would not match correctly against @[] and not recognize the IPv6
2987 PH/02 The ipliteral router was not recognizing addresses of the form user@
2988 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2990 PH/03 Added disable_ipv6.
2992 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2993 decoded form, because it is most often used to construct To: headers
2994 lines in autoreplies, and the decoded form may well be syntactically
2995 invalid. However, $reply_address has leading white space removed, and all
2996 newlines turned into spaces so that the autoreply transport does not
2999 PH/05 If group was specified without a user on a router, and no group or user
3000 was specified on a transport, the group from the router was ignored.
3002 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3003 for visible compile-time settings that can be used to change these
3004 numbers, for those that want even more. Backwards compatibility with old
3005 spool files has been maintained. However, going back to a previous Exim
3006 release will lost any variables that are in spool files.
3008 PH/07 Two small changes when running in the test harness: increase delay when
3009 passing a TCP/IP connection to a new process, in case the original
3010 process has to generate a bounce, and remove special handling of
3011 127.0.0.2 (sic), which is no longer necessary.
3013 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3014 be the same on different OS.
3016 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3019 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3020 whether --show-vars was specified or not
3022 JJ/02 exipick: Added support for new ACL variable spool format introduced
3025 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3026 syntactically invalid From: or Reply-to: line, and a filter used this to
3027 generate an autoreply, and therefore failed to obtain an address for the
3028 autoreply, Exim could try to deliver to a non-existent relative file
3029 name, causing unrelated and misleading errors. What now happens is that
3030 it logs this as a hard delivery error, but does not attempt to create a
3033 PH/11 The exinext utility has a -C option for testing purposes, but although
3034 the given file was scanned by exinext itself; it wasn't being passed on
3035 when Exim was called.
3037 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3038 an end-of-file indication when reading a command response.
3040 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3041 compiled. In many other places in Exim, IPv6 addresses are always
3042 recognized, so I have changed this. It also means that IPv4 domain
3043 literals of the form [IPV4:n.n.n.n] are now always recognized.
3045 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3046 used if the router is not running as root, for example, when verifying at
3047 ACL time, or when using -bh. The debugging output from this situation was
3048 non-existent - all you got was a failure to exec. I have made two
3051 (a) Failures to set uid/gid, the current directory, or a process leader
3052 in a subprocess such as that created by queryprogram now generate
3053 suitable debugging output when -d is set.
3055 (b) The queryprogram router detects when it is not running as root,
3056 outputs suitable debugging information if -d is set, and then runs
3057 the subprocess without attempting to change uid/gid.
3059 PH/15 Minor change to Makefile for building test_host (undocumented testing
3062 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3063 additional section of a DNS packet that returns MX or SRV records.
3064 Instead, it always explicitly searches for A/AAAA records. This avoids
3065 major problems that occur when a DNS server includes only records of one
3066 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3067 fixed another bug: if SRV records were looked up and the corresponding
3068 address records were *not* found in the additional section, the port
3069 values from the SRV records were lost.
3071 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3072 using the correct key (the original address) when searching the retry
3073 rules in order to find which one to use for generating the retry hint.
3075 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3076 adding the default one. Similarly, if it contains a Reply-To: header, the
3077 errors_reply_to option, if set, is not used.
3079 PH/19 When calculating a retry time, Exim used to measure the "time since
3080 failure" by looking at the "first failed" field in the retry record. Now
3081 it does not use this if it is later than than the arrival time of the
3082 message. Instead it uses the arrival time. This makes for better
3083 behaviour in cases where some deliveries succeed, thus re-setting the
3084 "first failed" field. An example is a quota failure for a huge message
3085 when small messages continue to be delivered. Without this change, the
3086 "time since failure" will always be short, possible causing more frequent
3087 delivery attempts for the huge message than are intended.
3088 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3090 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3091 $1, $2, $3) because the numerical variables can be reset during some
3092 expansion items (e.g. "match"), thereby losing the authentication data.
3094 PH/21 Make -bV show the size of off_t variables so that the test suite can
3095 decide whether to run tests for quotas > 2G.
3097 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3098 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3099 filecount value is greater than 2G or if a quota value is greater than 2G
3100 on a system where the size of off_t is not greater than 4, a panic error
3103 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3104 never match. The debug and -bh output now contains an explicit error
3105 message indicating a malformed IPv4 address or mask.
3107 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3108 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3109 PH/23 above applies.
3111 PH/25 Do not write to syslog when running in the test harness. The only
3112 occasion when this arises is a failure to open the main or panic logs
3113 (for which there is an explicit test).
3115 PH/26 Added the /no_tell option to "control=freeze".
3117 PH/27 If a host name lookup failed very early in a connection, for example, if
3118 the IP address matched host_lookup and the reverse lookup yielded a name
3119 that did not have a forward lookup, an error message of the form "no IP
3120 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3121 could be logged. Now it outputs the IP address instead of "NULL".
3123 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3124 allows the sender and the authenticated sender to be set when
3125 submitting a message from within Exim. Since child_open_exim() is
3126 documented for local_scan(), the new function should be too.
3128 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3129 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3130 results in an empty string is now treated as unset.
3132 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3134 PH/31 Added sender_verify_fail logging option.
3136 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3137 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3138 tidied the source and removed it altogether.
3140 PH/33 When a queue run was abandoned because the load average was too high, a
3141 log line was always written; now it is written only if the queue_run log
3142 selector is set. In addition, the log line for abandonment now contains
3143 information about the queue run such as the pid. This is always present
3144 in "start" and "stop" lines but was omitted from the "abandon" line.
3146 PH/34 Omit spaces between a header name and the colon in the error message that
3147 is given when verify = headers_syntax fails (if there are lots of them,
3148 the message gets confusing).
3150 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3151 names, as there are now some PTR records that contain slashes. This check
3152 is only to protect against broken name servers that fall over on strange
3153 characters, so the fact that it applies to all lookups doesn't matter.
3155 PH/36 Now that the new test suite is complete, we can remove some of the
3156 special code in Exim that was needed for the old test suite. For example,
3157 sorting DNS records because real resolvers return them in an arbitrary
3158 order. The new test suite's fake resolver always returns records in the
3161 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3162 bounces) except when queue_only is set, to avoid logging races between
3163 the different processes.
3165 PH/38 Panic-die if .include specifies a non-absolute path.
3167 PH/39 A tweak to the "H" retry rule from its user.
3169 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3170 a label. They prevented compilation on older perls.
3172 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3173 a warning to be raised on newish perls.
3175 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3176 on queue. Changes to match documented behaviour of showing count of
3177 messages matching specified criteria.
3179 PH/40 Changed the default ident timeout from 30s to 5s.
3181 PH/41 Added support for the use of login_cap features, on those BSD systems
3182 that have them, for controlling the resources used by pipe deliveries.
3184 PH/42 The content-scanning code uses fopen() to create files in which to put
3185 message data. Previously it was not paying any attention to the mode of
3186 the files. Exim runs with umask(0) because the rest of the code creates
3187 files with open(), and sets the required mode explicitly. Thus, these
3188 files were ending up world-writeable. This was not a big issue, because,
3189 being within the spool directory, they were not world-accessible. I have
3190 created a function called modefopen, which takes an additional mode
3191 argument. It sets umask(777), creates the file, chmods it to the required
3192 mode, then resets the umask. All the relevant calls to fopen() in the
3193 content scanning code have been changed to use this function.
3195 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3196 to 24 hours. This avoids potential overflow problems when processing G
3197 and H retry rules. I suspect nobody ever tinkers with this value.
3199 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3201 PH/45 When the plaintext authenticator is running as a client, the server's
3202 challenges are checked to ensure they are valid base64 strings. By
3203 default, the authentication attempt is cancelled if an invalid string is
3204 received. Setting client_ignore_invalid_base64 true ignores these errors.
3205 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3206 they are received. Thus, the responses can be made to depend on the
3207 challenges. If an invalid string is ignored, an empty string is placed in
3210 PH/46 Messages that are created by the autoreply transport now contains a
3211 References: header, in accordance with RFCs 2822 and 3834.
3213 PH/47 Added authenticated_sender_force to the smtp transport.
3215 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3217 PH/49 Installed latest patch from the Sieve maintainer.
3219 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3220 was also set, the appendfile transport was unnecessarily scanning a
3221 directory of message files (e.g. for maildir delivery) to find the count
3222 of files (along with the size), even though it did not need this
3223 information. It now does the scan only if it needs to find either the
3224 size of the count of files.
3226 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3228 PH/52 Two bugs concerned with error handling when the smtp transport is
3231 (i) Exim was not creating retry information for temporary errors given
3232 for individual recipients after the DATA command when the smtp transport
3233 was used in LMTP mode. This meant that they could be retried too
3234 frequently, and not timed out correctly.
3236 (ii) Exim was setting the flag that allows error details to be returned
3237 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3238 recipients that were returned after the DATA command.
3240 PH/53 This is related to PH/52, but is more general: for any failing address,
3241 when detailed error information was permitted to be returned to the
3242 sender, but the error was temporary, then after the final timeout, only
3243 "retry timeout exceeded" was returned. Now it returns the full error as
3244 well as "retry timeout exceeded".
3246 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3247 do this, and (what is worse) MTAs that accept it.
3249 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3250 will now be deprecated.
3252 PH/56 New os.c-cygwin from the Cygwin maintainer.
3254 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3255 formats (previously only available in exim formats via -bpr, -bpru,
3256 and -bpra. Now also available in native and exiqgrep formats)
3258 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3259 with very large, slow to parse queues
3261 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3263 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3265 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3266 responses to authentication challenges, though it was showing the
3267 challenges; (ii) I've removed the CR characters from the debug output for
3270 PH/58 Allow for the insertion of a newline as well as a space when a string
3271 is turned into more than one encoded-word during RFC 2047 encoding. The
3272 Sieve code now uses this.
3274 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3275 data_4xx, lost_connection, tls_required.
3277 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3278 message was being sent as an SMTP response.
3280 PH/61 Add -l and -k options to exicyclog.
3282 PH/62 When verifying, if an address was redirected to one new address, so that
3283 verification continued, and the new address failed or deferred after
3284 having set something in $address_data, the value of $address_data was not
3285 passed back to the ACL. This was different to the case when no
3286 redirection occurred. The value is now passed back in both cases.
3288 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3289 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3290 use login_cap.h, so on its own it isn't the distinguishing feature. The
3291 new name refers directly to the setclassresources() function.
3293 PH/65 Added configuration files for NetBSD3.
3295 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3297 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3298 is preferred over IPv4.
3300 PH/68 The bounce_return_message and bounce_return_body options were not being
3301 honoured for bounces generated during the reception of non-SMTP messages.
3302 In particular, this applied to messages rejected by the ACL. This bug has
3303 been fixed. However, if bounce_return_message is true and bounce_return_
3304 body is false, the headers that are returned for a non-SMTP message
3305 include only those that have been read before the error was detected.
3306 (In the case of an ACL rejection, they have all been read.)
3308 PH/69 The HTML version of the specification is now built in a directory called
3309 spec_html instead of spec.html, because the latter looks like a path with
3310 a MIME-type, and this confuses some software.
3312 PH/70 Catch two compiler warnings in sieve.c.
3314 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3315 function verify_get_ident() calls ip_connect() to connect a socket, but
3316 if the "connect()" function timed out, ip_connect() used to close the
3317 socket. However, verify_get_ident() also closes the socket later, and in
3318 between Exim writes to the log, which may get opened at this point. When
3319 the socket was closed in ip_connect(), the log could get the same file
3320 descriptor number as the socket. This naturally causes chaos. The fix is
3321 not to close the socket in ip_connect(); the socket should be closed by
3322 the function that creates it. There was only one place in the code where
3323 this was missing, in the iplookup router, which I don't think anybody now
3324 uses, but I've fixed it anyway.
3326 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3327 well as to direct DNS lookups. Otherwise the handling of names in host
3328 lists is inconsistent and therefore confusing.
3334 PH/01 Two changes to the default runtime configuration:
3336 (1) Move the checks for relay_from_hosts and authenticated clients from
3337 after to before the (commented out) DNS black list checks.
3339 (2) Add control=submission to the relay_from_hosts and authenticated
3340 clients checks, on the grounds that messages accepted by these
3341 statements are most likely to be submissions.
3343 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3345 (1) Generate an error if the third argument for the ${prvs expansion is
3348 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3351 (3) Reset the variables that are obtained from the first argument of
3352 ${prvscheck and used in the second argument before leaving the code,
3353 because their memory is reclaimed, so using them afterwards may do
3356 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3357 one (it's much easier than Tom thought :-).
3359 (5) Because of (4), we can now allow for the use of $prvscheck_result
3360 inside the third argument.
3362 PH/03 For some reason, the default setting of PATH when running a command from
3363 a pipe transport was just "/usr/bin". I have changed it to
3366 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3367 anything to be listed in the output from -bV.
3369 PH/05 When a filter generated an autoreply, the entire To: header line was
3370 quoted in the delivery log line, like this:
3372 => >A.N.Other <ano@some.domain> <original@ddress> ...
3374 This has been changed so that it extracts the operative address. There
3375 may be more than one such address. If so, they are comma-separated, like
3378 => >ano@some.domain,ona@other.domain <original@ddress> ...
3380 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3381 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3382 not being looked up in the rDNS to get a host name, Exim was showing the
3383 IP address twice in Received: lines, even though the IP addresses were
3384 identical. For example:
3386 Received: from [1.2.3.4] (helo=[1.2.3.4])
3388 However, if the real host name was known, it was omitting the HELO data
3389 if it matched the actual IP address. This has been tidied up so that it
3390 doesn't show the same IP address twice.
3392 PH/07 When both +timestamp and +memory debugging was on, the value given by
3393 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3394 called by the debug printing, thereby overwriting the timestamp buffer.
3395 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3397 PH/08 When the original message was included in an autoreply transport, it
3398 always said "this is a copy of the message, including all the headers",
3399 even if body_only or headers_only was set. It now gives an appropriate
3402 PH/09 Applied a patch from the Sieve maintainer which:
3404 o fixes some comments
3405 o adds the (disabled) notify extension core
3406 o adds some debug output for the result of if/elsif tests
3407 o points to the current vacation draft in the documentation
3408 and documents the missing references header update
3412 o fixes a bug in processing the envelope test (when testing
3413 multiple envelope elements, the last element determined the
3416 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3417 Electronic Mail") by including:
3419 Auto-submitted: auto-generated
3421 in the messages that it generates (bounce messages and others, such as
3422 warnings). In the case of bounce messages for non-SMTP messages, there was
3423 also a typo: it was using "Auto_submitted" (underscore instead of
3424 hyphen). Since every message generated by Exim is necessarily in response
3425 to another message, thes have all been changed to:
3427 Auto-Submitted: auto-replied
3429 in accordance with these statements in the RFC:
3431 The auto-replied keyword:
3433 - SHOULD be used on messages sent in direct response to another
3434 message by an automatic process,
3436 - MUST NOT be used on manually-generated messages,
3438 - MAY be used on Delivery Status Notifications (DSNs) and Message
3439 Disposition Notifications (MDNs),
3441 - MUST NOT be used on messages generated by automatic or periodic
3442 processes, except for messages which are automatic responses to
3445 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3446 to the default Received: header definition.
3448 PH/12 Added log selector acl_warn_skipped (default on).
3450 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3451 variables because (a) they are in the wrong storage pool and (b) even if
3452 they were copied, it wouldn't work properly because of the caching.
3454 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3455 checking when decoding. Apparently there are clients that generate
3456 overlong encoded strings. Why am I not surprised?
3458 PH/15 If the first argument of "${if match_address" was not empty, but did not
3459 contain an "@" character, Exim crashed. Now it writes a panic log message
3460 and treats the condition as false.
3462 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3464 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3465 extension "envelope-auth". The code is finished and in agreement with
3466 other implementations, but there is no documentation so far and in fact,
3467 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3468 not changing the active code.
3470 Print executed "if" and "elsif" statements when debugging is used. This
3471 helps a great deal to understand what a filter does.
3473 Document more things not specified clearly in RFC3028. I had all this
3474 sorted out, when out of a sudden new issues came to my mind. Oops."
3476 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3479 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3480 canonical form (as documented). However, after a host name lookup from
3481 the IP address, check_host() was doing a simple string comparison with
3482 addresses acquired from the DNS when checking that the found name did
3483 have the original IP as one of its addresses. Since any found IPv6
3484 addresses are likely to be in abbreviated form, the comparison could
3485 fail. Luckily, there already exists a function for doing the comparison
3486 by converting both addresses to binary, so now that is used instead of
3487 the text comparison.
3489 PH/20 There was another similar case to PH/19, when a complete host name was
3490 given in a host list; looking up its IP address could give an abbreviated
3491 form, whereas the current host's name might or might not be abbreviated.
3492 The same fix has been applied.
3498 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3499 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3502 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3503 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3505 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3507 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3508 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3509 Also, for "Auto-Submitted", treat anything other than "no" as
3510 non-personal, in accordance with RFC 3834. (Previously it treated
3511 anything starting "auto-" as non-personal.)
3513 TF/01 The control=submission/name=... option had a problem with syntax
3514 errors if the name included a slash character. The /name= option
3515 now slurps the rest of the string, so it can include any characters
3516 but it must come last in the list of options (after /sender_retain
3519 PH/05 Some modifications to the interface to the fake nameserver for the new
3527 TK/01 Added the "success_on_redirect" address verification option. See
3528 NewStuff for rationale and an example.
3530 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3532 PH/02 Patch to exigrep to allow it to work on syslog lines.
3534 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3535 fread() to skip over the body file's header line, because in Cygwin the
3536 header line is locked and is inaccessible.
3538 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3539 co-exist for some time) to make it clear that it is the Exim ID that is
3540 referenced, not the Message-ID: header line.
3542 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3543 string_format() function, because snprintf() does not exist on all
3546 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3547 ${stat: expansion item.
3549 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3550 protocol synchronization error", to keep the pedants happy.
3552 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3553 well as for IRIX systems, when gcc is being used. See the host.c source
3556 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3558 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3561 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3562 transport and to the smtp transport in LMTP mode.
3564 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3566 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3567 run a filter in a subprocess. This could lead to confusion in subsequent
3568 lookups in the parent process. There should also be a search_tidyup() at
3569 the end of the subprocess.
3571 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3572 only if the host matched helo_try_verify_hosts, which caused the
3573 verification to occur when the EHLO/HELO command was issued. The ACL just
3574 tested the remembered result. Now, if a previous verification attempt has
3575 not happened, "verify = helo" does it there and then.
3577 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3579 TK/03 Fix log output including CR from clamd.
3581 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3582 exist provoked a memory error which could cause a segfault.
3584 PH/15 Installed PCRE 6.2
3586 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3588 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3589 of the problem. Specifically, suggested +O2 rather than +O1 for the
3592 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3594 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3595 record, the port was not correctly logged when the outgoing_port log
3596 selector was set (it logged the transort's default port).
3598 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3599 fallback_hosts, and "hosts" in the smtp transport.
3601 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3602 host errors such as "Connection refused".
3604 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3605 authentication with radiusclient 0.4.9:
3607 - Error returned from rc_read_config was caught wrongly
3608 - Username/password not passed on to radius server due to wrong length.
3610 The presumption is that some radiusclient API changes for 4.51/PH/17
3611 were not taken care of correctly. The code is still untested by me (my
3612 Linux distribution still has 0.3.2 of radiusclient), but it was
3613 contributed by a Radius user.
3615 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3616 expanding the "port" option of the smtp transport.
3618 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3619 while reading a MIME header. Thanks to Tom Hughes for a patch.
3621 PH/24 Include config.h inside local_scan.h so that configuration settings are
3624 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3625 that in an ACL for RCPT (for example), you can examine exactly what was
3628 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3629 commands, but it was not correctly comparing the address with the actual
3630 client host address. Thus, it would show the EHLO address in Received:
3631 header lines when this was not necessary.
3633 PH/27 Added the % operator to ${eval:}.
3635 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3636 it should be ignoring failures (because with -C, for example, it has lost
3637 privilege). It wasn't ignoring creation failures other than "already
3640 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3643 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3644 because an input line was too long, either on its own, or by virtue of
3645 too many continuations, the temporary file was not being removed, and the
3646 return code was incorrect.
3648 PH/31 Missing "BOOL" in function definition in filtertest.c.
3650 PH/32 Applied Sieve patches from the maintainer.
3652 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3654 PH/33 Added "verify = not_blind".
3656 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3657 Local/Makefile (with some defaults set). These are used in built scripts
3658 such as exicyclog, but they have never been used in the exim_install
3659 script (though there are many overriding facilities there). I have
3660 arranged that the exim_install script now takes note of these two
3663 PH/35 Installed configuration files for Dragonfly.
3665 PH/36 When a locally submitted message by a trusted user did not contain a
3666 From: header, and the sender address was obtained from -f or from an SMTP
3667 MAIL command, and the trusted user did not use -F to supply a sender
3668 name, $originator_name was incorrectly used when constructing a From:
3669 header. Furthermore, $originator_name was used for submission mode
3670 messages from external hosts without From: headers in a similar way,
3671 which is clearly wrong.
3673 PH/37 Added control=suppress_local_fixups.
3675 PH/38 When log_selector = +received_sender was set, and the addition of the
3676 sender made the log line's construction buffer exactly full, or one byte
3677 less than full, an overflow happened when the terminating "\n" was
3680 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3681 when the result of a list match is failure because a DNS lookup failed.
3683 PH/40 RM_COMMAND is now used in the building process.
3685 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3686 the "build-* directories that it finds.
3688 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3689 address in a domain literal was a prefix of an interface address.
3691 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3692 when verifying a sender address, unless rewrite_headers is false.
3694 PH/44 Wrote a long comment about why errors_to addresses are verified as
3695 recipients, not senders.
3697 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3698 the ratelimit ACL was added.
3700 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3702 PH/46 Added extra information about PostgreSQL errors to the error string.
3704 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3705 avoiding the need to install special zones in a real server. This is
3706 backwards compatible; if it can't find the fake resolver, it drops back.
3707 Thus, both old and new test suites can be run.
3709 TF/02 Added util/ratelimit.pl
3711 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3712 clock is set back in time.
3714 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3715 Candler <B.Candler@pobox.com>.
3717 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3718 OK for addresses that are the result of redirections.
3720 PH/48 A number of further additions for the benefit of the new test suite,
3721 including a fake gethostbyname() that interfaces to the fake DNS resolver
3724 TF/06 The fix for widen_domains has also been applied to qualify_single and
3725 search_parents which are the other dnslookup options that can cause
3728 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3731 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3733 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3734 service" if it encountered an MX record with an empty target hostname.
3735 The message is now "an MX or SRV record indicated no SMTP service".
3737 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3738 if the DNS of the sending site is misconfigured. This is quite a
3739 common situation. This change restores the behaviour of treating a
3740 helo verification defer as a failure.
3742 PH/51 If self=fail was set on a router, the bounce message did not include the
3743 actual error message.
3749 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3751 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3752 command itself ended in error, the underlying message about the transport
3753 filter timeout was being overwritten with the pipe command error. Now the
3754 underlying error message should be appended to the second error message.
3756 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3758 PH/02 Reduce the amount of output that "make" produces by default. Full output
3759 can still be requested.
3761 PH/03 The warning log line about a condition test deferring for a "warn" verb
3762 was being output only once per connection, rather than after each
3763 occurrence (because it was using the same function as for successful
3764 "warn" verbs). This seems wrong, so I have changed it.
3766 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3767 it should not have, which might have caused a crash in the right
3768 circumstances, but probably never did.
3770 PH/04 Installed a modified version of Tony Finch's patch to make submission
3771 mode fix the return path as well as the Sender: header line, and to
3772 add a /name= option so that you can make the user's friendly name appear
3775 TF/03 Added the control = fakedefer ACL modifier.
3777 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3778 Mark Lowes for thorough testing.
3780 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3782 TK/03 Merged latest SRS patch from Miles Wilton.
3784 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3785 with the definition in sysexits.h (which is #included earlier).
3786 Fortunately, Exim does not actually use EX_OK. The code used to try to
3787 preserve the sysexits.h value, by assuming that macro definitions were
3788 scanned for macro replacements. I have been disabused of this notion,
3789 so now the code just undefines EX_OK before #including unistd.h.
3791 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3792 in the smtp transport. When a block could not be written in a single
3793 write() function, the timeout was being re-applied to each part-write.
3794 This seems wrong - if the receiver was accepting one byte at a time it
3795 would take for ever. The timeout is now adjusted when this happens. It
3796 doesn't have to be particularly precise.
3798 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3799 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3801 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3802 without a domain if the check to <postmaster@domain> fails.
3804 SC/01 Eximstats: added -xls and the ability to specify output files
3805 (patch written by Frank Heydlauf).
3807 SC/02 Eximstats: use FileHandles for outputting results.
3809 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3811 SC/04 Eximstats: fixed display of large numbers with -nvr option
3813 SC/05 Eximstats: fixed merging of reports with empty tables.
3815 SC/06 Eximstats: added the -include_original_destination flag
3817 SC/07 Eximstats: removed tabs and trailing whitespace.
3819 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3821 TK/06 MBOX spool code: Add real "From " MBOX separator line
3822 so the .eml file is really in mbox format (even though
3823 most programs do not really care). Patch from Alex Miller.
3825 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3826 The latter is generated from $received_to and is only set if the
3827 message has one envelope recipient. SA can use these headers,
3828 obviously out-of-the-box. Patch from Alex Miller.
3830 PH/08 The ${def test on a variable was returning false if the variable's
3831 value was "0", contrary to what the specification has always said!
3832 The result should be true unless the variable is empty.
3834 PH/09 The syntax error of a character other than { following "${if
3835 def:variable_name" (after optional whitespace) was not being diagnosed.
3836 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3837 accidental colon was present, for example, could give incorrect results.
3839 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3840 result is used (not including appendfile, where other changes are about
3843 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3844 This involved changing a lot of size variables from int to off_t. It
3845 should work with maildirs and everything.
3847 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3848 spamd dying while we are connected to it.
3850 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3853 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3854 function for BDB 4.3.
3856 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3858 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3859 the sources for PCRE have been re-arranged and more files are now
3862 PH/15 The code I had for printing potentially long long variables in PH/11
3863 above was not the best (it lost precision). The length of off_t variables
3864 is now inspected at build time, and an appropriate printing format (%ld
3865 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3866 to be "long long int" or "long int". This is needed for the internal
3867 formatting function string_vformat().
3869 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3870 the configuration file to be ":syslog", then the script "guesses" where
3871 the logs files are, rather than using the compiled in default. In our
3872 case the guess is not the same as the compiled default, so the script
3873 suddenly stopped working when I started to use syslog. The patch checks
3874 to see if log_file_path is "". If so, it attempts to read it from exim
3875 with no configuration file to get the compiled in version, before it
3876 falls back to the previous guessing code."
3878 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3879 implementing BATV in an Exim configuration. See NewStuff for the gory
3882 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3883 Makefile that are specific to HP-UX.
3885 PH/18 If the "use_postmaster" option was set for a recipient callout together
3886 with the "random" option, the postmaster address was used as the MAIL
3887 FROM address for the random test, but not for the subsequent recipient
3888 test. It is now used for both.
3890 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3891 patch removes a few documentation additions to RFC 3028, because the
3892 latest draft now contains them. It adds the new en;ascii-case comparator
3893 and a new error check for 8bit text in MIME parts. Comparator and
3894 require names are now matched exactly. I enabled the subaddress
3895 extension, but it is not well tested yet (read: it works for me)."
3897 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3898 rework some of the code of TK/09 above to avoid the hardwired use of
3899 "%lld" and "long long". Replaced the call to snprintf() with a call to
3902 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3903 records point to non-existent hosts", "retry timeout exceeded", and
3904 "retry time not reached for any host after a long failure period".
3906 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3907 experimental DomainKeys support:
3909 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3910 (2) On an error such as an illegally used "control", the wrong name for
3911 the control was given.
3913 These problems did NOT occur unless DomainKeys support was compiled.
3915 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3917 PH/24 Added ${if match_ip condition.
3919 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3920 databases so that it will be absolutely obvious if a crash occurs in the
3921 DB library. This is a regular occurrence (often caused by mis-matched
3924 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3925 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3926 picked up on a user's system that detects such things. There doesn't seem
3927 to be a gcc warning option for this - only an attribute that has to be
3928 put on the function's prototype. It seems that in Fedora Core 4 they have
3929 set this on a number of new functions. No doubt there will be more in due
3932 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3933 specify a transport. However, if an address that was verified by such a
3934 router was the subject of a callout, Exim crashed because it tried to
3935 read the rcpt_include_affixes from the non-existent transport. Now it
3936 just assumes that the setting of that option is false. This bug was
3937 introduced by 4.51/PH/31.
3939 PH/28 Changed -d+all to exclude +memory, because that information is very
3940 rarely of interest, but it makes the output a lot bigger. People tend to
3941 do -d+all out of habit.
3943 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3944 code in os-type was giving problems when libc.so lives in lib64, like on
3947 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3948 aren't the modern standard, and it seems that some systems' include files
3949 don't always have them. Exim was already checking for some of the newer
3950 ones like T_AAAA, and defining it itself. I've added checks for all the
3951 record types that Exim uses.
3953 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3954 not automatically generating a new one, as it is supposed to. This
3955 prevented TLS from working. If the file did exist, but contained invalid
3956 data, a new version was generated, as expected. It was only the case of a
3957 non-existent file that was broken.
3959 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3960 with a change in libdomainkeys > 0.64.
3962 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3963 from DNS. If the selector record carries the flag, it now has
3964 precedence over the domain-wide flag.
3966 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3968 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3969 the use of a port name that isn't defined in /etc/services) occurred, the
3970 message was deferred as in a normal delivery, and thus remained on the
3971 spool, instead of being failed because of the mua_wrapper setting. This
3972 is now fixed, and I tidied up some of the mua_wrapper messages at the
3975 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3976 the messages in a hash of arrays rather than using individual hashes.
3977 This is a bit cleaner and results in dramatic memory savings, albeit
3978 at a slight CPU cost.
3980 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3981 as requested by Marc Sherman.
3983 SC/10 Eximstats: added histograms for user specified patterns as requested
3986 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3988 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3989 fopen() in the content-scanning modules that did not already have it.
3995 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3996 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3998 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4000 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4002 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4003 to test Sieve filters that use "vacation".
4005 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4006 that changes the way the GnuTLS parameters are stored in the cache file.
4007 The new format can be generated externally. For backward compatibility,
4008 if the data in the cache doesn't make sense, Exim assumes it has read an
4009 old-format file, and it generates new data and writes a new file. This
4010 means that you can't go back to an older release without removing the
4013 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4014 work if there are any delivery delays because "one_time" forces the
4015 parent to be marked "delivered", so its unseen clone is never tried
4016 again. For this reason, Exim now forbids the simultaneous setting of
4019 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4020 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4021 entry if you want to know the details. The fix, however, neglected to
4022 consider the case when local delivery batching is involved. The test for
4023 "previously delivered" was not happening when checking to see if an
4024 address could be batched with a previous (undelivered) one; under
4025 certain circumstances this could lead to multiple deliveries to the same
4028 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4029 in its include files, and this causes problems building Exim.
4031 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4032 header_syntax) but Exim was just ignoring anything given after a slash.
4033 In particular, this caused confusion with an attempt to use "verify =
4034 reverse_host_lookup/defer_ok". An error is now given when options are
4035 supplied for verify items that do not have them. (Maybe reverse_host_
4036 lookup should have a defer_ok option, but that's a different point.)
4038 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4039 defined by RFC 821) to 2048, because there were problems with some AUTH
4040 commands, and RFC 1869 says the size should be increased for extended
4041 SMTP commands that take arguments.
4043 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4046 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4047 "unknown" error; now it says that the functionality isn't in the binary.
4049 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4050 an address' error message when a string expansion fails (syntax or
4051 whatever). Otherwise the password may appear in the log. Following change
4052 PH/42 below, there is no longer a chance of it appearing in a bounce
4055 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4057 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4058 of its addresses was ever tried. (Bugzilla bug #2.)
4060 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4061 the result incorrectly in the debug output. (It correctly added a newline
4062 to what was transported.)
4064 TF/01 Added $received_time.
4066 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4067 commented out examples of how to interface to a virus scanner and to
4068 SpamAssassin. Also added commented examples of av_scanner and
4069 spamd_address settings.
4071 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4072 and controls are allowed in which ACLs. There were a couple of minor
4073 errors. Some of the entries in the conditions table (which is a table of
4074 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4075 negation of where the condition IS allowed.
4077 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4079 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4080 header file does not have a version number, so I've had to invent a new
4081 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4082 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4083 radiusclient), but it was contributed by a Radius user.
4085 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4086 files or directories, for OpenSSL.
4088 PH/19 When an Exim process that is running as root has to create an Exim log
4089 file, it does so in a subprocess that runs as exim:exim so as to get the
4090 ownership right at creation (otherwise, other Exim processes might see
4091 the file with the wrong ownership). There was no test for failure of this
4092 fork() call, which would lead to the process getting stuck as it waited
4093 for a non-existent subprocess. Forks do occasionally fail when resources
4094 run out. I reviewed all the other calls to fork(); they all seem to check
4097 PH/20 When checking for unexpected SMTP input at connect time (before writing
4098 the banner), Exim was not dealing correctly with a non-positive return
4099 from the read() function. If the client had disconnected by this time,
4100 the result was a log entry for a synchronization error with an empty
4101 string after "input=" when read() returned zero. If read() returned -1
4102 (an event I could not check), uninitialized data bytes were printed.
4103 There were reports of junk text (parts of files, etc) appearing after
4106 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4108 PH/22 Added support for macro redefinition, and (re)definition in between
4109 driver and ACL definitions.
4111 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4112 forgetting to use the resulting value; it was using the unexpanded value.
4114 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4115 hadn't been configured. The fix is from Juergen Kreileder, who
4116 understands it better than I do:
4118 "Here's what I see happening with three configured cyrus_sasl
4119 authenticators configured (plain, login, cram-md5):
4121 On startup auth_cyrus_sasl_init() gets called for each of these.
4122 This means three calls to sasl_listmech() without a specified mech_list.
4123 => SASL tests which mechs of all available mechs actually work
4124 => three warnings about OTP not working
4125 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4127 With the patch, sasl_listmech() also gets called three times. But now
4128 SASL's mech_list option is set to the server_mech specified in the the
4129 authenticator. Or in other words, the answer from sasl_listmech()
4130 gets limited to just the mech you're testing for (which is different
4132 => the return list contains just 'plain' or 'login', 'cram-md5' or
4133 nothing depending on the value of ob->server_mech.
4135 I've just tested the patch: Authentication still works fine,
4136 unavailable mechs specified in the exim configuration are still
4137 caught, and the auth.log warnings about OTP are gone."
4139 PH/25 When debugging is enabled, the contents of the command line are added
4140 to the debugging output, even when log_selector=+arguments is not
4143 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4144 answer is "GNU", and only if the return is "GNU/something" is the answer
4147 PH/27 $acl_verify_message is now set immediately after the failure of a
4148 verification in an ACL, and so is available in subsequent modifiers. In
4149 particular, the message can be preserved by coding like this:
4151 warn !verify = sender
4152 set acl_m0 = $acl_verify_message
4154 Previously, $acl_verify_message was set only while expanding "message"
4155 and "log_message" when a very denied access.
4157 PH/28 Modified OS/os.c-Linux with
4159 -#ifndef OS_LOAD_AVERAGE
4160 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4162 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4163 nomenclature these days.)
4165 PH/29 Installed patch from the Sieve maintainer that adds the options
4166 sieve_useraddress and sieve_subaddress to the redirect router.
4168 PH/30 In these circumstances:
4169 . Two addresses routed to the same list of hosts;
4170 . First host does not offer TLS;
4171 . First host accepts first address;
4172 . First host gives temporary error to second address;
4173 . Second host offers TLS and a TLS session is established;
4174 . Second host accepts second address.
4175 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4176 and peerdn, if requested) that were in fact used only for the second
4179 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4180 attention to any local part prefix or suffix that was matched by the
4181 router that accepted the address. It now behaves in the same way as it
4182 does for delivery: the affixes are removed from the local part unless
4183 rcpt_include_affixes is set on the transport.
4185 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4186 timeout during the DATA phase of an incoming message.
4188 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4189 applied a patch sanctioned by the Sieve maintainer.
4191 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4192 the uid or gid is negative. A case of a negative gid caused this to be
4193 noticed. The fix allows for either to be negative.
4195 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4196 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4199 PH/36 The change PH/12 above was broken. Fixed it.
4201 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4202 the grounds that routing the same address twice would always produce the
4203 same answer. This might have been true once, but it is certainly no
4204 longer true now. Routing a child address may depend on the previous
4205 routing that produced that child. Some complicated redirection strategies
4206 went wrong when messages had multiple recipients, and made Exim's
4207 behaviour dependent on the order in which the addresses were given.
4209 I have moved the duplicate checking until after the routing is complete.
4210 Exim scans the addresses that are assigned to local and remote
4211 transports, and removes any duplicates. This means that more work will be
4212 done, as duplicates will always all be routed, but duplicates are
4213 presumably rare, so I don't expect this is of any significance.
4215 For deliveries to pipes, files, and autoreplies, the duplicate checking
4216 still happens during the routing process, since they are not going to be
4219 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4220 It corrects a timeout issue with spamd. This is Ian's comment: "The
4221 background is that sometimes spamd either never reads data from a
4222 connection it has accepted, or it never writes response data. The exiscan
4223 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4224 blindly assumes that writes won't block so it may never time out."
4226 PH/39 Allow G after quota size as well as K and M.
4228 PH/40 The value set for $authenticated_id in an authenticator may not contain
4229 binary zeroes or newlines because the value is written to log lines and
4230 to spool files. There was no check on this. Now the value is run through
4231 the string_printing() function so that such characters are converted to
4232 printable escape sequences.
4234 PH/41 $message_linecount is a new variable that contains the total number of
4235 lines in the message. Compare $body_linecount, which is the count for the
4238 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4239 bounce and delay warning messages, except in certain special cases, which
4242 (a) An SMTP error message from a remote host;
4243 (b) A message specified in a :fail: redirection;
4244 (c) A message specified in a "fail" command in a system filter;
4245 (d) A message specified in a FAIL return from the queryprogram router;
4246 (e) A message specified by the cannot_route_message router option.
4248 In these cases only, Exim does include the error details in bounce and
4249 warning messages. There are also a few cases where bland messages such
4250 as "unrouteable address" or "local delivery error" are given.
4252 PH/43 $value is now also set for the "else" part of a ${run expansion.
4254 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4255 being worked on, but at least Exim now implements the latest version to
4258 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4259 process to complete was treated as a delivery failure, a timeout while
4260 writing the message to the pipe was logged, but erroneously treated as a
4261 successful delivery. Such timeouts include transport filter timeouts. For
4262 consistency with the overall process timeout, these timeouts are now
4263 treated as errors, giving rise to delivery failures by default. However,
4264 there is now a new Boolean option for the pipe transport called
4265 timeout_defer, which, if set TRUE, converts the failures into defers for
4266 both kinds of timeout. A transport filter timeout is now identified in
4269 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4270 systems where "make" and "gmake" are different, calling "gmake" at top
4271 level broke things. I've arranged for the value of $(MAKE) to be passed
4272 from the Makefile to this script so that it can call the same version of
4276 A note about Exim versions 4.44 and 4.50
4277 ----------------------------------------
4279 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4280 changes of various kinds. As a consequence, a big documentation update was
4281 needed. This delayed the release for rather longer than seemed good, especially
4282 in the light of a couple of (minor) security issues. Therefore, the changes
4283 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4284 release. So 4.44 and 4.50 are in effect two different branches that both start
4287 I have left the 4.50 change log unchanged; it contains all the changes since
4288 4.43. The change log for 4.44 is below; many of its items are identical to
4289 those for 4.50. This seems to be the most sensible way to preserve the
4290 historical information.
4296 1. Minor wording change to the doc/README.SIEVE file.
4298 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4299 computation of the current number of files was incorrect.
4301 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4302 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4305 4. Give more explanation in the error message when the command for a transport
4306 filter fails to execute.
4308 5. There are several places where Exim runs a non-Exim command in a
4309 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4310 was being done only for the command run by the queryprogram router. It is
4311 now done for all such subprocesses. The other cases are: ${run, transport
4312 filters, and the commands run by the lmtp and pipe transports.
4314 6. Added CONFIGURE_GROUP build-time option.
4316 7. Some older OS have a limit of 256 on the maximum number of file
4317 descriptors. Exim was using setrlimit() to set 1000 as a large value
4318 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4319 systems. I've change it so that if it can't get 1000, it tries for 256.
4321 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4322 was an oversight, and furthermore, ever since the addition of extra
4323 controls (e.g. 4.43/32), the checks on when to allow different forms of
4324 "control" were broken. There should now be diagnostics for all cases when a
4325 control that does not make sense is encountered.
4327 9. Added the /retain_sender option to "control=submission".
4329 10. $recipients is now available in the predata ACL (oversight).
4331 11. Tidy the search cache before the fork to do a delivery from a message
4332 received from the command line. Otherwise the child will trigger a lookup
4333 failure and thereby defer the delivery if it tries to use (for example) a
4334 cached ldap connection that the parent has called unbind on.
4336 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4337 of $address_data from the recipient verification was clobbered by the
4338 sender verification.
4340 13. The value of address_data from a sender verification is now available in
4341 $sender_address_data in subsequent conditions in the ACL statement.
4343 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4345 15. Added a new option "connect=<time>" to callout options, to set a different
4348 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4349 was its contents. (It was OK if the option was not defined at all.)
4351 17. A "Completed" log line is now written for messages that are removed from
4352 the spool by the -Mrm option.
4354 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4355 information about exactly what failed.
4357 19. Added -dd to debug only the daemon process.
4359 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4360 handles timeouts, both on the server side and network timeouts. Renamed the
4361 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4363 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4364 It is now set to "smtps".
4366 22. $host_address is now set to the target address during the checking of
4367 ignore_target_hosts.
4369 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4370 being passed; this would have caused $sender_host_name to have been used if
4371 matching the list had actually called for a host name (not very likely,
4372 since this list is usually IP addresses). A host name is now passed as
4375 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4376 code that specifies a non-restarting handler (typically sigaction() in
4377 modern systems) in an attempt to fix a rare and obscure crash bug.
4379 25. Narrowed the window for a race in the daemon that could cause it to ignore
4380 SIGCHLD signals. This is not a major problem, because they are used only to
4381 wake it up if nothing else does.
4383 26. A malformed maildirsize file could cause Exim to calculate negative values
4384 for the mailbox size or file count. Odd effects could occur as a result.
4385 The maildirsize information is now recalculated if the size or filecount
4388 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4389 support for a long time. Removed HAVE_SYS_VFS_H.
4391 28. Installed the latest version of exipick from John Jetmore.
4393 29. In an address list, if the pattern was not a regular expression, an empty
4394 subject address (from a bounce message) matched only if the pattern was an
4395 empty string. Non-empty patterns were not even tested. This was the wrong
4396 because it is perfectly reasonable to use an empty address as part of a
4397 database query. An empty address is now tested by patterns that are
4398 lookups. However, all the other forms of pattern expect the subject to
4399 contain a local part and a domain, and therefore, for them, an empty
4400 address still always fails if the pattern is not itself empty.
4402 30. Exim went into a mad DNS loop when attempting to do a callout where the
4403 host was specified on an smtp transport, and looking it up yielded more
4404 than one IP address.
4406 31. Re-factored the code for checking spool and log partition space into a
4407 function that finds that data and another that does the check. The former
4408 is then used to implement four new variables: $spool_space, $log_space,
4409 $spool_inodes, and $log_inodes.
4411 32. The RFC2047 encoding function was originally intended for short strings
4412 such as real names; it was not keeping to the 75-character limit for
4413 encoded words that the RFC imposes. It now respects the limit, and
4414 generates multiple encoded words if necessary. To be on the safe side, I
4415 have increased the buffer size for the ${rfc2047: expansion operator from
4418 33. It is now permitted to omit both strings after an "if" condition; if the
4419 condition is true, the result is "true". As before, when the second string
4420 is omitted, a false condition yields an empty string. This makes it less
4421 cumbersome to write custom ACL and router conditions.
4423 34. Failure to deliver a bounce message always caused it to be frozen, even if
4424 there was an errors_to setting on the router. The errors_to setting is now
4427 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4428 canonical form (fully expanded) before being placed in
4429 $sender_host_address.
4431 36. The table in the code that translates DNS record types into text (T_A to
4432 "A" for instance) was missing entries for NS and CNAME. It is just possible
4433 that this could have caused confusion if both these types were looked up
4434 for the same domain, because the text type is used as part of Exim's
4435 per-process caching. But the chance of anyone hitting this buglet seems
4438 37. The dnsdb lookup has been extended in a number of ways.
4440 (1) There is a new type, "zns", which walks up the domain tree until it
4441 finds some nameserver records. It should be used with care.
4443 (2) There is a new type, "mxh", which is like "mx" except that it returns
4444 just the host names, not the priorities.
4446 (3) It is now possible to give a list of domains (or IP addresses) to be
4447 looked up. The behaviour when one of the lookups defers can be
4448 controlled by a keyword.
4450 (4) It is now possible to specify the separator character for use when
4451 multiple records are returned.
4453 38. The dnslists ACL condition has been extended: it is now possible to supply
4454 a list of IP addresses and/or domains to be looked up in a particular DNS
4457 39. Added log_selector=+queue_time_overall.
4459 40. When running the queue in the test harness, wait just a tad after forking a
4460 delivery process, to get repeatability of debugging output.
4462 41. Include certificate and key file names in error message when GnuTLS fails
4463 to set them up, because the GnuTLS error message doesn't include the name
4464 of the failing file when there is a problem reading it.
4466 42. Allow both -bf and -bF in the same test run.
4468 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4470 44. The "Exiscan patch" is now merged into the mainline Exim source.
4472 45. Sometimes the final signoff response after QUIT could fail to get
4473 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4474 < 0 before doing a fflush(). This bug looks as though it goes back to the
4475 introduction of TLS in release 3.20, but "sometimes" must have been rare
4476 because the tests only now provoked it.
4478 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4479 (this can affect the format of dates).
4481 47. exim_tidydb, when checking for the continued existence of a message for
4482 which it has found a message-specific retry record, was not finding
4483 messages that were in split spool directories. Consequently, it was
4484 deleting retry records that should have stayed in existence.
4486 48. Steve fixed some bugs in eximstats.
4488 49. The SPA authentication driver was not abandoning authentication and moving
4489 on to the next authenticator when an expansion was forced to fail,
4490 contradicting the general specification for all authenticators. Instead it
4491 was generating a temporary error. It now behaves as specified.
4493 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4494 (the order specifies the preference for clients). The order is now AES256,
4495 AES128, 3DES, ARCFOUR128.
4497 51. Small patch to Sieve code - explicitly set From: when generating an
4500 52. Exim crashed if a remote delivery caused a very long error message to be
4501 recorded - for instance if somebody sent an entire SpamAssassin report back
4502 as a large number of 550 error lines. This bug was coincidentally fixed by
4503 increasing the size of one of Exim's internal buffers (big_buffer) that
4504 happened as part of the Exiscan merge. However, to be on the safe side, I
4505 have made the code more robust (and fixed the comments that describe what
4508 53. Now that there can be additional text after "Completed" in log lines (if
4509 the queue_time_overall log selector is set), a one-byte patch to exigrep
4510 was needed to allow it to recognize "Completed" as not the last thing in
4513 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4514 patch that reportedly fixes this has been added. I am not expert enough to
4515 create a test for it. This is what the patch creator wrote:
4517 "I found a little strange behaviour of ldap code when working with
4518 Windows 2003 AD Domain, where users was placed in more than one
4519 Organization Units. When I tried to give exim partial DN, the exit code
4520 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4521 But simultaneously result of request was absolutely normal ldap result,
4522 so I produce this patch..."
4524 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4525 REFERENCE, so I have modified the code to exclude the patch when that macro
4528 55. Some experimental protocols are using DNS PTR records for new purposes. The
4529 keys for these records are domain names, not reversed IP addresses. The
4530 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4531 leaves it alone. Component reversal etc. now happens only for IP addresses.
4534 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4536 57. Double the size of the debug message buffer (to 2048) so that more of very
4537 long debug lines gets shown.
4539 58. The exicyclog utility now does better if the number of log files to keep
4540 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4542 59. Two changes related to the smtp_active_hostname option:
4544 (1) $smtp_active_hostname is now available as a variable.
4545 (2) The default for smtp_banner uses $smtp_active_hostname instead
4546 of $primary_hostname.
4548 60. The host_aton() function is supposed to be passed a string that is known
4549 to be a valid IP address. However, in the case of IPv6 addresses, it was
4550 not checking this. This is a hostage to fortune. Exim now panics and dies
4551 if the condition is not met. A case was found where this could be provoked
4552 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4553 components; fortuitously, this particular loophole had already been fixed
4554 by change 4.50/55 above.
4556 If there are any other similar loopholes, the new check in host_aton()
4557 itself should stop them being exploited. The report I received stated that
4558 data on the command line could provoke the exploit when Exim was running as
4559 exim, but did not say which command line option was involved. All I could
4560 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4561 running as the user.
4564 61. There was a buffer overflow vulnerability in the SPA authentication code
4565 (which came originally from the Samba project). I have added a test to the
4566 spa_base64_to_bits() function which I hope fixes it.
4569 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4570 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4572 63. The daemon start-up calls getloadavg() while still root for those OS that
4573 need the first call to be done as root, but it missed one case: when
4574 deliver_queue_load_max is set with deliver_drop_privilege. This is
4575 necessary for the benefit of the queue runner, because there is no re-exec
4576 when deliver_drop_privilege is set.
4578 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4579 This has been fixed.
4581 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4582 was in use, was not putting the data itself into the right store pool;
4583 consequently, it could be overwritten for a subsequent message in the same
4584 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4587 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4589 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4590 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4591 the function were treating the return as a boolean value, which happened to
4592 work because 0=false and not-0=true, but is not correct code.
4594 68. The host_aton() function was not handling scoped IPv6 addresses (those
4595 with, for example, "%eth0" on the end) correctly.
4597 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4598 negated items (that is, ~something) in unsigned ints. Some compilers
4599 apparently mutter when there is no cast.
4601 70. If an address verification called from an ACL failed, and did not produce a
4602 user-specific message (i.e. there was only a "system" message), nothing was
4603 put in $acl_verify_message. In this situation, it now puts the system
4606 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4607 session; change 4.31/43 added the unwanted input to the log line - except
4608 that it did not do this in the start of session case. It now does.
4610 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4611 This is wrong and can cause the other end to generate a synchronization
4612 error if it is another Exim or anything else that does the synchronization
4613 check. A QUIT command is no longer sent after a timeout.
4615 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4616 during host lookups.
4618 74. The defer_ok option of callout verification was not working if it was used
4619 when verifying addresses in header lines, that is, for this case:
4621 verify = header_sender/callout=defer_ok
4623 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4624 those file descriptors could be used for SMTP connections. If anything
4625 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4626 could be sent to the SMTP client, causing chaos. The daemon now opens
4627 stdin, stdout, and stderr to /dev/null when it puts itself into the
4630 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4631 log by default. The user can override this with suitable Perl magic.
4633 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4634 the log message when discard triggers, was not working for the DATA ACL or
4635 for the non-SMTP ACL.
4637 78. Error message wording change in sieve.c.
4639 79. If smtp_accept_max_per_host was set, the number of connections could be
4640 restricted to fewer than expected, because the daemon was trying to set up
4641 a new connection before checking whether the processes handling previous
4642 connections had finished. The check for completed processes is now done
4643 earlier. On busy systems, this bug wouldn't be noticed because something
4644 else would have woken the daemon, and it would have reaped the completed
4647 80. If a message was submitted locally by a user whose login name contained one
4648 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4649 It caused a spool format error. I have fixed the spool reading code. A
4650 related problem was that the "from" clause in the Received: line became
4651 illegal because of the space(s). It is now covered by ${quote_local_part.
4653 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4656 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4658 83. Patch from Sieve maintainer to fix unterminated string problem in
4659 "vacation" handling.
4661 84. Some minor changes to the Linux configuration files to help with other
4662 OS variants using glibc.
4664 85. One more patch for Sieve to update vacation handling to latest spec.
4667 ----------------------------------------------------
4668 See the note above about the 4.44 and 4.50 releases.
4669 ----------------------------------------------------
4675 1. Change 4.43/35 introduced a bug that caused file counts to be
4676 incorrectly computed when quota_filecount was set in an appendfile
4679 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4680 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4683 3. Give more explanation in the error message when the command for a transport
4684 filter fails to execute.
4686 4. There are several places where Exim runs a non-Exim command in a
4687 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4688 was being done only for the command run by the queryprogram router. It is
4689 now done for all such subprocesses. The other cases are: ${run, transport
4690 filters, and the commands run by the lmtp and pipe transports.
4692 5. Some older OS have a limit of 256 on the maximum number of file
4693 descriptors. Exim was using setrlimit() to set 1000 as a large value
4694 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4695 systems. I've change it so that if it can't get 1000, it tries for 256.
4697 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4698 was an oversight, and furthermore, ever since the addition of extra
4699 controls (e.g. 4.43/32), the checks on when to allow different forms of
4700 "control" were broken. There should now be diagnostics for all cases when a
4701 control that does not make sense is encountered.
4703 7. $recipients is now available in the predata ACL (oversight).
4705 8. Tidy the search cache before the fork to do a delivery from a message
4706 received from the command line. Otherwise the child will trigger a lookup
4707 failure and thereby defer the delivery if it tries to use (for example) a
4708 cached ldap connection that the parent has called unbind on.
4710 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4711 of $address_data from the recipient verification was clobbered by the
4712 sender verification.
4714 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4715 was its contents. (It was OK if the option was not defined at all.)
4717 11. A "Completed" log line is now written for messages that are removed from
4718 the spool by the -Mrm option.
4720 12. $host_address is now set to the target address during the checking of
4721 ignore_target_hosts.
4723 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4724 being passed; this would have caused $sender_host_name to have been used if
4725 matching the list had actually called for a host name (not very likely,
4726 since this list is usually IP addresses). A host name is now passed as
4729 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4730 code that specifies a non-restarting handler (typically sigaction() in
4731 modern systems) in an attempt to fix a rare and obscure crash bug.
4733 15. Narrowed the window for a race in the daemon that could cause it to ignore
4734 SIGCHLD signals. This is not a major problem, because they are used only to
4735 wake it up if nothing else does.
4737 16. A malformed maildirsize file could cause Exim to calculate negative values
4738 for the mailbox size or file count. Odd effects could occur as a result.
4739 The maildirsize information is now recalculated if the size or filecount
4742 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4743 support for a long time. Removed HAVE_SYS_VFS_H.
4745 18. Updated exipick to current release from John Jetmore.
4747 19. Allow an empty sender to be matched against a lookup in an address list.
4748 Previously the only cases considered were a regular expression, or an
4751 20. Exim went into a mad DNS lookup loop when doing a callout where the
4752 host was specified on the transport, if the DNS lookup yielded more than
4755 21. The RFC2047 encoding function was originally intended for short strings
4756 such as real names; it was not keeping to the 75-character limit for
4757 encoded words that the RFC imposes. It now respects the limit, and
4758 generates multiple encoded words if necessary. To be on the safe side, I
4759 have increased the buffer size for the ${rfc2047: expansion operator from
4762 22. Failure to deliver a bounce message always caused it to be frozen, even if
4763 there was an errors_to setting on the router. The errors_to setting is now
4766 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4767 canonical form (fully expanded) before being placed in
4768 $sender_host_address.
4770 24. Updated eximstats to version 1.33
4772 25. Include certificate and key file names in error message when GnuTLS fails
4773 to set them up, because the GnuTLS error message doesn't include the name
4774 of the failing file when there is a problem reading it.
4776 26. Expand error message when OpenSSL has problems setting up cert/key files.
4779 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4780 (this can affect the format of dates).
4782 28. exim_tidydb, when checking for the continued existence of a message for
4783 which it has found a message-specific retry record, was not finding
4784 messages that were in split spool directories. Consequently, it was
4785 deleting retry records that should have stayed in existence.
4787 29. eximstats updated to version 1.35
4788 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4789 1.35 - bugfix such that pie charts by volume are generated correctly
4791 30. The SPA authentication driver was not abandoning authentication and moving
4792 on to the next authenticator when an expansion was forced to fail,
4793 contradicting the general specification for all authenticators. Instead it
4794 was generating a temporary error. It now behaves as specified.
4796 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4797 (the order specifies the preference for clients). The order is now AES256,
4798 AES128, 3DES, ARCFOUR128.
4800 31. Small patch to Sieve code - explicitly set From: when generating an
4803 32. Exim crashed if a remote delivery caused a very long error message to be
4804 recorded - for instance if somebody sent an entire SpamAssassin report back
4805 as a large number of 550 error lines. This bug was coincidentally fixed by
4806 increasing the size of one of Exim's internal buffers (big_buffer) that
4807 happened as part of the Exiscan merge. However, to be on the safe side, I
4808 have made the code more robust (and fixed the comments that describe what
4811 33. Some experimental protocols are using DNS PTR records for new purposes. The
4812 keys for these records are domain names, not reversed IP addresses. The
4813 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4814 leaves it alone. Component reversal etc. now happens only for IP addresses.
4817 34. The host_aton() function is supposed to be passed a string that is known
4818 to be a valid IP address. However, in the case of IPv6 addresses, it was
4819 not checking this. This is a hostage to fortune. Exim now panics and dies
4820 if the condition is not met. A case was found where this could be provoked
4821 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4822 components; fortuitously, this particular loophole had already been fixed
4823 by change 4.50/55 or 4.44/33 above.
4825 If there are any other similar loopholes, the new check in host_aton()
4826 itself should stop them being exploited. The report I received stated that
4827 data on the command line could provoke the exploit when Exim was running as
4828 exim, but did not say which command line option was involved. All I could
4829 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4830 running as the user.
4833 35. There was a buffer overflow vulnerability in the SPA authentication code
4834 (which came originally from the Samba project). I have added a test to the
4835 spa_base64_to_bits() function which I hope fixes it.
4838 36. The daemon start-up calls getloadavg() while still root for those OS that
4839 need the first call to be done as root, but it missed one case: when
4840 deliver_queue_load_max is set with deliver_drop_privilege. This is
4841 necessary for the benefit of the queue runner, because there is no re-exec
4842 when deliver_drop_privilege is set.
4844 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4845 was in use, was not putting the data itself into the right store pool;
4846 consequently, it could be overwritten for a subsequent message in the same
4847 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4850 38. Sometimes the final signoff response after QUIT could fail to get
4851 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4852 < 0 before doing a fflush(). This bug looks as though it goes back to the
4853 introduction of TLS in release 3.20, but "sometimes" must have been rare
4854 because the tests only now provoked it.
4860 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4861 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4862 Now it returns an integer. A number of calls were still expecting a T/F
4863 return. Fortuitously, in all cases, the tests worked in OK situations,
4864 which is the norm. However, things would have gone wrong on any write
4865 failures on the smtp file descriptor. This function is used when sending
4866 messages over SMTP and also when doing verify callouts.
4868 2. When Exim is called to do synchronous delivery of a locally submitted
4869 message (the -odf or -odi options), it no longer closes stderr before doing
4872 3. Implemented the mua_wrapper option.
4874 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4876 5. Implemented the functions header_remove(), header_testname(),
4877 header_add_at_position(), and receive_remove_recipient(), and exported them
4880 6. If an ACL "warn" statement specified the addition of headers, Exim already
4881 inserted X-ACL-Warn: at the start if there was no header name. However, it
4882 was not making this test for the second and subsequent header lines if
4883 there were newlines in the string. This meant that an invalid header could
4884 be inserted if Exim was badly configured.
4886 7. Allow an ACL "warn" statement to add header lines at the start or after all
4887 the Received: headers, as well as at the end.
4889 8. Added the rcpt_4xx retry error code.
4891 9. Added postmaster_mailfrom=xxx to callout verification option.
4893 10. Added mailfrom=xxxx to the callout verification option, for verify=
4896 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4897 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4899 12. Inserted some casts to stop certain compilers warning when using pointer
4900 differences as field lengths or precisions in printf-type calls (mostly
4901 affecting debugging statements).
4903 13. Added optional readline() support for -be (dynamically loaded).
4905 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4906 same clock tick as a message's arrival, so that its received time was the
4907 same as the "first fail" time on the retry record, and that message
4908 remained on the queue past the ultimate address timeout, every queue runner
4909 would try a delivery (because it was past the ultimate address timeout) but
4910 after another failure, the ultimate address timeout, which should have then
4911 bounced the address, did not kick in. This was a "< instead of <=" error;
4912 in most cases the first failure would have been in the next clock tick
4913 after the received time, and all would be well.
4915 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4916 being recognized when the domain list was tested by the match_domain
4917 condition in an expansion string.
4919 16. Added the ${str2b64: operator.
4921 17. Exim was always calling setrlimit() to set a large limit for the number of
4922 processes, without checking whether the existing limit was already
4923 adequate. (It did check for the limit on file descriptors.) Furthermore,
4924 errors from getrlimit() and setrlimit() were being ignored. Now they are
4925 logged to the main and panic logs, but Exim does carry on, to try to do its
4926 job under whatever limits there are.
4928 18. Imported PCRE 5.0.
4930 19. Trivial typo in log message " temporarily refused connection" (the leading
4933 20. If the log selector return_path_on_delivery was set and an address was
4934 redirected to /dev/null, the delivery process crashed because it assumed
4935 that a return path would always be set for a "successful" delivery. In this
4936 case, the whole delivery is bypassed as an optimization, and therefore no
4939 21. Internal re-arrangement: the function for sending a challenge and reading
4940 a response while authentication was assuming a zero-terminated challenge
4941 string. It's now changed to take a pointer and a length, to allow for
4942 binary data in such strings.
4944 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4946 23. Exim was not respecting finduser_retries when seeking the login of the
4947 uid under which it was called; it was always trying 10 times. (The default
4948 setting of finduser_retries is zero.) Also, it was sleeping after the final
4949 failure, which is pointless.
4951 24. Implemented tls_on_connect_ports.
4953 25. Implemented acl_smtp_predata.
4955 26. If the domain in control=submission is set empty, Exim assumes that the
4956 authenticated id is a complete email address when it generates From: or
4957 Sender: header lines.
4959 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4960 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4961 chown and chgrp in /bin and hostname in /usr/bin.
4963 28. Exim was keeping the "process log" file open after each use, just as it
4964 does for the main log. This opens the possibility of it remaining open for
4965 long periods when the USR1 signal hits a daemon. Occasional processlog
4966 errors were reported, that could have been caused by this. Anyway, it seems
4967 much more sensible not to leave this file open at all, so that is what now
4970 29. The long-running daemon process does not normally write to the log once it
4971 has entered its main loop, and it closes the log before doing so. This is
4972 so that log files can straightforwardly be renamed and moved. However,
4973 there are a couple of unusual error situations where the daemon does write
4974 log entries, and I had neglected to close the log afterwards.
4976 30. The text of an SMTP error response that was received during a remote
4977 delivery was being truncated at 512 bytes. This is too short for some of
4978 the long messages that one sometimes sees. I've increased the limit to
4981 31. It is now possible to make retry rules that apply only when a message has a
4982 specific sender, in particular, an empty sender.
4984 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4985 it possible to be selective about when SMTP synchronization is enforced.
4987 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4989 32. Implemented hosts_connection_nolog.
4991 33. Added an ACL for QUIT.
4993 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4996 35. Added mailbox_size and mailbox_filecount to appendfile.
4998 36. Added control = no_multiline_responses to ACLs.
5000 37. There was a bug in the logic of the code that waits for the clock to tick
5001 in the case where the clock went backwards by a substantial amount such
5002 that the microsecond fraction of "now" was more than the microsecond
5003 fraction of "then" (but the whole seconds number was less).
5005 38. Added support for the libradius Radius client library this is found on
5006 FreeBSD (previously only the radiusclient library was supported).
5012 1. When certain lookups returned multiple values in the form name=value, the
5013 quoting of the values was not always being done properly. Specifically:
5014 (a) If the value started with a double quote, but contained no whitespace,
5016 (b) If the value contained whitespace other than a space character (i.e.
5017 tabs or newlines or carriage returns) it was not quoted.
5018 This fix has been applied to the mysql and pgsql lookups by writing a
5019 separate quoting function and calling it from the lookup code. The fix
5020 should probably also be applied to nisplus, ibase and oracle lookups, but
5021 since I cannot test any of those, I have not disturbed their existing code.
5023 2. A hit in the callout cache for a specific address caused a log line with no
5024 reason for rejecting RCPT. Now it says "Previous (cached) callout
5025 verification failure".
5027 3. There was an off-by-one bug in the queryprogram router. An over-long
5028 return line was truncated at 256 instead of 255 characters, thereby
5029 overflowing its buffer with the terminating zero. As well as fixing this, I
5030 have increased the buffer size to 1024 (and made a note to document this).
5032 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5033 when Exim is waiting for an SMTP response from a remote server, Exim
5034 restarts its select() call on the socket, thereby resetting its timeout.
5035 This is not a problem when such interrupts are rare. Somebody set up a cron
5036 job to run exiwhat every 2 minutes, which is less than the normal select()
5037 timeout (5 or 10 minutes). This meant that the select() timeout never
5038 kicked in because it was always reset. I have fixed this by comparing the
5039 time when an interrupt arrives with the time at the start of the first call
5040 to select(). If more time than the timeout has elapsed, the interrupt is
5041 treated as a timeout.
5043 5. Some internal re-factoring in preparation for the addition of Sieve
5044 extensions (by MH). In particular, the "personal" test is moved to a
5045 separate function, and given an option for scanning Cc: and Bcc: (which is
5046 not set for Exim filters).
5048 6. When Exim created an email address using the login of the caller as the
5049 local part (e.g. when creating a From: or Sender: header line), it was not
5050 quoting the local part when it contained special characters such as @.
5052 7. Installed new OpenBSD configuration files.
5054 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5055 try to make them clearer.
5057 9. Callout options, other than the timeout value, were being ignored when
5058 verifying sender addresses in header lines. For example, when using
5060 verify = header_sender/callout=no_cache
5062 the cache was (incorrectly) being used.
5064 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5065 only the Cygwin environment.
5067 11. When return_path_on_delivery was set as a log selector, if different remote
5068 addresses in the same message used different return paths and parallel
5069 remote delivery occurred, the wrong values would sometimes be logged.
5070 (Whenever a remote delivery process finished, the return path value from
5071 the most recently started remote delivery process was logged.)
5073 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5074 lines when AUTH and/or TLS are in use. This is the "received protocol"
5075 field. Exim used to use "asmtp" for authenticated SMTP, without any
5076 indication (in the protocol name) for TLS use. Now it follows the RFC and
5077 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5078 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5079 names appear in log lines as well as in Received: header lines.
5081 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5082 extensions, and comparison tests, and to fix some bugs.
5084 14. Changes to the "personal" filter test:
5086 (1) The test was buggy in that it was just doing the equivalent of
5087 "contains" tests on header lines. For example, if a user's address was
5088 anne@some.where, the "personal" test would incorrectly be true for
5090 To: susanne@some.where
5092 This test is now done by extracting each address from the header in turn,
5093 and checking the entire address. Other tests that are part of "personal"
5094 are now done using regular expressions (for example, to check local parts
5095 of addresses in From: header lines).
5097 (2) The list of non-personal local parts in From: addresses has been
5098 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5099 taken from the Sieve specification recommendations.
5101 (3) If the message contains any header line starting with "List-" it is
5102 treated as non-personal.
5104 (4) The test for "circular" in the Subject: header line has been removed
5105 because it now seems ill-conceived.
5107 15. Minor typos in src/EDITME comments corrected.
5109 16. Installed latest exipick from John Jetmore.
5111 17. If headers_add on a router specified a text string that was too long for
5112 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5113 of string_sprintf() is now avoided.
5115 18. $message_body_size was not set (it was always zero) when running the DATA
5116 ACL and the local_scan() function.
5118 19. For the "mail" command in an Exim filter, no default was being set for
5119 the once_repeat time, causing a random time value to be used if "once" was
5120 specified. (If the value happened to be <= 0, no repeat happened.) The
5121 default is now 0s, meaning "never repeat". The "vacation" command was OK
5122 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5123 (I found it when inspecting the code).
5125 20. There is now an overall timeout for performing a callout verification. It
5126 defaults to 4 times the callout timeout, which applies to individual SMTP
5127 commands during the callout. The overall timeout applies when there is more
5128 than one host that can be tried. The timeout is checked before trying the
5129 next host. This prevents very long delays if there are a large number of
5130 hosts and all are timing out (e.g. when the network connections are timing
5131 out). The value of the overall timeout can be changed by specifying an
5132 additional sub-option for "callout", called "maxwait". For example:
5134 verify = sender/callout=5s,maxwait=20s
5136 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5137 to the end before writing, but this should make it even safer).
5139 22. Exim was forgetting that it had advertised PIPELINING for the second and
5140 subsequent messages on an SMTP connection. It was also not resetting its
5141 memory on STARTTLS and an internal HELO.
5143 23. When Exim logs an SMTP synchronization error within a session, it now
5144 records whether PIPELINING has been advertised or not.
5146 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5147 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5148 rather than long int.
5150 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5152 26. Added the never_mail option to autoreply.
5158 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5159 crash if the getsockname() call failed; this can happen if a connection is
5160 closed very soon after it is established. The problem was simply in the
5161 order in which certain operations were done, causing Exim to try to write
5162 to the SMTP stream before it had set up the file descriptor. The bug has
5163 been fixed by making things happen in the correct order.
5169 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5170 before the connection was closed, thus losing the rejection response.
5172 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5173 some early Solaris releases, but causes trouble in current releases where
5174 socklen_t is defined.
5176 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5179 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5182 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5183 that releasing the top of it at the end releases what was used for sub-
5184 expansions (unless the block got too big). However, discard this block if
5185 the first thing is a variable or header, so that we can use its block when
5186 it is dynamic (useful for very large $message_headers, for example).
5188 6. Lookups now cache *every* query, not just the most recent. A new, separate
5189 store pool is used for this. It can be recovered when all lookup caches are
5190 flushed. Lookups now release memory at the end of their result strings.
5191 This has involved some general refactoring of the lookup sources.
5193 7. Some code has been added to the store_xxx() functions to reduce the amount
5194 of flapping under certain conditions.
5196 8. log_incoming_interface used to affect only the <= reception log lines. Now
5197 it causes the local interface and port to be added to several more SMTP log
5198 lines, for example "SMTP connection from", and rejection lines.
5200 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5202 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5204 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5205 could be overwritten at the end of the current message (or the start of a
5206 new message if it was set in a HELO ACL). The value is now preserved for
5207 the duration of the SMTP connection.
5209 12. If a transport had a headers_rewrite setting, and a matching header line
5210 contained an unqualified address, that address was qualified, even if it
5211 did not match any rewriting rules. The underlying bug was that the values
5212 of the flags that permit the existence of unqualified sender and recipient
5213 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5214 non-local messages, and by -bnq for local messages) were not being
5215 preserved with the message after it was received.
5217 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5218 "next input=" as part of the text comprising the host identity instead of
5219 the correct text. The code was using the same buffer for two different
5220 strings. However, depending on which order the printing function evaluated
5221 its arguments, the bug did not always show up. Under Linux, for example, my
5222 test suite worked just fine.
5224 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5225 doesn't work with some older versions of Perl. It has been changed to "my",
5226 which in any case is probably the better facility to use.
5228 15. A really picky compiler found some instances of statements for creating
5229 error messages that either had too many or two few arguments for the format
5232 16. The size of the buffer for calls to the DNS resolver has been increased
5233 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5234 for addresses that have a lot of PTR records. This alleviates a problem; it
5235 does not fully solve it.
5237 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5238 buffer now truncates the list and logs the incident, which is the same
5239 action as happens when Exim is looking up a host name and its aliases.
5240 Previously in this situation something unpredictable would happen;
5241 sometimes it was "internal error: store_reset failed".
5243 18. If a server dropped the connection unexpectedly when an Exim client was
5244 using GnuTLS and trying to read a response, the client delivery process
5245 crashed while trying to generate an error log message.
5247 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5248 string, for example:
5250 warn message = H1: something\nH2: something
5252 the text was added as a single header line from Exim's point of view
5253 though it ended up OK in the delivered message. However, searching for the
5254 second and subsequent header lines using $h_h2: did not work. This has been
5255 fixed. Similarly, if a system filter added multiple headers in this way,
5256 the routers could not see them.
5258 20. Expanded the error message when iplsearch is called with an invalid key to
5259 suggest using net-iplsearch in a host list.
5261 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5262 ACLs are no longer actually imposed (and a message to that effect is
5265 22. If a "gecos" field in a passwd entry contained escaped characters, in
5266 particular, if it contained a \" sequence, Exim got it wrong when building
5267 a From: or a Sender: header from that name. A second bug also caused
5268 incorrect handling when an unquoted " was present following a character
5269 that needed quoting.
5271 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5272 was not being matched caselessly.
5274 24. Arranged for all hyphens in the exim.8 source to be escaped with
5277 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5278 properly. Recipient callout cache records were still being keyed to include
5279 the sender, even when use_sender was set false. This led to far more
5280 callouts that were necessary. The sender is no longer included in the key
5281 when use_sender is false.
5283 26. Added "control = submission" modifier to ACLs.
5285 27. Added the ${base62d: operator to decode base 62 numbers.
5287 28. dnsdb lookups can now access SRV records.
5289 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5290 the configuration file.
5292 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5293 (-v) mode. This makes the output for a verbose queue run more intelligible.
5295 31. Added a use_postmaster feature to recipient callouts.
5297 32. Added the $body_zerocount variable, containing the number of binary zero
5298 bytes in the message body.
5300 33. The time of last modification of the "new" subdirectory is now used as the
5301 "mailbox time last read" when there is a quota error for a maildir
5304 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5306 35. Added +ignore_unknown as a special item in host lists.
5308 36. Code for decoding IPv6 addresses in host lists is now included, even if
5309 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5310 address was recognized as an IP address, but was then not correctly decoded
5311 into binary, causing unexpected and incorrect effects when compared with
5318 1. Very minor rewording of debugging text in manualroute to say "list of
5319 hosts" instead of "hostlist".
5321 2. If verify=header_syntax was set, and a header line with an unqualified
5322 address (no domain) and a large number of spaces between the end of the
5323 name and the colon was received, the reception process suffered a buffer
5324 overflow, and (when I tested it) crashed. This was caused by some obsolete
5325 code that should have been removed. The fix is to remove it!
5327 3. When running in the test harness, delay a bit after writing a bounce
5328 message to get a bit more predictability in the log output.
5330 4. Added a call to search_tidyup() just before forking a reception process. In
5331 theory, someone could use a lookup in the expansion of smtp_accept_max_
5332 per_host which, without the tidyup, could leave open a database connection.
5334 5. Added the variables $recipient_data and $sender_data which get set from a
5335 lookup success in an ACL "recipients" or "senders" condition, or a router
5336 "senders" option, similar to $domain_data and $local_part_data.
5338 6. Moved the writing of debug_print from before to after the "senders" test
5341 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5342 problems for message scanning, either using a data ACL, or using
5343 local_scan() because the Received: header was not generated till after they
5344 were called (in order to set the time as the time of reception completion).
5345 I have revised the way this works. The header is now generated after the
5346 body is received, but before the ACL or local_scan() are called. After they
5347 are run, the timestamp in the header is updated.
5353 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5354 before starting a queue runner without re-exec. This happened only when
5355 deliver_drop_privilege was set or when the Exim user was set to root. The
5356 effect of the bug was that timeouts during subsequent deliveries caused
5357 crashes instead of being properly handled. The handler is now left at its
5358 default (and expected) setting.
5360 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5361 message, again when deliver_drop_privilege is set or Exim is run as root.
5362 The bug described in (1) was not present in this case, but the tidying up
5363 of the other signals was missing. I have made the two cases consistent.
5365 3. The ignore_target_hosts setting on a manualroute router was being ignored
5366 for hosts that were looked up using the /MX notation.
5368 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5371 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5372 operated on the sender address. After changing the $sender_address to <>
5373 for the sender address verify, Exim was re-instated it as the original
5374 (before rewriting) address, but remembering that it had rewritten it, so it
5375 wasn't rewriting it again. This bug also had the effect of breaking the
5376 sender address verification caching when the sender address was rewritten.
5378 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5379 This has been changed so that if the ip literal address matches
5380 ignore_target_hosts, the router declines.
5382 7. Added expansion conditions match_domain, match_address, and match_local_
5383 part (NOT match_host).
5385 8. The placeholder for the Received: header didn't have a length field set.
5387 9. Added code to Exim itself and to exim_lock to test for a specific race
5388 condition that could lead to file corruption when using MBX delivery. The
5389 issue is with the lockfile that is created in /tmp. If this file is removed
5390 after a process has opened it but before that process has acquired a lock,
5391 there is the potential for a second process to recreate the file and also
5392 acquire a lock. This could lead to two Exim processes writing to the file
5393 at the same time. The added code performs the same test as UW imapd; it
5394 checks after acquiring the lock that its file descriptor still refers to
5395 the same named file.
5397 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5398 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5401 11. Added the smtp_active_hostname option. If used, this will typically be made
5402 to depend on the incoming interface address. Because $interface_address is
5403 not set up until the daemon has forked a reception process, error responses
5404 that can happen earlier (such as "too many connections") no longer contain
5407 12. If an expansion in a condition on a "warn" statement fails because a lookup
5408 defers, the "warn" statement is abandoned, and the next ACL statement is
5409 processed. Previously this caused the whole ACL to be aborted.
5411 13. Added the iplsearch lookup type.
5413 14. Added ident_timeout as a log selector.
5415 15. Added tls_certificate_verified as a log selector.
5417 16. Added a global option tls_require_ciphers (compare the smtp transport
5418 option of the same name). This controls incoming TLS connections.
5420 17. I finally figured out how to make tls_require_ciphers do a similar thing
5421 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5422 before starting the TLS session.
5424 18. Tabs are now shown as \t in -bP output.
5426 19. If the log selector return_path_on_delivery was set, Exim crashed when
5427 bouncing a message because it had too many Received: header lines.
5429 20. If two routers both had headers_remove settings, and the first one included
5430 a superfluous trailing colon, the final name in the first list and the
5431 first name in the second list were incorrectly joined into one item (with a
5432 colon in the middle).
5438 1. Added -C and -D options to the exinext utility, mainly to make it easier
5439 to include in the automated testing, but these could be helpful when
5440 multiple configurations are in use.
5442 2. The exinext utility was not formatting the output nicely when there was
5443 an alternate port involved in the retry record key, nor when there was a
5444 message id as well (for retries that were specific to a specific message
5445 and a specific host). It was also confused by IPv6 addresses, because of
5446 the additional colons they contain. I have fixed the IPv4 problem, and
5447 patched it up to do a reasonable job for IPv6.
5449 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5450 delivery, the log line now contains "pipelined" if PIPELINING was used.
5452 4. An SMTP transport process used to panic and die if the bind() call to set
5453 an explicit outgoing interface failed. This has been changed; it is now
5454 treated in the same way as a connect() failure.
5456 5. A reference to $sender_host_name in the part of a conditional expansion
5457 that was being skipped was still causing a DNS lookup. This no longer
5460 6. The def: expansion condition was not recognizing references to header lines
5461 that used bh_ and bheader_.
5463 7. Added the _cache feature to named lists.
5465 8. The code for checking quota_filecount in the appendfile transport was
5466 allowing one more file than it should have been.
5468 9. For compatibility with Sendmail, the command line option
5476 and sets the incoming protocol and host name (for trusted callers). The
5477 host name and its colon can be omitted when only the protocol is to be set.
5478 Note the Exim already has two private options, -pd and -ps, that refer to
5479 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5480 "s", but I don't think that's a major issue.
5482 10. A number of refactoring changes to the code, none of which should affect
5485 (a) The number of logging options was getting close to filling up the
5486 32-bit word that was used as a bit map. I have split them into two classes:
5487 those that are passed in the argument to log_write(), and those that are
5488 only ever tested independently outside of that function. These are now in
5489 separate 32-bit words, so there is plenty of room for expansion again.
5490 There is no change in the user interface or the logging behaviour.
5492 (b) When building, for example, log lines, the code previously used a
5493 macro that called string_cat() twice, in order to add two strings. This is
5494 not really sufficiently general. Furthermore, there was one instance where
5495 it was actually wrong because one of the argument was used twice, and in
5496 one call a function was used. (As it happened, calling the function twice
5497 did not affect the overall behaviour.) The macro has been replaced by a
5498 function that can join an arbitrary number of extra strings onto a growing
5501 (c) The code for expansion conditions now uses a table and a binary chop
5502 instead of a serial search (which was left over from when there were very
5503 few conditions). Also, it now recognizes conditions like "pam" even when
5504 the relevant support is not compiled in: a suitably worded error message is
5505 given if an attempt is made to use such a condition.
5507 11. Added ${time_interval:xxxxx}.
5509 12. A bug was causing one of the ddress fields not to be passed back correctly
5510 from remote delivery subprocesses. The field in question was not being
5511 subsequently used, so this caused to problems in practice.
5513 13. Added new log selectors queue_time and deliver_time.
5515 14. Might have fixed a bug in maildirsizefile handling that threw up
5516 "unexpected character" debug warnings, and recalculated the data
5517 unnecessarily. In any case, I expanded the warning message to give more
5520 15. Added the message "Restricted characters in address" to the statements in
5521 the default ACL that block characters like @ and % in local parts.
5523 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5524 Three changes have been made:
5526 (a) There was a serious bug; a negative response to MAIL caused the whole
5527 recipient domain to be cached as invalid, thereby blocking all messages
5528 to all local parts at the same domain, from all senders. This bug has
5529 been fixed. The domain is no longer cached after a negative response to
5530 MAIL if the sender used is not empty.
5532 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5535 (c) A new callout option, "use_sender" has been added for people who want
5536 the modified behaviour.
5542 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5545 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5546 indeed breaks things for older releases.
5548 3. Added additional logging to the case where there is a problem reading data
5549 from a filter that is running in a subprocess using a pipe, in order to
5550 try to track down a specific problem.
5552 4. Testing facility fudge: when running in the test harness and attempting
5553 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5554 getting "No route to host". Convert this to a timeout.
5556 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5559 6. Some OS don't have socklen_t but use size_t instead. This affects the
5560 fifth argument of getsockopt() amongst other things. This is now
5561 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5562 can be set for individual OS. I have set it for SunOS5, OSF1, and
5563 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5564 some earlier ones do not.
5566 7. Change 4.30/15 was not doing the test caselessly.
5568 8. The standard form for an IPv6 address literal was being rejected by address
5569 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5570 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5571 this, as well as the form without the "IPv6" on the front (but only when
5572 address literals are enabled, of course).
5574 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5576 10. Exim crashed if a message with an empty sender address specified by -f
5577 encountered a router with an errors_to setting. This could be provoked only
5578 by a command such as
5582 where an empty string was supplied; "<>" did not hit this bug.
5584 11. Installed PCRE release 4.5.
5586 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5587 remained set. It is now erased.
5589 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5590 times from message ids (which are base 36 rather than the normal 62).
5592 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5593 were being counted as actual protocol errors, and logged if the log
5594 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5595 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5596 and DATA following a set of rejected RCPTs do not count as protocol errors.
5597 In other words, Exim assumes they were pipelined, though this may not
5598 actually be the case. Of course, in all cases the client gets an
5599 appropriate error code.
5601 15. If a lookup fails in an ACL condition, a message about the failure may
5602 be available; it is used if testing the ACL cannot continue, because most
5603 such messages specify what the cause of the deferral is. However, some
5604 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5605 that caused an old message to be retained and used if a later statement
5606 caused a defer, replacing the real cause of the deferral.
5608 16. If an IP address had so many PTR records that the DNS lookup buffer
5609 was not large enough to hold them, Exim could crash while trying to process
5610 the truncated data. It now detects and logs this case.
5612 17. Further to 4.21/58, another change has been made: if (and only if) the
5613 first line of a message (the first header line) ends with CRLF, a bare LF
5614 in a subsequent header line has a space inserted after it, so as not to
5615 terminate the header.
5617 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5618 unnecessarily, used atoi() instead of strtol(), and didn't check the
5619 termination when getting file sizes from file names by regex.
5621 19. Completely re-implemented the support for maildirsize files, in the light
5622 of a number of problems with the previous contributed implementation
5623 (4.30/29). In particular:
5625 . If the quota is zero, the maildirsize file is maintained, but no quota is
5628 . If the maildir directory does not exist, it is created before any attempt
5629 to write a maildirsize file.
5631 . The quota value in the file is just a cache; if the quota is changed in
5632 the transport, the new value overrides.
5634 . A regular expression is available for excluding directories from the
5637 20. The autoreply transport checks the characters in options that define the
5638 message's headers; it allows continued headers, but it was checking with
5639 isspace() after an embedded newline instead of explicitly looking for a
5642 21. If all the "regular" hosts to which an address was routed had passed their
5643 expiry times, and had not reached their retry times, the address was
5644 bounced, even if fallback hosts were defined. Now Exim should go on to try
5647 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5648 equivalent code in the SMTP transport. Some hosts send humungous responses
5649 to HELO/EHLO, more than 1024 it seems.
5651 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5652 gives compiler warnings in some environments. I've now done it "properly",
5655 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5656 (because of problems with the built-in one) was declared to return uschar *
5657 instead of char *, causing compiler failure.
5659 25. Fixed a file descriptor leak when processing alias/forward files.
5661 26. Fixed a minor format string issue in dbfn.c.
5663 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5665 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5666 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5669 29. When a sender address is verified, it is cached, to save repeating the test
5670 when there is more than one recipient in a message. However, when the
5671 verification involves a callout, it is possible for different callout
5672 options to be set for different recipients. It is too complicated to keep
5673 track of this in the cache, so now Exim always runs a verification when a
5674 callout is required, relying on the callout cache for the optimization.
5675 The overhead is duplication of the address routing, but this should not be
5678 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5679 to be verified with callout=postmaster, and the main callout worked but the
5680 postmaster check failed, the verification correctly failed. However, if a
5681 subsequent RCPT command asked for sender verification *without* the
5682 postmaster check, incorrect caching caused this verification also to fail,
5685 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5686 it was not caching the DNS options (qualify_single, search_parents) that
5687 were used when the lookup failed. A subsequent lookup with different
5688 options therefore always gave the same answer, though there were cases
5689 where it should not have. (Example: a "domains = !$mx_any" option on a
5690 dnslookup router: the "domains" option is always processed without any
5691 widening, but the router might have qualify_single set.) Now Exim uses the
5692 cached value only when the same options are set.
5694 32. Added John Jetmore's "exipick" utility to the distribution.
5696 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5697 than a timeout (e.g. a certificate is required, and is not provided), an
5698 Exim server now closes the connection immediately. Previously it waited for
5699 the client to close - but if the client is SSL, it seems that they each
5700 wait for each other, leading to a delay before one of them times out.
5702 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5703 maintained 0.8.x compatibility because I don't think many are using it, and
5704 it is clearly obsolete.
5706 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5709 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5710 client certificate was expired. A simple patch fixes this, though I don't
5711 understand the full logic of why the verify callback is called multiple
5714 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5715 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5716 which causes problems with some clients (such as the Certicom SSL Plus
5717 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5718 disables the coutermeasure allowing Eudora to connect."
5720 38. Exim was not checking that a write() to a log file succeeded. This could
5721 lead to Bad Things if a log got too big, in particular if it hit a file
5722 size limit. Exim now panics and dies if it cannot write to a log file, just
5723 as it does if it cannot open a log file.
5725 39. Modified OS/Makefile-Linux so that it now contains
5727 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5729 The two -D definitions ensure that Exim is compiled with large file
5730 support, which makes it possible to handle log files that are bigger than
5733 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5734 instance) a domain was checked against a named list that involved a lookup,
5735 causing $domain_data to be set, then another domain was checked against the
5736 same list, then the first domain was re-checked, the value of $domain_data
5737 after the final check could be wrong. In particular, if the second check
5738 failed, it could be set empty. This bug probably also applied to
5741 41. The strip_trailing_dot option was not being applied to the address given
5742 with the -f command-line option.
5744 42. The code for reading a message's header from the spool was incrementing
5745 $received_count, but never initializing it. This meant that the value was
5746 incorrect (doubled) while delivering a message in the same process in which
5747 it was received. In the most common configuration of Exim, this never
5748 happens - a fresh exec is done - but it can happen when
5749 deliver_drop_privilege is set.
5751 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5752 it now includes up to 150 characters of the unexpected data in the log
5755 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5756 and building data strings. The size of both of these buffers was 10 000
5757 bytes - far larger than anybody would *ever* want, thought I. Needless to
5758 say, somebody hit the limit. I have increased the maximum line length to
5759 20 000 and the maximum data length of concatenated lines to 100 000. I have
5760 also fixed two bugs, because there was no checking on these buffers. Tsk,
5761 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5762 buffer is too small.
5764 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5765 lsearch lookups. Now it does.
5767 46. When parsing a route_list item in a manualroute router, a fixed-length
5768 buffer was used for the list of hosts. I made this 1024 bytes long,
5769 thinking that nobody would ever have a list of hosts that long. Wrong.
5770 Somebody had a whole pile of complicated expansion conditions, and the
5771 string was silently truncated, leading to an expansion error. It turns out
5772 that it is easier to change to an unlimited length (owing to other changes
5773 that have happened since this code was originally written) than to build
5774 structure for giving a limitation error. The length of the item that
5775 expands into the list of hosts is now unlimited.
5777 47. The lsearch lookup could not handle data where the length of text line was
5778 more than 4095 characters. Such lines were truncated, leading to shortened
5779 data being returned. It should now handle lines of any length.
5781 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5782 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5785 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5786 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5787 no longer happens in comment lines. A list of replacements is now placed
5788 at the head of all of the source files, except those whose only change is
5789 to replace PERL_COMMAND in the very first #! line.
5791 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5792 messages on the queue, with a bottom-up merge sort, using code contributed
5793 by Michael Haardt. This should make operations like -bp somewhat faster on
5794 large queues. It won't affect queue runners, except when queue_run_in_order
5797 51. Installed eximstats 1.31 in the distribution.
5799 52. Added support for SRV lookups to the dnslookup router.
5801 53. If an ACL referred to $message_body or $message_body_end, the value was not
5802 reset for any messages that followed in the same SMTP session.
5804 54. The store-handling optimization for building very long strings was not
5805 differentiating between the different store pools. I don't think this
5806 actually made any difference in practice, but I've tidied it.
5808 55. While running the routers to verify a sender address, $sender_address
5809 was still set to the sender address. This is wrong, because when routing to
5810 send a bounce to the sender, it would be empty. Therefore, I have changed
5811 it so that, while verifying a sender address, $sender_address is set to <>.
5812 (There is no change to what happens when verifying a recipient address.)
5814 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5815 target A or AAAA records (if not already returned) without resetting the
5816 qualify_single or search_parents options of the DNS resolver. These are
5817 inappropriate in this case because the targets of MX and SRV records must
5818 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5819 target that, when qualified, matched something in the local domain. These
5820 two options are now turned off when doing these lookups.
5822 57. It seems that at least some releases of Reiserfs (which does not have the
5823 concept of a fixed number of inodes) returns zero and not -1 for the
5824 number of available inodes. This interacted badly with check_spool_inodes,
5825 which assumed that -1 was the "no such thing" setting. What I have done is
5826 to check that the total number of inodes is greater than zero before doing
5827 the test of how many are available.
5829 58. When a "warn" ACL statement has a log_message modifier, the message is
5830 remembered, and not repeated. This is to avoid a lot of repetition when a
5831 message has many recipients that cause the same warning to be written.
5832 However, Exim was preserving the list of already written lines for an
5833 entire SMTP session, which doesn't seem right. The memory is now reset if a
5834 new message is started.
5836 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5837 debugging output unless log_rewrite was also set.
5839 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5840 of (int)(handle) when we know that handle contains (void *)(-1).
5842 61. The Exim daemon panic-logs an error return when it closes the incoming
5843 connection. However "connection reset by peer" seems to be common, and
5844 isn't really an error worthy of noting specially, so that particular error
5847 62. When Exim is trying to find all the local interfaces, it used to panic and
5848 die if the ioctl to get the interface flags failed. However, it seems that
5849 on at least one OS (Solaris 9) it is possible to have an interface that is
5850 included in the list of interfaces, but for which you get a failure error
5851 for this call. This happens when the interface is not "plumbed" into a
5852 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5853 failure of the "get flags" call assumes that the interface is down.
5855 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5856 makes life easier for people who are doing arithmetic on fields extracted
5857 from dates, where you often get leading zeros that should not be
5858 interpreted as octal.
5860 64. Added qualify_domain to the redirect router, to override the global
5863 65. If a pathologically long header line contained very many addresses (the
5864 report of this problem mentioned 10 000) and each of them was rewritten,
5865 Exim could use up a very large amount of memory. (It kept on making new
5866 copies of the header line as it rewrote, and never released the old ones.)
5867 At the expense of a bit more processing, the header rewriting function has
5868 been changed so that it no longer eats memory in this way.
5870 66. The generation of the Received: header has been moved from the time that a
5871 message starts to be received, to the time that it finishes. The timestamp
5872 in the Received: header should now be very close to that of the <= log
5873 line. There are two side-effects of this change:
5875 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5876 logged header lines no longer include the local Received: line, because
5877 it has not yet been created. The same applies to a copy of the message
5878 that is returned to a non-SMTP sender when a message is rejected.
5880 (b) When a filter file is tested using -bf, no additional Received: header
5881 is added to the test message. After some thought, I decided that this
5884 This change does not affect the value of $received_for. It is still set
5885 after address rewriting, but before local_scan() is called.
5887 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5889 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5890 gave an unhelpful panic error message, and a defer error. I have managed to
5891 change this behaviour so that it now rejects any supplied certificate,
5892 which seems right, as the list of acceptable certificates is empty.
5894 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5895 gave an unhelpful defer error. I have not managed to make this reject any
5896 supplied certificates, but the error message it gives is "no certificate
5897 supplied", which is not helpful.
5899 70. exigrep's output now also includes lines that are not associated with any
5900 message, but which match the given pattern. Implemented by a patch from
5901 Martin Sluka, which also tidied up the Perl a bit.
5903 71. Recipient callout verification, like sender verification, was using <> in
5904 the MAIL FROM command. This isn't really the right thing, since the actual
5905 sender may affect whether the remote host accepts the recipient or not. I
5906 have changed it to use the actual sender in the callout; this means that
5907 the cache record is now keyed on a recipient/sender pair, not just the
5908 recipient address. There doesn't seem to be a real danger of callout loops,
5909 since a callout by the remote host to check the sender would use <>.
5910 [SEE ABOVE: changed after hitting problems.]
5912 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5913 temporary errors. However, in the case of such a code being given after
5914 the end of a data transmission (i.e. after ".") Exim was failing to write
5915 a retry record for the message. (Yes, there was some broken host that was
5916 actually sending 8xx at this point.)
5918 73. An unknown lookup type in a host list could cause Exim to panic-die when
5919 the list was checked. (An example that provoked this was putting <; in the
5920 middle of a list instead of at the start.) If this happened during a DATA
5921 ACL check, a -D file could be left lying around. This kind of configuration
5922 error no longer causes Exim to die; instead it causes a defer error. The
5923 incident is still logged to the main and panic logs.
5925 74. Buglet left over from Exim 3 conversion. The message "too many messages
5926 in one connection" was written to the rejectlog but not the mainlog, except
5927 when address rewriting (yes!) was being logged.
5929 75. Added write_rejectlog option.
5931 76. When a system filter was run not as root (that is, when system_filter_user
5932 was set), the values of the $n variables were not being returned to the
5933 main process; thus, they were not subsequently available in the $sn
5936 77. Added +return_path_on_delivery log selector.
5938 78. A connection timeout was being treated differently from recipients deferred
5939 when testing hosts_max_try with a message that was older than the host's
5940 retry timeout. (The host should not be counted, thus allowing all hosts to
5941 be tried at least once before bouncing.) This may have been the cause of an
5942 occasionally reported bug whereby a message would remain on the queue
5943 longer than the retry timeout, but would be bounced if a delivery was
5944 forced. I say "may" because I never totally pinned down the problem;
5945 setting up timeout/retry tests is difficult. See also the next item.
5947 79. The ultimate address timeout was not being applied to errors that involved
5948 a combination of host plus message (for example, a timeout on a MAIL
5949 command). When an address resolved to a number of possible hosts, and they
5950 were not all tried for each delivery (e.g. because of hosts_max_try), a
5951 message could remain on the queue longer than the retry timeout.
5953 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5956 81. Fixed an obscure SMTP outgoing bug which required at least the following
5957 conditions: (a) there was another message waiting for the same server;
5958 (b) the server returned 5xx to all RCPT commands in the first message so
5959 that the message was not completed; (c) the server dropped the connection
5960 or gave a negative response to the RSET that Exim sends to abort the
5961 transaction. The observed case was a dropped connection after DATA that had
5962 been sent in pipelining mode. That is, the server had advertised PIPELINING
5963 but was not implementing it correctly. The effect of the bug was incorrect
5964 behaviour, such as trying another host, and this could lead to a crash.
5970 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5971 and daemon.c were passed as pointers to ints; they should have been
5972 pointers to socklen_t variables (which are typically unsigned ints).
5974 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5977 3. Fixed a really odd bug that affected only the testing scheme; patching a
5978 certain fixed string in the binary changed the value of another string that
5979 happened to be identical to the end of the original first string.
5981 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5982 name", it returns that address as the IP address. On some operating
5983 systems (e.g. Solaris), it also passes back the IP address string as the
5984 "host name". However, on others (e.g. Linux), it passes back an empty
5985 string. Exim wasn't checking for this, and was changing the host name to an
5986 empty string, assuming it had been canonicalized.
5988 5. Although rare, it is permitted to have more than one PTR record for a given
5989 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5990 all the names associated with an address, because they do in Solaris.
5991 However, it seems that they do not in Linux for data that comes from the
5992 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5993 I found this out when I moved to a new Linux workstation and tried to run
5994 the Exim test suite.
5996 To get round this problem I have changed the code so that it now does its
5997 own call to the DNS to look up PTR records when searching for a host name.
5998 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5999 addresses that are only in /etc/hosts are still found.
6001 This behaviour is, however, controlled by an option called host_lookup_
6002 order, which defaults to "bydns:byaddr". If people want to use the other
6003 order, or indeed, just use one or the other means of lookup, they can
6004 specify it in this variable.
6006 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6007 some operating systems, this comes back from gethostbyaddr() as an empty
6008 string, and this is what Exim used to test for. However, it seems that in
6009 other systems, "." is yielded. Exim now tests for this case too.
6011 7. The values of check_spool_space and check_log_space are now held internally
6012 as a number of kilobytes instead of an absolute number of bytes. If a
6013 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6014 kilobyte. This means that much larger values can be stored.
6016 8. Exim monitor: an attempt to get the action menu when not actually pointing
6017 at a message produces an empty menu entitled "No message selected". This
6018 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6019 no entries in it ("Shell widget menu has zero width and/or height"). So I
6020 have added a single, blank menu entry in this case.
6022 9. Added ${quote_local_part.
6024 10. MIME decoding is now applied to the contents of Subject: header lines when
6027 11. Now that a reference to $sender_host_address automatically causes a reverse
6028 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6029 host lookup before query-style lookups in lists that might use this
6030 variable. This has therefore been abolished, and the "net-" prefix is no
6031 longer necessary for query-style lookups.
6033 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6034 have been a typo for LFLAGS, so it has been changed.
6036 13. The install script calls Exim with "-C /dev/null" in order to find the
6037 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6038 to be output. However, since Exim outputs its version number before the
6039 error, it didn't break the script. It just looked ugly. I fixed this by
6040 always allowing "-C /dev/null" if the caller is root.
6042 14. Ignore overlarge ACL variable number when reading spool file - insurance
6043 against a later release with more variables having written the file.
6045 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6046 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6047 this, as well as the form without the "IPv6" on the front.
6049 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6050 OS/Makefile-Darwin file.
6052 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6053 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6055 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6056 variable. However, some OS use ints for inodes. I've added cast to long int
6057 to get rid of the compiler warning.
6059 19. I had forgotten to lock out "/../" in configuration file names when
6060 ALT_CONFIG_PREFIX was set.
6062 20. Routers used for verification do not need to specify transports. However,
6063 if such a router generated a host list, and callout was configured, Exim
6064 crashed, because it could not find a port number from the (non-existent)
6065 transport. It now assumes port 25 in this circumstance.
6067 21. Added the -t option to exigrep.
6069 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6070 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6071 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6072 them is set, LOOKUP_LSEARCH is forced.
6074 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6076 24. Added sender and host information to the "rejected by local_scan()" log
6077 line; previously there was no indication of these.
6079 25. Added .include_if_exists.
6081 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6082 Linux. It turns out that not all file systems support this. Apparently some
6083 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6084 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6085 not supported on the file descriptor, is now ignored when Exim is trying to
6086 sync a directory. This applies only to Linux.
6088 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6090 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6091 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6094 29. Added support for maildirsize files from supplied patch (modified a bit).
6096 30. The use of :fail: followed by an empty string could lead Exim to respond to
6097 sender verification failures with (e.g.):
6099 550 Verification failed for <xxx>
6100 550 Sender verify failed
6102 where the first response line was missing the '-' that indicates it is not
6103 the final line of the response.
6105 31. The loop for finding the name of the user that called Exim had a hardwired
6106 limit of 10; it now uses the value of finduser_retries, which is used for
6107 all other user lookups.
6109 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6112 33. Exim was neglecting to zero errno before one call of strtol() when
6113 expanding a string and expecting an integer value. On some systems this
6114 resulted in spurious "integer overflow" errors. Also, it was casting the
6115 result into an int without checking.
6117 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6118 did not work. The code looks as if it has *never* worked, though it appears
6119 to have been documented since at least release 1.62. I have made it work.
6121 35. The "timeout_DNS" error in retry rules, also documented since at least
6122 1.62, also never worked. As it isn't clear exactly what this means, and
6123 clearly it isn't a major issue, I have abolished the feature by treating it
6124 as "timeout", and writing a warning to the main and panic logs.
6126 36. The display of retry rules for -brt wasn't always showing the error code
6129 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6130 timeout_connect_A, timeout_connect_MX.
6132 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6133 to the empty sender.
6135 39. The daemon was not analysing the content of -oX till after it had closed
6136 stderr and disconnected from the controlling terminal. This meant that any
6137 syntax errors were only noted on the panic log, and the return code from
6138 the command was 0. By re-arranging the code a little, I've made the
6139 decoding happen first, so such errors now appear on stderr, and the return
6140 code is 1. However, the actual setting up of the sockets still happens in
6141 the disconnected process, so errors there are still only recorded on the
6144 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6145 connections (as happens on some IP stacks) was logged at start up time as
6146 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6147 it from "IPv6 and IPv4", which means that two separate sockets are being
6150 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6151 says whether AF_INET or AF_INET6 was passed as an argument.
6153 42. Exiwhat output was messed up when time zones were included in log
6156 43. Exiwhat now gives more information about the daemon's listening ports,
6157 and whether -tls-on-connect was used.
6159 44. The "port" option of the smtp transport is now expanded.
6161 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6162 silently ignored. Now an error message is written to the main and panic
6165 46. There's a new ACL modifier called "logwrite" which writes to a log file
6166 as soon as it is encountered.
6168 47. Added $local_user_uid and $local_user_gid at routing time.
6170 48. Exim crashed when trying to verify a sender address that was being
6173 49. Exim was recognizing only a space character after ".include". It now also
6174 recognizes a tab character.
6176 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6177 extracting the relevant information from the specification. The man page no
6178 longer contains scrambled data for the -d option, and I've added a section
6179 at the front about calling Exim under different names.
6181 51. Added "extra_headers" argument to the "mail" command in filter files.
6183 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6186 53. Installed eximstats 1.29.
6188 54. Added transport_filter_timeout as a generic transport option.
6190 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6191 Cc: header lines. This was required by RFC 822, but it not required by RFC
6194 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6195 incoming messages that did not have them. Now it does so only if the
6196 message originates locally, that is, if there is no associated remote host
6197 address. When Resent- header lines are present, this applies to the Resent-
6198 lines rather than the non-Resent- lines.
6200 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6201 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6203 58. Messages for configuration errors now include the name of the main
6204 configuration file - useful now that there may be more than one file in a
6205 list (.included file names were always shown).
6207 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6208 for those rare installations that do not start the daemon as root or run it
6209 setuid root. I've cut out the call to initgroups() if the daemon is not
6212 60. The Exim user and group can now be bound into the binary as text strings
6213 that are looked up at the start of Exim's processing.
6215 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6217 62. Added $mailstore_basename variable.
6219 63. Installed patch to sieve.c from Michael Haardt.
6221 64. When Exim failed to open the panic log after failing to open the main log,
6222 the original message it was trying to log was written to stderr and debug
6223 output, but if they were not available (the usual case in production), it
6224 was lost. Now it is written to syslog before the two lines that record the
6225 failures to open the logs.
6227 65. Users' Exim filters run in subprocesses under the user's uid. It is
6228 possible for a "deliver" command or an alias in a "personal" command to
6229 provoke an address rewrite. If logging of address rewriting is configured,
6230 this fails because the process is not running as root or exim. There may be
6231 a better way of dealing with this, but for the moment (because 4.30 needs
6232 to be released), I have disabled address rewrite logging when running a
6233 filter in a non-root, non-exim process.
6239 1. The buildconfig auxiliary program wasn't quoting the value set for
6240 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6241 not defined. This bug was masked in 4.22 by the effect that was fixed in
6244 2. Some messages that were rejected after a message id was allocated were
6245 shown as "incomplete" by exigrep. It no longer does this for messages that
6246 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6248 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6249 have allow_domain_literals set, the ID did not get logged in the <= line.
6250 Domain literals are now always recognized in Message-ID: header lines.
6252 4. The first argument for a ${extract expansion item is the key name or field
6253 number. Leading and trailing spaces in this item were not being ignored,
6254 causing some misleading effects.
6256 5. When deliver_drop_privilege was set, single queue runner processes started
6257 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6258 same command in the process it spins off) were not dropping privilege.
6260 6. When the daemon running as "exim" started a queue runner, it always
6261 re-executed Exim in the spun-off process. This is a waste of effort when
6262 deliver_drop_privilege is set. The new process now just calls the
6263 queue-runner function directly.
6269 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6272 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6273 ignored. Though the use of -oP was forcing the writing of a pid file, it
6274 was always written to the default place.
6276 3. If the message "no IP address found for host xxxx" is generated during
6277 incoming verification, it is now followed by identification of the incoming
6278 connection (so you can more easily find what provoked it).
6280 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6282 5. Added some features to "harden" Exim a bit more against certain attacks:
6284 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6285 be put in Local/Makefile. This is like the never_users runtime option,
6286 but it cannot be overridden. The default setting is "root".
6288 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6289 prefix string with which any file named in a -C command line option
6292 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6293 is retained for -C and -D only if the caller of Exim is root. Without
6294 it, the exim user may also use -C and -D and retain privilege.
6296 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6297 command line option is disabled.
6299 6. Macro names set by the -D option must start with an upper case letter, just
6300 like macro names defined in the configuration file.
6302 7. Added "dereference=" facility to LDAP.
6304 8. Two instances of the typo "uknown" in the source files are fixed.
6306 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6307 the Configure-Makefile script screwed up while processing it.
6309 10. Incorporated PCRE 4.4.
6311 11. The SMTP synchronization check was not operating right at the start of an
6312 SMTP session. For example, it could not catch a HELO sent before the client
6313 waited for the greeting. There is now a check for outstanding input at the
6314 point when the greeting is written. Because of the duplex, asynchronous
6315 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6316 way, but not yet received, when the check is performed.
6318 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6319 on TCP/IP sockets, because this apparently causes some broken clients to
6322 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6323 unchanged) from the Cygwin maintainer.
6325 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6326 was supported instead of testing for mbx. Effectively a typo.
6328 15. The spa authenticator server code was not checking that the input it
6329 received was valid base64.
6331 16. The debug output line for the "set" modifier in ACLs was not showing the
6332 name of the variable that was being set.
6334 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6336 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6337 lookup on its own. Something else was needed to trigger the lookup. For
6338 example, a match in host_lookup or the need for a host name in a host list.
6339 Now, if $sender_host_name is referenced and the host name has not yet been
6340 looked up, a lookup is performed. If the lookup fails, the variable remains
6341 empty, and $host_lookup_failed is set to "1".
6343 19. Added "eqi" as a case-independent comparison operator.
6345 20. The saslauthd authentication condition could segfault if neither service
6346 nor realm was specified.
6348 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6349 error message that was logged was misleading, and incoming SMTP
6350 connections were dropped. The message is now more accurate, and temporary
6351 errors are given to SMTP connections.
6353 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6354 (except RSET) with a 421 error, until QUIT is received. However, it was
6355 failing to send a response to QUIT.
6357 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6358 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6359 ACL is now run after the helo_try_verify_hosts code.
6361 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6362 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6363 case-independent, so other case variants are also recognized). Apparently
6364 some systems use these upper case variants.
6366 25. If more than two messages were waiting for the same host, and a transport
6367 filter was specified for the transport, Exim sent two messages over the
6368 same TCP/IP connection, and then failed with "socket operation on non-
6369 socket" when it tried to send the third.
6371 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6373 27. The extern definition of crypt16() in expand.c was not being excluded when
6374 the OS had its own crypt16() function.
6376 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6377 as a preferred synonym for return_size_limit, both as an option and as an
6380 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6382 30. Changed the default configuration ACL to relax the local part checking rule
6383 for addresses that are not in any local domains. For these addresses,
6384 slashes and pipe symbols are allowed within local parts, but the sequence
6385 /../ is explicitly forbidden.
6387 31. SPA server authentication was not clearing the challenge buffer before
6390 32. log_message in a "warn" ACL statement was writing to the reject log as
6391 well as to the main log, which contradicts the documentation and doesn't
6392 seem right (because no rejection is happening). So I have stopped it.
6394 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6395 However, I am unable to do any testing of this.
6397 34. Fixed an infelicity in the appendfile transport. When checking directories
6398 for a mailbox, to see if any needed to be created, it was accidentally
6399 using path names with one or more superfluous leading slashes; tracing
6400 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6402 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6403 contents are added to the log line that is written for every discarded
6404 recipient. (Previously a log_message setting was ignored.)
6406 36. The ${quote: operator now quotes the string if it is empty.
6408 37. The install script runs exim in order to find its version number. If for
6409 some reason other than non-existence or emptiness, which it checks, it
6410 could not run './exim', it was installing it with an empty version number,
6411 i.e. as "exim-". This error state is now caught, and the installation is
6414 38. An argument was missing from the function that creates an error message
6415 when Exim fails to connect to the socket for saslauthd authentication.
6416 This could cause Exim to crash, or give a corrupted message.
6418 39. Added isip, isip4, and isip6 to ${if conditions.
6420 40. The ACL variables $acl_xx are now saved with the message, and can be
6421 accessed later in routers, transports, and filters.
6423 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6424 strings in the file are not string-expanded.
6426 42. If a MAIL command specified a SIZE value that was too large to fit into an
6427 int variable, the check against message_size_limit failed. Such values are
6428 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6429 day this will have to be increased, but I don't think I want to be around
6430 when emails are that large.
6437 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6438 iconv() is not standard in FreeBSD.
6440 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6441 IPv6 enabled. The observed symptom was a segmentation fault on return from
6442 the function os_common_find_running_interfaces() in src/os.c.
6444 3. In the check_special_case() function in daemon.c I had used "errno" as an
6445 argument name, which causes warnings on some systems. This was basically a
6446 typo, since it was named "eno" in the comments!
6448 4. The code that waits for the clock to tick (at a resolution of some fraction
6449 of a second) so as to ensure message-id uniqueness was always waiting for
6450 at least one whole tick, when it could have waited for less. [This is
6451 almost certainly not relevant at current processor speeds, where it is
6452 unlikely to ever wait at all. But we try to future-proof.]
6454 5. The function that sleeps for a time interval that includes fractions of a
6455 second contained a race. It did not block SIGALRM between setting the
6456 timer, and suspending (a couple of lines later). If the interval was short
6457 and the sigsuspend() was delayed until after it had expired, the suspension
6458 never ended. On busy systems this could lead to processes getting stuck for
6461 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6462 process, before it forks any delivery processes. The open lookup caching
6463 mechanism meant that the open file or database connection was passed into
6464 the delivery process. The problem was that delivery processes always tidy
6465 up cached lookup data. This could cause a problem for the next delivery
6466 process started by the queue runner, because the external queue runner
6467 process does not know about the closure. So the next delivery process
6468 still has data in the lookup cache. In the case of a file lookup, there was
6469 no problem because closing a file descriptor in a subprocess doesn't affect
6470 the parent. However, if the lookup was caching a connection to a database,
6471 the connection was closed, and the second delivery process was likely to
6472 see errors such as "PGSQL: query failed: server closed the connection
6473 unexpectedly". The problem has been fixed by closing all cached lookups
6474 in a queue runner before running a delivery process.
6476 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6477 seem to have the "const" qualifier which it has on other OS. I've
6480 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6481 *of the same type* (client or server) with the same public name that an
6482 error should be diagnosed.
6484 9. When Exim looked up a host name for an IP address, but failed to find the
6485 original IP address when looking up the host name (a safety check), it
6486 output the message "<ip address> does not match any IP for NULL", which was
6487 confusing, to say the least. The bug was that the host name should have
6488 appeared instead of "NULL".
6490 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6491 user that is built into the binary, and the -C or -D options is used, root
6492 privilege is dropped before the configuration file is read. In addition,
6493 logging is switched to stderr instead of the normal log files. If the
6494 configuration then re-defines the Exim user, the unprivileged environment
6495 is probably not what is expected, so Exim logs a panic warning message (but
6498 However, if deliver_drop_privilege is set, the unprivileged state may well
6499 be exactly what is intended, so the warning has been cut out in that case,
6500 and Exim is allowed to try to write to its normal log files.
6506 1. smtp_return_error_details was not giving details for temporary sender
6507 or receiver verification errors.
6509 2. Diagnose a configuration error if two authenticators have the same public
6512 3. Exim used not to create the message log file for a message until the first
6513 delivery attempt. This could be confusing when incoming messages were held
6514 for policy or load reasons. The message log file is now created at the time
6515 the message is received, and an initial "Received" line is written to it.
6517 4. The automatically generated man page for command line options had a minor
6518 bug that caused no ill effects; however, a more serious problem was that
6519 the procedure for building the man page automatically didn't always
6520 operate. Consequently, release 4.20 contains an out-of-date version. This
6521 shouldn't happen again.
6523 5. When building Exim with embedded Perl support, the script that builds the
6524 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6525 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6527 6. The freeze_tell option was not being used for messages that were frozen on
6528 arrival, either by an ACL or by local_scan().
6530 7. Added the smtp_incomplete_transaction log selector.
6532 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6533 was accepting AUTH without a new EHLO.
6535 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6536 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6539 10. Logging of TCP/IP connections (when configured) now happens in the main
6540 daemon process instead of the child process, so that the TCP/IP connection
6541 count is more accurate (but it can never be perfect).
6543 11. The use of "drop" in a nested ACL was not being handled correctly in the
6544 outer ACL. Now, if condition failure induced by the nested "drop" causes
6545 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6546 or "require"), the connection is dropped.
6548 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6549 that yield "discard" can now be used with an "accept" or a "discard" verb,
6550 but an error is generated for any others (because I can't see a useful way
6551 to define what should happen).
6553 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6554 are now processed in the same way as lines in the Exim configuration file.
6555 In particular, continuation lines are supported.
6557 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6559 15. Added -ti meaning -t -i.
6561 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6562 domains, and warn by logging if others are found.
6564 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6565 returned from GIFCONF when Exim is trying to find the list of interfaces on
6566 a host. The code in os.c has been modified to copy each ifreq to an aligned
6567 structure in all cases.
6569 Also, in some cases, the returned ifreq's were being copied to a 'struct
6570 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6571 means the last couple of bytes of an IPv6 address could be chopped if the
6572 ifreq contained only a normal sockaddr (14 bytes storage).
6574 18. Named domain lists were not supported in the hosts_treat_as_local option.
6575 An entry such as +xxxx was not recognized, and was treated as a literal
6578 19. Ensure that header lines added by a DATA ACL are included in the reject log
6579 if the ACL subsequently rejects the message.
6581 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6582 MD5 (which is deprecated).
6584 21. When testing a filter file using -bf, Exim was writing a message when it
6585 took the sender from a "From " line in the message, but it was not doing so
6586 when it took $return_path from a Return-Path: header line. It now does.
6588 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6589 with a valid header line field name (a series of printing characters
6590 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6592 23. Changed "disc" in the source to "disk" to conform to the documentation and
6593 the book and for uniformity.
6595 24. Ignore Sendmail's -Ooption=value command line item.
6597 25. When execve() failed while trying to run a command in a pipe transport,
6598 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6599 could be confused with a return value of 69 from the command itself. This
6600 has been changed to 127, the value the shell returns if it is asked to run
6601 a non-existent command. The wording for the related log line suggests a
6602 non-existent command as the problem.
6604 26. If received_header_text expands to an empty string, do not add a Received:
6605 header line to the message. (Well, it adds a token one on the spool, but
6606 marks it "old" so that it doesn't get used or transmitted.)
6608 27. Installed eximstats 1.28 (addition of -nt option).
6610 28. There was no check for failure on the call to getsockname() in the daemon
6611 code. This can fail if there is a shortage of resources on the system, with
6612 ENOMEM, for example. A temporary error is now given on failure.
6614 29. Contrary to the C standard, it seems that in some environments, the
6615 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6616 program. Exim now does this explicitly; it affects the formatting of
6617 timestamps using strftime().
6619 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6620 complaints. I've now initialized all the variables, to avoid this.
6622 32. Header lines added by a system filter were not being "seen" during
6623 transport-time rewrites.
6625 33. The info_callback() function passed to OpenSSL is set up with type void
6626 (*)(SSL *, int, int), as described somewhere. However, when calling the
6627 function (actually a macro) that sets it up, the type void(*)() is
6628 expected. I've put in a cast to prevent warnings from picky compilers.
6630 34. If a DNS black list lookup found a CNAME record, but there were no A
6631 records associated with the domain it pointed at, Exim crashed.
6633 35. If a DNS black list lookup returned more than one A record, Exim ignored
6634 all but the first. It now scans all returned addresses if a particular IP
6635 value is being sought. In this situation, the contents of the
6636 $dnslist_value variable are a list of all the addresses, separated by a
6639 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6640 to accept a host name and all its aliases if the forward lookup for any of
6641 them yielded the IP address of the incoming connection. Now it accepts only
6642 those names whose forward lookup yields the correct IP address. Any other
6643 names are discarded. This closes a loophole whereby a rogue DNS
6644 administrator could create reverse DNS records to break through a
6645 wildcarded host restriction in an ACL.
6647 37. If a user filter or a system filter that ran in a subprocess used any of
6648 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6649 the wrong values were passed to the pipe command ($thisaddress had the
6650 value of $0, $0 had the value of $1, etc). This bug was introduced by
6651 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6653 38. Improved the line breaking for long SMTP error messages from ACLs.
6654 Previously, if there was no break point between 40 and 75 characters, Exim
6655 left the rest of the message alone. Two changes have been made: (a) I've
6656 reduced the minimum length to 35 characters; (b) if it can't find a break
6657 point between 35 and 75 characters, it looks ahead and uses the first one
6658 that it finds. This may give the occasional overlong line, but at least the
6659 remaining text gets split now.
6661 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6662 file descriptors might be low, and that setting 1000 would always raise it.
6663 It turns out that in some environments, the limit is already over 1000 and
6664 that lowering it causes trouble. So now Exim takes care not to decrease it.
6666 40. When delivering a message, the value of $return_path is set to $sender_
6667 address at the start of routing (routers may change the value). By an
6668 oversight, this default was not being set up when an address was tested by
6669 -bt or -bv, which affected the outcome if any router or filter referred to
6672 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6673 log only when "message" or "log_message" are set. However, if one of the
6674 conditions was an address verification, or a call to a nested ACL, the
6675 messages generated by the underlying test were being passed through. This
6676 no longer happens. The underlying message is available in $acl_verify_
6677 message for both "message" and "log_message" expansions, so it can be
6678 passed through if needed.
6680 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6681 new expansion $bh_ to give the encoded byte string without charset
6682 translation. Translation happens only if iconv() is available; HAVE_ICONV
6683 indicates this at build time. HEADERS_CHARSET gives the charset to
6684 translate to; headers_charset can change it in the configuration, and
6685 "headers charset" can change it in an individual filter file.
6687 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6688 that creates RFC 2047 encoded "words" labels them as that charset instead
6689 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6690 expansion operator; (ii) when Exim creates a From: line for a local
6691 message; (iii) when a header line is rewritten to include a "phrase" part.
6693 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6694 buggy, causing it to skip the first lines of messages whose message ID
6695 ended in 'D'. This would not have bitten before Exim release 4.14, because
6696 message IDs were unlikely to end in 'D' before then. The effect was to have
6697 incorrect size information for certain domains.
6699 45. #include "config.h" was missing at the start of the crypt16.c module. This
6700 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6703 46. If there was a timeout during a "random" callout check, Exim treated it as
6704 a failure of the random address, and carried on sending RSET and the real
6705 address. If the delay was just some slowness somewhere, the response to the
6706 original RCPT would be taken as a response to RSET and so on, causing
6707 mayhem of various kinds.
6709 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6710 when I implemented it. It didn't allow for the fact that some option values
6711 may legitimately be negative (e.g. size_addition), and it didn't even do
6712 the right test for positive values.
6714 48. Domain names in DNS records are case-independent. Exim always looks them up
6715 in lower case. Some resolvers return domain names in exactly the case they
6716 appear in the zone file, that is, they may contain uppercase letters. Not
6717 all resolvers do this - some return always lower case. Exim was treating a
6718 change of case by a resolver as a change of domain, similar to a widening
6719 of a domain abbreviation. This triggered its re-routing code and so it was
6720 trying to route what was effectively the same domain again. This normally
6721 caused routing to fail (because the router wouldn't handle the domain
6722 twice). Now Exim checks for this case specially, and just changes the
6723 casing of the domain that it ultimately uses when it transmits the message
6726 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6729 50. If a filter generated a file delivery with a non-absolute name (possible if
6730 no home directory exists for the router), the forbid_file option was not
6733 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6734 the existing equality matching.
6736 52. Exim was using ints instead of ino_t variables in some places where it was
6737 dealing with inode numbers.
6739 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6740 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6741 called TMPDIR, and if it finds it is different, it changes its value.
6743 54. The smtp_printf() function is now made available to local_scan() so
6744 additional output lines can be written before returning. There is also an
6745 smtp_fflush() function to enable the detection of a dropped connection.
6746 The variables smtp_input and smtp_batched_input are exported to
6749 55. Changed the default runtime configuration: the message "Unknown user"
6750 has been removed from the ACL, and instead placed on the localuser router,
6751 using the cannot_route_message feature. This means that any verification
6752 failures that generate their own messages won't get overridden. Similarly,
6753 the "Unrouteable address" message that was in the ACL for unverifiable
6754 relay addresses has also been removed.
6756 56. Added hosts_avoid_esmtp to the smtp transport.
6758 57. The exicyclog script was not checking for the esoteric option
6759 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6760 will work only if exicyclog is run under the appropriate euid.
6762 58. Following a discussion on the list, the rules by which Exim recognises line
6763 endings on incoming messages have been changed. The -dropcr and drop_cr
6764 options are now no-ops, retained only for backwards compatibility. The
6765 following line terminators are recognized: LF CRLF CR. However, special
6766 processing applies to CR:
6768 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6769 nor a local message in the state where . is a terminator.
6771 (ii) If a bare CR is encountered in a header line, an extra space is added
6772 after the line terminator so as not to end the header. The reasoning
6773 behind this is that bare CRs in header lines are most likely either
6774 to be mistakes, or people trying to play silly games.
6776 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6777 was being incorrectly given as 18 bytes larger than it should have been.
6778 This is a VOB (very old bug).
6780 60. This may never have affected anything current, but just in case it has:
6781 When the local host is found other than at the start of a list of hosts,
6782 the local host, those with the same MX, and any that follow, are discarded.
6783 When the list in question was part of a longer list of hosts, the following
6784 hosts (not currently being processed) were also being discarded. This no
6785 longer happens. I'm not sure if this situation could ever has previously
6788 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6791 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6792 header. This is something that is recommended in a new Internet Draft, and
6793 is something that is documented as being done by Sendmail. There are two
6794 possible values. For messages generated by the autoreply transport, Exim
6797 Auto-Submitted: auto-replied
6799 whereas for all other generated messages (e.g. bounces) it adds
6801 Auto-Submitted: auto-generated
6803 63. The "personal" condition in filters now includes a test for the
6804 Auto-Submitted: header. If it contains the string "auto-" the message it
6805 not considered personal.
6807 64. Added rcpt_include_affixes as a generic transport option.
6809 65. Added queue_only_override (default true).
6811 66. Added the syslog_duplication option.
6813 67. If what should have been the first header line of a message consisted of
6814 a space followed by a colon, Exim was mis-interpreting it as a header line.
6815 It isn't of course - it is syntactically invalid and should therefore be
6816 treated as the start of the message body. The misbehaviour could have
6817 caused a number of strange effects, including loss of data in subsequent
6818 header lines, and spool format errors.
6820 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6821 client host had authenticated. This control can now be exercised by an ACL
6822 for more flexibility.
6824 69. By default, callouts do not happen when testing with -bh. There is now a
6825 variant, -bhc, which does actually run the callout code, including
6826 consulting and updating the callout cache.
6828 70. Added support for saslauthd authentication, courtesy of Alexander
6831 71. If statvfs() failed on the spool or log directories while checking their
6832 size for availability, Exim confusingly gave the error "space shortage".
6833 Furthermore, in debugging mode it crashed with a floating point exception.
6834 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6835 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6836 serious problem, Exim now writes to the main and panic logs when this
6837 happens, with details of the failure. It then refuses to accept the
6838 incoming message, giving the message "spool directory problem" or "log
6839 directory problem" with a 421 code for SMTP messages.
6841 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6842 0, 1, and 2 exist, because some OS complain for execs without them (see
6843 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6844 descriptors. However, the code omitted to check that the open succeeded,
6845 causing mysterious errors if for some reason the permissions on /dev/null
6846 got screwed. Now Exim writes a message to the main and panic logs, and
6847 bombs out if it can't open /dev/null.
6849 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6850 interact so that it is all more flexible. It is supposed to remain
6851 backwards compatible. Also added extra_local_interfaces.
6853 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6854 to bomb out with an assertion failure - to the client this appears as a
6855 connection drop. This problem occurs in the part of the code that was taken
6856 from the Samba project. Fortunately, the assertion is in a very simple
6857 function, so I have fixed this by reproducing the function inline in the
6858 one place where it is called, and arranging for authentication to fail
6859 instead of killing the process with assert().
6861 75. The SPA client code was not working when the server requested OEM rather
6862 than Unicode encoding.
6864 76. Added code to make require_files with a specific uid setting more usable in
6865 the case where statting the file as root fails - usually a non-root-mounted
6866 NFS file system. When this happens and the failure is EACCES, Exim now
6867 forks a subprocess and does the per-uid checking as the relevant uid.
6869 77. Added process_log_path.
6871 78. If log_file_path was not explicitly set, a setting of check_log_space or
6872 check_log_inodes was ignored.
6874 79. If a space check for the spool or log partitions fails, the incident is now
6875 logged. Of course, in the latter case the data may get lost...
6877 80. Added the %p formatting code to string_format() so that it can be used to
6878 print addresses in debug_print(). Adjusted all the address printing in the
6879 debugging in store.c to use %p rather than %d.
6881 81. There was a concern that a line of code in smtp_in.c could overflow a
6882 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6883 initially expressed, the concern was not well-founded, because trailing
6884 spaces are removed early. However, if the trailing spaces were followed by
6885 a NULL, they did not get removed, so the overflow was possible. Two fixes
6888 (a) I re-wrote the offending code in a cleaner fashion.
6889 (b) If an incoming SMTP command contains a NULL character, it is rejected
6892 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6893 runs initgroups(), so that if the Exim user is in any additional groups,
6894 they will be used during message reception.
6900 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob