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.
26 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
27 than -2003 did; needs libidn2 in addition to libidn.
29 JH/02 The path option on a pipe transport is now expanded before use.
31 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
32 Patch provided by "Björn", documentation fix added too.
34 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
35 missing a wire-to-host endian conversion.
37 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
38 close after a BDAT command line could be taken as a following command,
39 giving a synch failure. Fix by only checking for synch immediately
40 before acknowledging the chunk.
42 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
43 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
44 macro. Patches provided by Josh Soref.
46 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
47 Previously we did not; the RFC seems ambiguous and VRFY is not listed
48 by IANA as a service extension. However, John Klensin suggests that we
51 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
52 the dkim code may be unix-mode line endings rather than smtp wire-format
53 CRLF, so prepend a CR to any bare LF.
55 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
56 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
58 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
59 the first were themselves being wrongly included in the feed into dkim
60 processing; with most chunk sizes in use this resulted in an incorrect
61 body hash calculated value.
63 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
64 DKIM signature block, for verification. Although advised against by
65 standards it is specifically not ruled illegal.
67 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
69 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
70 missing a body hash (the bh= tag).
72 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
73 It seems that HAProxy sends the Proxy Protocol information in clear and
74 only then does a TLS startup, so do the same.
76 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
77 TCP connections (such as for Spamd) unless the daemon successfully set
78 Fast Open mode on its listening sockets. This fixes breakage seen on
79 too-old kernels or those not configured for Fast Open, at the cost of
80 requiring both directions being enabled for TFO, and TFO never being used
81 by non-daemon-related Exim processes.
83 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
84 endings, at least on the first header line. Try to canonify any that get
85 past that check, despite the cost.
87 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
88 now limited to an arbitrary five deep, while parsing addresses with the
89 strip_excess_angle_brackets option enabled.
91 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
92 instead leave the unprompted TLS handshake in socket buffer for the
93 TLS library to consume.
95 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
97 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
99 JH/16 Drop variables when they go out of scope. Memory management drops a whole
100 region in one operation, for speed, and this leaves assigned pointers
101 dangling. Add checks run only under the testsuite which checks all
102 variables at a store-reset and panics on a dangling pointer; add code
103 explicitly nulling out all the variables discovered. Fixes one known
104 bug: a transport crash, where a dangling pointer for $sending_ip_address
105 originally assigned in a verify callout, is re-used.
107 PP/06 Drop '.' from @INC in various Perl scripts.
109 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
111 PP/08 Reduce a number of compilation warnings under clang; building with
112 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
113 should be warning-free.
115 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
117 HS/01 Fix portability problems introduced by PP/08 for platforms where
118 realloc(NULL) is not equivalent to malloc() [SunOS et al].
120 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
121 chunk. This allows us to accept broken chunked messages. We need a more
122 general solution here.
124 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
125 already-broken messages in the queue.
127 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
133 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
134 supports it and a size is available (ie. the sending peer gave us one).
136 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
137 years of being deprecated). The replacements are the ACLs
138 acl_smtp_mime and acl_not_smtp_mime.
140 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
141 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
142 or even an in-clear connection were permitted. Now, if the host lookup
143 was dnssec and dane was requested then the host is only used if the
144 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
145 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
146 if one fails this test.
147 This means that a poorly-configured remote DNS will make it incommunicado;
148 but it protects against a DNS-interception attack on it.
150 JH/04 Bug 1810: make continued-use of an open smtp transport connection
151 non-noisy when a race steals the message being considered.
153 JH/05 If main configuration option tls_certificate is unset, generate a
154 self-signed certificate for inbound TLS connections.
156 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
157 in rewrites and routers.
159 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
160 and logged a warning sing 4.83; now they are a configuration file error.
162 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
163 (lacking @domain). Apply the same qualification processing as RCPT.
165 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
167 JH/10 Support ${sha256:} applied to a string (as well as the previous
170 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
171 a cutthrough deliver is pending, as we always want to make a connection.
172 This also avoids re-routing the message when later placing the cutthrough
173 connection after a verify cache hit.
174 Do not update it with the verify result either.
176 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
177 when routing results in more than one destination address.
179 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
180 signing (which inhibits the cutthrough capability). Previously only
181 the presence of an option was tested; now an expansion evaluating as
182 empty is permissible (obviously it should depend only on data available
183 when the cutthrough connection is made).
185 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
186 the relevant preceding SMTP command did not note the pipelining mode.
188 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
189 Previously they were not counted.
191 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
192 as one having no matching records. Previously we deferred the message
193 that needed the lookup.
195 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
196 distinguished as "(=".
198 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
199 for missing MX records. Previously it only worked for missing A records.
201 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
203 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
204 after the data-go-ahead and data-ack. Patch from Jason Betts.
206 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
207 even for a "none" policy. Patch from Tony Meyer.
209 JH/22 Fix continued use of a connection for further deliveries. If a port was
210 specified by a router, it must also match for the delivery to be
213 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
214 When one of the recipients of a message was unsuitable for the connection
215 (has no matching addresses), we lost track of needing to mark it
216 deferred. As a result mail would be lost.
218 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
220 JH/25 Decoding ACL controls is now done using a binary search; the source code
221 takes up less space and should be simpler to maintain. Merge the ACL
222 condition decode tables also, with similar effect.
224 JH/26 Fix problem with one_time used on a redirect router which returned the
225 parent address unchanged. A retry would see the parent address marked as
226 delivered, so not attempt the (identical) child. As a result mail would
229 JH/27 Fix a possible security hole, wherein a process operating with the Exim
230 UID can gain a root shell. Credit to http://www.halfdog.net/ for
231 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
234 JH/28 Enable {spool,log} filesystem space and inode checks as default.
235 Main config options check_{log,spool}_{inodes,space} are now
236 100 inodes, 10MB unless set otherwise in the configuration.
238 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
239 Previously it only applied to the main-section connection policy
242 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
244 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
245 by me. Added RFC7919 DH primes as an alternative.
247 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
248 are not in the system include path.
250 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
251 GnuTLS, when a session startup failed (eg because the client disconnected)
252 Exim did stdio operations after fclose. This was exposed by a recent
253 change which nulled out the file handle after the fclose.
255 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
256 signed directly by the cert-signing cert, rather than an intermediate
257 OCSP-signing cert. This is the model used by LetsEncrypt.
259 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
261 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
262 an incoming connection.
264 HS/02 Bug 1802: Do not half-close the connection after sending a request
267 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
268 fallback to "prime256v1".
270 JH/34 SECURITY: Use proper copy of DATA command in error message.
271 Could leak key material. Remotely exploitable. CVE-2016-9963.
277 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
278 and 3.4.4 - once the server is enabled to respond to an OCSP request
279 it does even when not requested, resulting in a stapling non-aware
280 client dropping the TLS connection.
282 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
283 support variable-length bit vectors. No functional change.
285 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
286 The I= interface field on outgoing lines is now after the H= remote
287 host field, same as incoming lines. There is a separate
288 outgoing_interface log selector which allows you to disable the
291 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
292 If not running log_selector +smtp_connection the mainlog would be held
293 open indefinitely after a "too many connections" event, including to a
294 deleted file after a log rotate. Leave the per net connection logging
295 leaving it open for efficiency as that will be quickly detected by the
296 check on the next write.
298 HS/01 Bug 1671: Fix post transport crash.
299 Processing the wait-<transport> messages could crash the delivery
300 process if the message IDs didn't exist for some reason. When
301 using 'split_spool_directory=yes' the construction of the spool
302 file name failed already, exposing the same netto behaviour.
304 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
305 mime_regex ACL conditions.
307 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
308 to DSN fail messages (bounces): remote IP, remote greeting, remote response
309 to HELO, local diagnostic string.
311 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
312 log line to debug. Even when configured with a tls authenticator many
313 client connections are expected to not authenticate in this way, so
314 an authenticate fail is not an error.
316 HS/02 Add the Exim version string to the process info. This way exiwhat
317 gives some more detail about the running daemon.
319 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
320 matter for fast-change records such as DNSBLs.
322 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
323 retry record, even if constant. There may be multiple transports with
324 different interface settings and the retry behaviour needs to be kept
327 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
329 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
331 JH/10 Bug 840: fix log_defer_output option of pipe transport
333 JH/11 Bug 830: use same host for all RCPTS of a message, even under
334 hosts_randomize. This matters a lot when combined with mua_wrapper.
336 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
337 ${quote_pgsql:<string>} operator.
339 JH/13 Bug 1708: avoid misaligned access in cached lookup.
341 JH/14 Change header file name for freeradius-client. Relevant if compiling
342 with Radius support; from the Gentoo tree and checked under Fedora.
344 JH/15 Bug 1712: Introduce $prdr_requested flag variable
346 JH/16 Bug 1714: Permit an empty string as expansion result for transport
347 option transport_filter, meaning no filtering.
349 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
351 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
352 defaults to "*" (all hosts). The variable is now available when not built
353 with TLS, default unset, mainly to enable keeping the testsuite sane.
354 If a server certificate is not supplied (via tls_certificate) an error is
355 logged, and clients will find TLS connections fail on startup. Presumably
356 they will retry in-clear.
357 Packagers of Exim are strongly encouraged to create a server certificate
358 at installation time.
360 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
361 with the $config_file variable.
363 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
364 in transport context, after the attempt, and per-recipient. The latter type
365 is per host attempted. The event data is the error message, and the errno
366 information encodes the lookup type (A vs. MX) used for the (first) host,
367 and the trailing two digits of the smtp 4xx response.
369 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
370 to write to mainlog (or rejectlog, paniclog) in the window between file
371 creation and permissions/ownership being changed. Particularly affects
372 installations where exicyclog is run as root, rather than exim user;
373 result is that the running daemon panics and dies.
375 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
377 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
378 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
379 "pri" and "weight". Note that the previous implicit priority given by the
380 list order is no longer honoured.
382 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
385 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
386 by defining SUPPORT_SOCKS.
388 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
389 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
390 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
391 variables are renamed to proxy_{local,external}_{address,port}.
393 JH/27 Move Internationalisation support from Experimental to mainline, enabled
394 for a build by defining SUPPORT_I18N
396 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
397 of the query string, and make ${quote_redis:} do that quoting.
399 JH/29 Move Events support from Experimental to mainline, enabled by default
400 and removable for a build by defining DISABLE_EVENT.
402 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
404 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
405 cached by the daemon.
407 JH/32 Move Redis support from Experimental to mainline, enabled for a build
408 by defining LOOKUP_REDIS. The libhiredis library is required.
410 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
411 keys are given for lookup.
413 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
414 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
415 only supported when built with TLS support. The PolarSSL SHA routines
416 are still used when the TLS library is too old for convenient support.
418 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
419 openssl_options), for security. OpenSSL forces this from version 1.1.0
420 server-side so match that on older versions.
422 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
423 allocation for $value could be released as the expansion processing
424 concluded, but leaving the global pointer active for it.
426 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
427 and to use the domains and local_parts ACL conditions.
429 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
430 incorrectly not doubled on cutthrough transmission, hence seen as a
431 body-termination at the receiving system - resulting in truncated mails.
432 Commonly the sender saw a TCP-level error, and retransmitted the message
433 via the normal store-and-forward channel. This could result in duplicates
434 received - but deduplicating mailstores were liable to retain only the
435 initial truncated version.
437 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
439 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
441 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
442 we're in there, support oversigning also; bug 1309.
444 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
446 HS/04 Add support for keep_environment and add_environment options.
448 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
449 either intentional arithmetic overflow during PRNG, or testing config-
452 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
453 delivery resulted in actual delivery. Cancel cutthrough before DATA
456 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
457 rejects a recipient: pass the reject to the originator.
459 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
460 Many were false-positives and ignorable, but it's worth fixing the
463 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
464 for the new environment-manipulation done at startup. Move the routines
465 from being local to tls.c to being global via the os.c file.
467 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
468 an extract embedded as result-arg for a map, the first arg for extract
469 is unavailable so we cannot tell if this is a numbered or keyed
470 extraction. Accept either.
476 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
479 JH/02 The smtp transport option "multi_domain" is now expanded.
481 JH/03 The smtp transport now requests PRDR by default, if the server offers
484 JH/04 Certificate name checking on server certificates, when exim is a client,
485 is now done by default. The transport option tls_verify_cert_hostnames
486 can be used to disable this per-host. The build option
487 EXPERIMENTAL_CERTNAMES is withdrawn.
489 JH/05 The value of the tls_verify_certificates smtp transport and main options
490 default to the word "system" to access the system default CA bundle.
491 For GnuTLS, only version 3.0.20 or later.
493 JH/06 Verification of the server certificate for a TLS connection is now tried
494 (but not required) by default. The verification status is now logged by
495 default, for both outbound TLS and client-certificate supplying inbound
498 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
501 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
502 Status Notification (bounce) messages are now MIME format per RFC 3464.
503 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
504 under the control of the dsn_advertise_hosts option, and routers may
505 have a dsn_lasthop option.
507 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
508 default, modifiable by a malware= option. The list separator for
509 the options can now be changed in the usual way. Bug 68.
511 JH/10 The smtp_receive_timeout main option is now expanded before use.
513 JH/11 The incoming_interface log option now also enables logging of the
514 local interface on delivery outgoing connections.
516 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
517 if the interface and destination host and port all match.
519 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
522 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
523 Patch from Andrew Lewis.
525 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
526 now supports optional time-restrictions, weighting, and priority
527 modifiers per server. Patch originally by <rommer@active.by>.
529 JH/16 The spamd_address main option now supports a mixed list of local
530 and remote servers. Remote servers can be IPv6 addresses, and
531 specify a port-range.
533 JH/17 Bug 68: The spamd_address main option now supports an optional
534 timeout value per server.
536 JH/18 Bug 1581: Router and transport options headers_add/remove can
537 now have the list separator specified.
539 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
542 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
545 JH/21 Support for the A6 type of dns record is withdrawn.
547 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
548 rather than the verbs used.
550 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
551 from 255 to 1024 chars.
553 JH/24 Verification callouts now attempt to use TLS by default.
555 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
556 are generic router options now. The defaults didn't change.
558 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
559 Original patch from Alexander Shikoff, worked over by JH.
561 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
562 files if ZCAT_COMMAND is not executable.
564 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
566 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
568 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
569 Normally benign, it bites when the pair was led to by a CNAME;
570 modern usage is to not canonicalize the domain to a CNAME target
571 (and we were inconsistent anyway for A-only vs AAAA+A).
573 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
575 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
576 when evaluating $sender_host_dnssec.
578 JH/31 Check the HELO verification lookup for DNSSEC, adding new
579 $sender_helo_dnssec variable.
581 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
583 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
585 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
587 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
588 documented as working, but never had. Support all but $spam_report.
590 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
591 added for tls authenticator.
593 HS/03 Add perl_taintmode main config option
599 TL/01 When running the test suite, the README says that variables such as
600 no_msglog_check are global and can be placed anywhere in a specific
601 test's script, however it was observed that placement needed to be near
602 the beginning for it to behave that way. Changed the runtest perl
603 script to read through the entire script once to detect and set these
604 variables, reset to the beginning of the script, and then run through
605 the script parsing/test process like normal.
607 TL/02 The BSD's have an arc4random API. One of the functions to induce
608 adding randomness was arc4random_stir(), but it has been removed in
609 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
610 function when detected.
612 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
613 cause callback expansion.
615 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
616 syntax errors in an expansion can be treated as a string instead of
617 logging or causing an error, due to the internal use of bool_lax
618 instead of bool when processing it.
620 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
621 server certificates when making smtp deliveries.
623 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
625 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
627 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
629 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
630 Merged patch from Sebastian Wiedenroth.
632 JH/05 Fix results-pipe from transport process. Several recipients, combined
633 with certificate use, exposed issues where response data items split
634 over buffer boundaries were not parsed properly. This eventually
635 resulted in duplicates being sent. This issue only became common enough
636 to notice due to the introduction of connection certificate information,
637 the item size being so much larger. Found and fixed by Wolfgang Breyha.
639 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
640 size buffer was used, resulting in syntax errors when an expansion
643 JH/07 Add support for directories of certificates when compiled with a GnuTLS
644 version 3.3.6 or later.
646 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
647 is EXPERIMENTAL_EVENT, the main-configuration and transport options
648 both become "event_action", the variables become $event_name, $event_data
649 and $event_defer_errno. There is a new variable $verify_mode, usable in
650 routers, transports and related events. The tls:cert event is now also
651 raised for inbound connections, if the main configuration event_action
654 TL/06 In test suite, disable OCSP for old versions of openssl which contained
655 early OCSP support, but no stapling (appears to be less than 1.0.0).
657 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
658 server certificate names available under the smtp transport option
659 "tls_verify_cert_hostname" now do not permit multi-component wildcard
662 JH/10 Time-related extraction expansions from certificates now use the main
663 option "timezone" setting for output formatting, and are consistent
664 between OpenSSL and GnuTLS compilations. Bug 1541.
666 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
667 encoded parameter in the incoming message. Bug 1558.
669 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
670 include certificate info, eximon was claiming there were spoolfile
673 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
675 JH/14 Log delivery-related information more consistently, using the sequence
676 "H=<name> [<ip>]" wherever possible.
678 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
679 are problematic for Debian distribution, omit them from the release
682 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
684 JH/16 Fix string representation of time values on 64bit time_t architectures.
687 JH/17 Fix a null-indirection in certextract expansions when a nondefault
688 output list separator was used.
693 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
694 checkers that were complaining about end of non-void function with no
697 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
698 This was a regression introduced in 4.83 by another bugfix.
700 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
702 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
703 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
709 TF/01 Correctly close the server side of TLS when forking for delivery.
711 When a message was received over SMTP with TLS, Exim failed to clear up
712 the incoming connection properly after forking off the child process to
713 deliver the message. In some situations the subsequent outgoing
714 delivery connection happened to have the same fd number as the incoming
715 connection previously had. Exim would try to use TLS and fail, logging
716 a "Bad file descriptor" error.
718 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
719 utilities have not been installed.
721 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
722 temporary space as the ACL may create new global variables.
724 TL/01 LDAP support uses per connection or global context settings, depending
725 upon the detected version of the libraries at build time.
727 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
728 to extract and use the src ip:port in logging and expansions as if it
729 were a direct connection from the outside internet. PPv2 support was
730 updated based on HAProxy spec change in May 2014.
732 JH/02 Add ${listextract {number}{list}{success}{fail}}.
734 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
735 Properly escape header and check for NULL return.
737 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
740 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
742 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
743 characters in header names, implemented as a verify condition.
744 Contributed by Michael Fischer v. Mollard.
746 TL/05 Rename SPF condition results err_perm and err_temp to standardized
747 results permerror and temperror. Previous values are deprecated but
748 still accepted. In a future release, err_perm and err_temp will be
749 completely removed, which will be a backward incompatibility if the
750 ACL tests for either of these two old results. Patch contributed by
751 user bes-internal on the mailing list.
753 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
755 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
756 selectors, in both main and reject logs.
758 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
761 JH/07 Add malware type "sock" for talking to simple daemon.
763 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
765 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
766 routers/transports under cutthrough routing.
768 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
769 numbers. Touch up "bool" conditional to keep the same definition.
771 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
773 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
775 JH/12 Expand items in router/transport headers_add or headers_remove lists
776 individually rather than the list as a whole. Bug 1452.
778 Required for reasonable handling of multiple headers_ options when
779 they may be empty; requires that headers_remove items with embedded
780 colons must have them doubled (or the list-separator changed).
782 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
783 view the policy declared in the DMARC record. Currently, $dmarc_status
784 is a combined value of both the record presence and the result of the
787 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
789 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
790 dnslookup router and the smtp transport (applying to the forward
793 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
794 of ldap servers used for a specific lookup. Patch provided by Heiko
797 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
798 New variable $lookup_dnssec_authenticated for observability.
800 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
801 Patch submitted by Lars Timman.
803 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
805 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
806 Requires trusted mode and valid format message id, aborts otherwise.
807 Patch contributed by Heiko Schlichting.
809 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
810 certextract with support for various fields. Bug 1358.
812 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
813 is requested by default, modifiable by smtp transport option
816 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
817 operate on certificate variables to give certificate fingerprints
818 Also new ${sha256:cert_variable}.
820 JH/23 The PRDR feature is moved from being Experimental into the mainline.
822 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
823 Christian Aistleitner.
825 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
827 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
828 file. Patch from Wolfgang Breyha.
830 JH/25 Expand the coverage of the delivery $host and $host_address to
831 client authenticators run in verify callout. Bug 1476.
833 JH/26 Port service names are now accepted for tls_on_connect_ports, to
834 align with daemon_smtp_ports. Bug 72.
836 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
837 support and error reporting did not work properly.
839 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
840 and is readable. Patch from Andrew Colin Kissa.
842 TL/14 Enhance documentation of ${run expansion and how it parses the
843 commandline after expansion, particularly in the case when an
844 unquoted variable expansion results in an empty value.
846 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
848 PP/02 Fix internal collision of T_APL on systems which support RFC3123
849 by renaming away from it. Addresses GH issue 15, reported by
852 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
854 TL/15 SECURITY: prevent double expansion in math comparison functions
855 (can expand unsanitized data). Not remotely exploitable.
862 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
864 PP/02 Make -n do something, by making it not do something.
865 When combined with -bP, the name of an option is not output.
867 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
870 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
871 $sender_host_name and config options to manage this, and basic check
874 PP/05 DSCP support for outbound connections and control modifier for inbound.
876 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
877 (Only plugin which currently uses this is kerberos4, which nobody should
878 be using, but we should make it available and other future plugins might
879 conceivably use it, even though it would break NAT; stuff *should* be
880 using channel bindings instead).
882 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
883 name; added for Sendmail compatibility; requires admin caller.
884 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
885 just ignore it); requires trusted caller.
886 Also parse but ignore: -Ac -Am -X<logfile>
889 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
891 TL/02 Add +smtp_confirmation as a default logging option.
893 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
894 Patch by Magnus Holmgren from 2007-02-20.
896 TL/04 Bugzilla 1281 - Spec typo.
897 Bugzilla 1283 - Spec typo.
898 Bugzilla 1290 - Spec grammar fixes.
900 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
902 TL/06 Add Experimental DMARC support using libopendmarc libraries.
904 TL/07 Fix an out of order global option causing a segfault. Reported to dev
905 mailing list by by Dmitry Isaikin.
907 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
909 JH/02 Support "G" suffix to numbers in ${if comparisons.
911 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
913 NM/01 Bugzilla 1197 - Spec typo
914 Bugzilla 1196 - Spec examples corrections
916 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
918 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
919 gnutls_enable_pkcs11, but renamed to more accurately indicate its
922 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
923 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
925 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
926 "acl {{name}{arg}...}", and optional args on acl condition
929 JH/05 Permit multiple router/transport headers_add/remove lines.
931 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
933 JH/07 Avoid using a waiting database for a single-message-only transport.
934 Performance patch from Paul Fisher. Bugzilla 1262.
936 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
939 JH/09 Add $headers_added variable, with content from use of ACL modifier
940 add_header (but not yet added to the message). Bugzilla 199.
942 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
943 Pulled from Bugzilla 817 by Wolfgang Breyha.
945 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
947 (nb: this is the same fix as in Exim 4.80.1)
949 JH/11 Add A= logging on delivery lines, and a client_set_id option on
952 JH/12 Add optional authenticated_sender logging to A= and a log_selector
955 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
957 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
958 advertise SMTP AUTH mechanism to us, instead of a generic
959 protocol violation error. Also, make Exim more robust to bad
960 data from the Dovecot auth socket.
962 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
964 When a queue runner is handling a message, Exim first routes the
965 recipient addresses, during which it prunes them based on the retry
966 hints database. After that it attempts to deliver the message to
967 any remaining recipients. It then updates the hints database using
970 So if a recipient address works intermittently, it can get repeatedly
971 deferred at routing time. The retry hints record remains fresh so the
972 address never reaches the final cutoff time.
974 This is a fairly common occurrence when a user is bumping up against
975 their storage quota. Exim had some logic in its local delivery code
976 to deal with this. However it did not apply to per-recipient defers
977 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
979 This change adds a proper retry rule check during routing so that the
980 final cutoff time is checked against the message's age. We only do
981 this check if there is an address retry record and there is not a
982 domain retry record; this implies that previous attempts to handle
983 the address had the retry_use_local_parts option turned on. We use
984 this as an approximation for the destination being like a local
985 delivery, as in LMTP.
987 I suspect this new check makes the old local delivery cutoff check
988 redundant, but I have not verified this so I left the code in place.
990 TF/02 Correct gecos expansion when From: is a prefix of the username.
992 Test 0254 submits a message to Exim with the header
996 When I ran the test suite under the user fanf2, Exim expanded
997 the header to contain my full name, whereas it should have added
998 a Resent-Sender: header. It erroneously treats any prefix of the
999 username as equal to the username.
1001 This change corrects that bug.
1003 GF/01 DCC debug and logging tidyup
1004 Error conditions log to paniclog rather than rejectlog.
1005 Debug lines prefixed by "DCC: " to remove any ambiguity.
1007 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1009 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1010 Bug spotted by Jeremy Harris; was flawed since initial commit.
1011 Would have resulted in OCSP responses post-SNI triggering an Exim
1012 NULL dereference and crash.
1014 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1016 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1017 Bug detection, analysis and fix by Samuel Thibault.
1018 Bugzilla 1331, Debian bug #698092.
1020 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1022 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1023 Server implementation by Todd Lyons, client by JH.
1024 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1025 config variable "prdr_enable" controls whether the server
1026 advertises the facility. If the client requests PRDR a new
1027 acl_data_smtp_prdr ACL is called once for each recipient, after
1028 the body content is received and before the acl_smtp_data ACL.
1029 The client is controlled by both of: a hosts_try_prdr option
1030 on the smtp transport, and the server advertisement.
1031 Default client logging of deliveries and rejections involving
1032 PRDR are flagged with the string "PRDR".
1034 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1035 fclose(). Diagnosis by Todd Lyons.
1037 PP/17 Update configure.default to handle IPv6 localhost better.
1038 Patch by Alain Williams (plus minor tweaks).
1041 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1042 This is now consistent with GnuTLS, and is now documented: the
1043 previous undocumented portable approach to treating the option as
1044 unset was to force an expansion failure. That still works, and
1045 an empty string is now equivalent.
1047 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1048 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1049 not performing validation itself.
1051 PP/20 Added force_command boolean option to pipe transport.
1052 Patch from Nick Koston, of cPanel Inc.
1054 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1057 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1059 PP/21 Fix eximon continuous updating with timestamped log-files.
1060 Broken in a format-string cleanup in 4.80, missed when I repaired the
1061 other false fix of the same issue.
1062 Report and fix from Heiko Schlichting.
1065 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1066 Report from Prashanth Katuri.
1068 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1069 It's SecureTransport, so affects any MacOS clients which use the
1070 system-integrated TLS libraries, including email clients.
1072 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1073 using a MIME ACL for non-SMTP local injection.
1074 Report and assistance in diagnosis by Warren Baker.
1076 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1078 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1080 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1081 last id that failed so it may be referenced in subsequent ACL's.
1083 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1086 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1087 ldap library initialization, allowing self-signed CA's to be
1088 used. Also properly sets require_cert option later in code by
1089 using NULL (global ldap config) instead of ldap handle (per
1090 session). Bug diagnosis and testing by alxgomz.
1092 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1093 the src/util/ subdirectory.
1095 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1096 renamed to Transport Post Delivery Action by Jeremy Harris, as
1099 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1100 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1101 redis_servers = needs to be configured which will be used by the redis
1102 lookup. Patch from Warren Baker, of The Packet Hub.
1104 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1106 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1107 hostname or reverse DNS when processing a host list. Used suggestions
1108 from multiple comments on this bug.
1110 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1112 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1113 Missed a few lines, added it to make the runtest require no keyboard
1116 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1117 contains upper case chars. Make router use caseful_local_part.
1119 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1120 support when GnuTLS has been built with p11-kit.
1126 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1128 This, or similar/improved, will also be change PP/11 of 4.82.
1134 PP/01 Handle short writes when writing local log-files.
1135 In practice, only affects FreeBSD (8 onwards).
1136 Bugzilla 1053, with thanks to Dmitry Isaikin.
1138 NM/01 Bugzilla 949 - Documentation tweak
1140 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1143 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1145 PP/02 Implemented gsasl authenticator.
1147 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1149 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1150 `pkg-config foo` for cflags/libs.
1152 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1153 with rest of GSASL and with heimdal_gssapi.
1155 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1156 `pkg-config foo` for cflags/libs for the TLS implementation.
1158 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1159 properties get this fed in as external SSF. A number of robustness
1160 and debugging improvements to the cyrus_sasl authenticator.
1162 PP/08 cyrus_sasl server now expands the server_realm option.
1164 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1165 Patch by Jeremy Harris.
1167 PP/10 Added dbmjz lookup type.
1169 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1171 PP/12 MAIL args handles TAB as well as SP, for better interop with
1172 non-compliant senders.
1173 Analysis and variant patch by Todd Lyons.
1175 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1176 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1177 Patch from Dirk Mueller <dmueller@suse.com>
1179 PP/13 tls_peerdn now print-escaped for spool files.
1180 Observed some $tls_peerdn in wild which contained \n, which resulted
1181 in spool file corruption.
1183 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1184 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1185 or write after TLS renegotiation, which otherwise led to messages
1188 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1189 as a tracking header (ie: a signed header comes before the signature).
1190 Patch from Wolfgang Breyha.
1192 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1193 comma-sep list; embedded commas doubled.
1195 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1197 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1199 Report and patch from Dmitry Banschikov.
1201 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1202 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1203 failures. We appear to now support TLS1.1+ with Exim.
1205 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1206 lets Exim select keys and certificates based upon TLS SNI from client.
1207 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1208 before an outbound SMTP session. New log_selector, +tls_sni.
1210 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1211 NULL dereference. Report and patch from Alun Jones.
1213 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1214 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1215 Not seeing resolver debug output on NetBSD, but suspect this is a
1216 resolver implementation change.
1218 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1219 Left warnings. Added "eximon gdb" invocation mode.
1221 PP/21 Defaulting "accept_8bitmime" to true, not false.
1223 PP/22 Added -bw for inetd wait mode support.
1225 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1226 locate the relevant includes and libraries. Made this the default.
1228 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1229 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1231 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1232 This may cause build issues on older platforms.
1234 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1235 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1236 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1237 Added SNI support via GnuTLS too.
1238 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1240 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1242 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1243 Applied second patch from Janne, implementing suggestion to default
1244 multiple-strings-in-record handling to match SPF spec.
1246 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1248 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1249 read-only, out of scope).
1250 Patch from Wolfgang Breyha, report from Stuart Northfield.
1252 PP/29 Fix three issues highlighted by clang analyser static analysis.
1253 Only crash-plausible issue would require the Cambridge-specific
1254 iplookup router and a misconfiguration.
1255 Report from Marcin Mirosław.
1257 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1259 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1260 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1261 As part of this, removing so much warning spew let me fix some minor
1262 real issues in debug logging.
1264 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1265 assignment on my part. Fixed.
1267 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1268 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1269 Janne Snabb (who went above and beyond: thank you).
1271 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1272 string otherwise requires a connection and a bunch more work and it's
1273 relatively easy to get wrong. Should also expose TLS library linkage
1276 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1277 64-bit ${eval} (JH/03).
1279 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1280 GNU libc to support some of the 64-bit stuff, should not lead to
1281 conflicts. Defined before os.h is pulled in, so if a given platform
1282 needs to override this, it can.
1284 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1285 protection layer was required, which is not implemented.
1286 Bugzilla 1254, patch from Wolfgang Breyha.
1288 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1289 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1290 tls_dhparam take prime identifiers. Also unbreak combination of
1291 OpenSSL+DH_params+TLSSNI.
1293 PP/39 Disable SSLv2 by default in OpenSSL support.
1299 PP/01 Solaris build fix for Oracle's LDAP libraries.
1300 Bugzilla 1109, patch from Stephen Usher.
1302 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1304 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1307 TF/02 Fix a couple more cases where we did not log the error message
1308 when unlink() failed. See also change 4.74-TF/03.
1310 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1311 lock up or crash if it happened to be inside a call to libc when it
1312 got a SIGUSR1 from exiwhat.
1314 The SIGUSR1 handler appends the current process status to the process
1315 log which is later printed by exiwhat. It used to use the general
1316 purpose logging code to do this, but several functions it calls are
1317 not safe for signals.
1319 The new output code in the SIGUSR1 handler is specific to the process
1320 log, and simple enough that it's easy to inspect for signal safety.
1321 Removing some special cases also simplifies the general logging code.
1322 Removing the spurious timestamps from the process log simplifies
1325 TF/04 Improved ratelimit ACL condition.
1327 The /noupdate option has been deprecated in favour of /readonly which
1328 has clearer semantics. The /leaky, /strict, and /readonly update modes
1329 are mutually exclusive. The update mode is no longer included in the
1330 database key; it just determines when the database is updated. (This
1331 means that when you upgrade Exim will forget old rate measurements.)
1333 Exim now checks that the per_* options are used with an update mode that
1334 makes sense for the current ACL. For example, when Exim is processing a
1335 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1336 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1337 must specify per_mail/readonly. If you omit the update mode it defaults to
1338 /leaky where that makes sense (as before) or /readonly where required.
1340 The /noupdate option is now undocumented but still supported for
1341 backwards compatibility. It is equivalent to /readonly except that in
1342 ACLs where /readonly is required you may specify /leaky/noupdate or
1343 /strict/noupdate which are treated the same as /readonly.
1345 A useful new feature is the /count= option. This is a generalization
1346 of the per_byte option, so that you can measure the throughput of other
1347 aggregate values. For example, the per_byte option is now equivalent
1348 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1350 The per_rcpt option has been generalized using the /count= mechanism
1351 (though it's more complicated than the per_byte equivalence). When it is
1352 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1353 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1354 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1355 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1356 Note that using per_rcpt with a non-readonly update mode in more than
1357 one ACL will cause the recipients to be double-counted. (The per_mail
1358 and per_byte options don't have this problem.)
1360 The handling of very low rates has changed slightly. If the computed rate
1361 is less than the event's count (usually one) then this event is the first
1362 after a long gap. In this case the rate is set to the same as this event's
1363 count, so that the first message of a spam run is counted properly.
1365 The major new feature is a mechanism for counting the rate of unique
1366 events. The new per_addr option counts the number of different
1367 recipients that someone has sent messages to in the last time period. It
1368 behaves like per_rcpt if all the recipient addresses are different, but
1369 duplicate recipient addresses do not increase the measured rate. Like
1370 the /count= option this is a general mechanism, so the per_addr option
1371 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1372 example, measure the rate that a client uses different sender addresses
1373 with the options per_mail/unique=$sender_address. There are further
1374 details in the main documentation.
1376 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1378 TF/06 Removed a few PCRE remnants.
1380 TF/07 Automatically extract Exim's version number from tags in the git
1381 repository when doing development or release builds.
1383 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1384 Bugzilla 879. Patch from Paul Fisher.
1386 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1387 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1390 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1392 PP/05 Variable $av_failed, true if the AV scanner deferred.
1393 Bugzilla 1078. Patch from John Horne.
1395 PP/06 Stop make process more reliably on build failure.
1396 Bugzilla 1087. Patch from Heiko Schlittermann.
1398 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1399 Bugzilla 1089. Patch from Heiko Schlittermann.
1401 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1402 Bugzilla 1131. Patch from Holger Weiß.
1404 PP/09 Handle IPv6 addresses with SPF.
1405 Bugzilla 860. Patch from Wolfgang Breyha.
1407 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1409 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1412 PP/11 match_* no longer expand right-hand-side by default.
1413 New compile-time build option, EXPAND_LISTMATCH_RHS.
1414 New expansion conditions, "inlist", "inlisti".
1416 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1418 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1420 PP/14 fix log_write() format string regression from TF/03.
1421 Bugzilla 1152. Patch from Dmitry Isaikin.
1427 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1429 PP/02 Harmonised TLS library version reporting; only show if debugging.
1430 Layout now matches that introduced for other libraries in 4.74 PP/03.
1432 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1434 PP/04 New "dns_use_edns0" global option.
1436 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1439 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1440 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1442 TK/01 Updated PolarSSL code to 0.14.2.
1443 Bugzilla 1097. Patch from Andreas Metzler.
1445 PP/07 Catch divide-by-zero in ${eval:...}.
1446 Fixes bugzilla 1102.
1448 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1451 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1452 format-string attack -- SECURITY: remote arbitrary code execution.
1454 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1455 time unintentionally subject to list matching rules, letting the header
1456 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1457 arbitrary string expansion). This allowed for information disclosure.
1459 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1460 INT_MIN/-1 -- value coerced to INT_MAX.
1466 NM/01 Workaround for PCRE version dependency in version reporting
1469 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1470 This fixes portability to compilers other than gcc, notably
1471 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1473 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1474 makefiles for portability to HP-UX and POSIX correctness.
1476 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1477 Also via indented variable definition in the Makefile.
1478 (Debugging by Oliver Heesakkers).
1480 PP/02 Restore caching of spamd results with expanded spamd_address.
1481 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1483 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1484 Improves build reliability. Fix from: Frank Elsner
1486 NM/02 Fix wide character breakage in the rfc2047 coding
1487 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1489 NM/03 Allow underscore in dnslist lookups
1490 Fixes bug 1026. Patch from Graeme Fowler
1492 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1493 Code patches from Adam Ciarcinski of NetBSD.
1495 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1498 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1499 is logged, to avoid truncation. Patch from John Horne.
1501 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1502 Patch from Jakob Hirsch.
1504 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1505 SQL string expansion failure details.
1506 Patch from Andrey Oktyabrski.
1508 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1509 Patch from Simon Arlott.
1511 PP/09 New lookups functionality failed to compile on old gcc which rejects
1512 extern declarations in function scope.
1513 Patch from Oliver Fleischmann
1515 PP/10 Use sig_atomic_t for flags set from signal handlers.
1516 Check getgroups() return and improve debugging.
1517 Fixed developed for diagnosis in bug 927 (which turned out to be
1520 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1521 Patch from Mark Zealey.
1523 PP/12 Bugzilla 1056: Improved spamd server selection.
1524 Patch from Mark Zealey.
1526 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1527 Based on patch from Heiko Schlittermann.
1529 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1530 Patch from Uwe Doering, sign-off by Michael Haardt.
1532 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1533 variable declaration deep within a block. Bug and patch from
1536 PP/15 lookups-Makefile IRIX compatibility coercion.
1538 PP/16 Make DISABLE_DKIM build knob functional.
1540 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1541 Patch by Simon Arlott
1543 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1544 variable argument macros. Our copy now differs from upstream.
1550 TF/01 Failure to get a lock on a hints database can have serious
1551 consequences so log it to the panic log.
1553 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1554 controlled using the smtp_confirmation log selector.
1556 TF/03 Include the error message when we fail to unlink a spool file.
1558 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1559 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1560 for maintaining out-of-tree patches for some time.
1562 PP/01 Bugzilla 139: Documentation and portability issues.
1563 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1564 Handle per-OS dynamic-module compilation flags.
1566 PP/02 Let /dev/null have normal permissions.
1567 The 4.73 fixes were a little too stringent and complained about the
1568 permissions on /dev/null. Exempt it from some checks.
1569 Reported by Andreas M. Kirchwitz.
1571 PP/03 Report version information for many libraries, including
1572 Exim version information for dynamically loaded libraries. Created
1573 version.h, now support a version extension string for distributors
1574 who patch heavily. Dynamic module ABI change.
1576 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1577 privilege escalation vulnerability whereby the Exim run-time user
1578 can cause root to append content of the attacker's choosing to
1581 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1584 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1585 If dropping privileges for untrusted macros, we disabled normal logging
1586 on the basis that it would fail; for the Exim run-time user, this is not
1587 the case, and it resulted in successful deliveries going unlogged.
1588 Fixed. Reported by Andreas Metzler.
1594 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1595 only prepend for the Resent-* case. Fixes regression introduced in
1596 Exim 4.70 by NM/22 for Bugzilla 607.
1598 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1599 increasing numbers of administrators be bitten by this.
1601 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1603 PP/03 Bugzilla 994: added openssl_options main configuration option.
1605 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1607 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1609 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1611 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1612 without a peer certificate, leading to a segfault because of an
1613 assumption that peers always have certificates. Be a little more
1614 paranoid. Problem reported by Martin Tscholak.
1616 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1617 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1618 NB: ClamAV planning to remove STREAM in "middle of 2010".
1619 CL also introduces -bmalware, various -d+acl logging additions and
1620 more caution in buffer sizes.
1622 PP/09 Implemented reverse_ip expansion operator.
1624 PP/10 Bugzilla 937: provide a "debug" ACL control.
1626 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1628 PP/12 Bugzilla 973: Implement --version.
1630 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1632 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1634 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1636 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1637 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1638 ignore trailing whitespace.
1640 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1642 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1645 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1646 Notification from Dr Andrew Aitchison.
1648 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1649 ExtendedDetectionInfo response format.
1650 Notification from John Horne.
1652 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1655 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1656 XSL and documented dependency on system catalogs, with examples of how
1659 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1662 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1663 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1664 configuration file which is writeable by the Exim user or group.
1666 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1667 of configuration files to cover files specified with the -C option if
1668 they are going to be used with root privileges, not just the default
1671 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1672 option (effectively making it always true).
1674 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1675 files to be used while preserving root privileges.
1677 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1678 that rogue child processes cannot use them.
1680 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1681 run-time user, instead of root.
1683 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1684 Exim run-time user without dropping privileges.
1686 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1687 result string, instead of calling string_vformat() twice with the same
1690 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1691 for other users. Others should always drop root privileges if they use
1692 -C on the command line, even for a whitelisted configure file.
1694 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1696 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1702 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1703 $data_path, and $header_path variables; fixed documentation bugs and
1706 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1707 exipick to access non-standard spools, including the "frozen" queue
1710 NM/01 Bugzilla 965: Support mysql stored procedures.
1711 Patch from Alain Williams
1713 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1715 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1716 Patch from Andreas Metzler
1718 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1719 Patch from Kirill Miazine
1721 NM/05 Bugzilla 671: Added umask to procmail example.
1723 JJ/03 installed exipick 20100323.0, fixing doc bug
1725 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1726 directory. Notification and patch from Dan Rosenberg.
1728 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1730 TK/02 Improve log output when DKIM signing operation fails.
1732 MH/01 Treat the transport option dkim_domain as a colon separated
1733 list, not as a single string, and sign the message with each element,
1734 omitting multiple occurences of the same signer.
1736 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1737 Bugzilla 985, 986. Patch by Simon Arlott
1739 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1740 Patch by Simon Arlott
1742 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1743 MBX locking. Notification from Dan Rosenberg.
1749 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1751 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1753 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1755 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1757 NM/04 Bugzilla 915: Flag broken perl installation during build.
1763 TK/01 Added patch by Johannes Berg that expands the main option
1764 "spamd_address" if it starts with a dollar sign.
1766 TK/02 Write list of recipients to X-Envelope-Sender header when building
1767 the mbox-format spool file for content scanning (suggested by Jakob
1770 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1771 (http://www.dcc-servers.net/) support via dccifd. Activated by
1772 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1774 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1775 by Mark Daniel Reidel <mr@df.eu>.
1777 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1778 When building exim an external PCRE library is now needed -
1779 PCRE is a system library on the majority of modern systems.
1780 See entry on PCRE_LIBS in EDITME file.
1782 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1783 conversation. Added nologin parameter to request.
1784 Patch contributed by Kirill Miazine.
1786 TF/01 Do not log submission mode rewrites if they do not change the address.
1788 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1790 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1791 log files in place. Contributed by Roberto Lima.
1793 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1795 TF/03 Bugzilla 615: When checking the local_parts router precondition
1796 after a local_part_suffix or local_part_prefix option, Exim now
1797 does not use the address's named list lookup cache, since this
1798 contains cached lookups for the whole local part.
1800 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1801 Robert Millan. Documentation is in experimental-spec.txt.
1803 TF/04 Bugzilla 668: Fix parallel build (make -j).
1805 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1807 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1808 Patch provided by Jan Srzednicki.
1810 TF/05 Leading white space used to be stripped from $spam_report which
1811 wrecked the formatting. Now it is preserved.
1813 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1814 that they are available at delivery time.
1816 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1818 TF/08 TLS error reporting now respects the incoming_interface and
1819 incoming_port log selectors.
1821 TF/09 Produce a more useful error message if an SMTP transport's hosts
1822 setting expands to an empty string.
1824 NM/06 Bugzilla 744: EXPN did not work under TLS.
1825 Patch provided by Phil Pennock.
1827 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1828 Patch provided by Richard Godbee.
1830 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1831 acl_smtp_notquit, added index entry.
1833 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1834 Patch provided by Eugene Bujak.
1836 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1837 accept(). Patch provided by Maxim Dounin.
1839 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1840 Patch provided by Phil Pennock.
1842 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1844 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1845 Patch provided by Brad "anomie" Jorsch.
1847 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1848 Patch provided by Dean Brooks.
1850 TK/05 Add native DKIM support (does not depend on external libraries).
1852 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1853 Patch provided by Graeme Fowler.
1855 NM/16 Bugzilla 851: Documentation example syntax fix.
1857 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1859 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1862 NM/19 Bugzilla 745: TLS version reporting.
1863 Patch provided by Phil Pennock.
1865 NM/20 Bugzilla 167: bool: condition support.
1866 Patch provided by Phil Pennock.
1868 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1869 clients. Patch provided by Phil Pennock.
1871 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1872 Patch provided by Brad "anomie" Jorsch.
1874 NM/23 Bugzilla 687: Fix misparses in eximstats.
1875 Patch provided by Heiko Schlittermann.
1877 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1878 Patch provided by Heiko Schlittermann.
1880 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1881 plus update to original patch.
1883 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1885 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1886 Patch provided by David Brownlee.
1888 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1890 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1892 NM/30 Bugzilla 888: TLS documentation bugfixes.
1894 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1896 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1897 Unlike the original bugzilla I have changed all shell scripts in src tree.
1899 NM/33 Bugzilla 898: Transport filter timeout fix.
1900 Patch by Todd Rinaldo.
1902 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1903 Patch by Serge Demonchaux.
1905 NM/35 Bugzilla 39: Base64 decode bug fixes.
1906 Patch by Jakob Hirsch.
1908 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1910 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1912 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1914 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1920 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1921 ALT-N's libdkim that I have put here:
1922 http://duncanthrax.net/exim-experimental/
1924 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1925 were called 'true' and it seems that C99 defines that as a reserved
1926 keyword to be used with 'bool' variable types. That means you could
1927 not include C99-style headers which use bools without triggering
1928 build errors in sieve.c.
1930 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1931 as mailq or other aliases. Changed the --help handling significantly
1932 to do whats expected. exim_usage() emits usage/help information.
1934 SC/01 Added the -bylocaldomain option to eximstats.
1936 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1938 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1940 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1946 PH/01 Another patch from the Sieve maintainer.
1948 PH/02 When an IPv6 address is converted to a string for single-key lookup
1949 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1950 dots are used instead of colons so that keys in lsearch files need not
1951 contain colons. This was done some time before quoting was made available
1952 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1953 (notated using the quote facility) so as to distinguish them from IPv4
1954 keys. This meant that lookups for IP addresses in host lists did not work
1955 for iplsearch lookups.
1957 This has been fixed by arranging for IPv6 addresses to be expressed with
1958 colons if the lookup type is iplsearch. This is not incompatible, because
1959 previously such lookups could never work.
1961 The situation is now rather anomalous, since one *can* have colons in
1962 ordinary lsearch keys. However, making the change in all cases is
1963 incompatible and would probably break a number of configurations.
1965 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1968 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1969 conversion specification without a maximum field width, thereby enabling
1970 a rogue spamd server to cause a buffer overflow. While nobody in their
1971 right mind would setup Exim to query an untrusted spamd server, an
1972 attacker that gains access to a server running spamd could potentially
1973 exploit this vulnerability to run arbitrary code as the Exim user.
1975 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1976 $primary_hostname instead of what libspf2 thinks the hosts name is.
1978 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1979 a directory entry by the name of the lookup key. Previously, if a
1980 symlink pointed to a non-existing file or a file in a directory that
1981 Exim lacked permissions to read, a lookup for a key matching that
1982 symlink would fail. Now it is enough that a matching directory entry
1983 exists, symlink or not. (Bugzilla 503.)
1985 PH/03 The body_linecount and body_zerocount variables are now exported in the
1988 PH/04 Added the $dnslist_matched variable.
1990 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1991 This means they are set thereafter only if the connection becomes
1994 PH/06 Added the client_condition to authenticators so that some can be skipped
1995 by clients under certain conditions.
1997 PH/07 The error message for a badly-placed control=no_multiline_responses left
1998 "_responses" off the end of the name.
2000 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2002 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2003 (without spaces) instead of just copying the configuration text.
2005 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2007 PH/11 Added $max_received_linelength.
2009 PH/12 Added +ignore_defer and +include_defer to host lists.
2011 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2012 way in which PCRE > 7.0 is built.
2014 PH/14 Implemented queue_only_load_latch.
2016 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2017 MAIL command. The effect was to mangle the value on 64-bit systems.
2019 PH/16 Another patch from the Sieve maintainer.
2021 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2023 PH/18 If a system quota error occurred while trying to create the file for
2024 a maildir delivery, the message "Mailbox is full" was not appended to the
2025 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2026 only to a quota excession during the actual writing of the file.
2028 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2029 characters?) which causes problems in log lines. The DN values are now
2030 passed through string_printing() before being added to log lines.
2032 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2033 and InterBase are left for another time.)
2035 PH/21 Added message_body_newlines option.
2037 PH/22 Guard against possible overflow in moan_check_errorcopy().
2039 PH/23 POSIX allows open() to be a macro; guard against that.
2041 PH/24 If the recipient of an error message contained an @ in the local part
2042 (suitably quoted, of course), incorrect values were put in $domain and
2043 $local_part during the evaluation of errors_copy.
2049 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2050 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2053 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2054 issue a MAIL command.
2056 PH/02 In an ACL statement such as
2058 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2060 if a client was not listed at all, or was listed with a value other than
2061 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2062 the condition was not true (as it should be), so access was not denied.
2063 The bug was that the ! inversion was incorrectly passed on to the second
2064 item. This has been fixed.
2066 PH/03 Added additional dnslists conditions == and =& which are different from
2067 = and & when the dns lookup returns more than one IP address.
2069 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2070 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2072 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2073 FSYNC, which compiles an option called disable_fsync that allows for
2074 bypassing fsync(). The documentation is heavily laced with warnings.
2076 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2078 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2079 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2080 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2081 including adding "make clean"; (3) Added -fPIC when compiling the test
2082 dynamically loaded module, to get rid of a warning.
2084 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2085 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2086 = 0s. The bug is otherwise harmless.
2088 PH/07 There was a bug in the dovecot authenticator such that the value of
2089 $auth1 could be overwritten, and so not correctly preserved, after a
2090 successful authentication. This usually meant that the value preserved by
2091 the server_setid option was incorrect.
2093 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2095 PH/09 Installed PCRE release 7.0.
2097 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2098 run for batched SMTP input. It is now run at the start of every message
2099 in the batch. While fixing this I discovered that the process information
2100 (output by running exiwhat) was not always getting set for -bs and -bS
2101 input. This is fixed, and it now also says "batched" for BSMTP.
2103 PH/11 Added control=no_pipelining.
2105 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2106 patch, slightly modified), and move the expansion of helo_data till after
2107 the connection is made in the smtp transport (so it can use these
2110 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2112 PH/14 Added log_selector = +pid.
2114 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2116 PH/16 Add ${if forany and ${if forall.
2118 PH/17 Added dsn_from option to vary the From: line in DSNs.
2120 PH/18 Flush SMTP output before performing a callout, unless control =
2121 no_callout_flush is set.
2123 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2124 was true (the default) a successful delivery failed to delete the retry
2125 item, thus causing premature timeout of the address. The bug is now
2128 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2130 PH/21 Long custom messages for fakedefer and fakereject are now split up
2131 into multiline responses in the same way that messages for "deny" and
2132 other ACL rejections are.
2134 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2135 with slight modification.
2137 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2138 draft, changing the syntax and factoring some duplicate code".
2140 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2141 for deliveries of the second and subsequent messages over the same SMTP
2144 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2145 ${reduce, with only minor "tidies".
2147 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2149 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2150 expansion side effects.
2152 PH/27 When a message times out after an over-quota error from an Exim-imposed
2153 quota, the bounce message says "mailbox is full". This message was not
2154 being given when it was a system quota that was exceeded. It now should
2157 MH/03 Made $recipients available in local_scan(). local_scan() already has
2158 better access to the recipient list through recipients_list[], but
2159 $recipients can be useful in postmaster-provided expansion strings.
2161 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2162 in the case of a MAIL command with additional options following the
2163 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2164 were accidentally chopped off.
2166 PH/29 SMTP synchronization checks are implemented when a command is read -
2167 there is a check that no more input is waiting when there shouldn't be
2168 any. However, for some commands, a delay in an ACL can mean that it is
2169 some time before the response is written. In this time, more input might
2170 arrive, invalidly. So now there are extra checks after an ACL has run for
2171 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2172 pipelining has not been advertised.
2174 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2176 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2177 returned in the SMTP response when smtp_return_error_details was false.
2178 This has been fixed.
2180 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2181 instead of the C I/O that was originally supplied, because problems were
2182 reported on Solaris.
2184 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2185 Exim which did not show up earlier: it was assuming that a call to
2186 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2187 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2188 was a macro that became an assignment, so it seemed to work. This has
2189 changed to a proper function call with a void return, hence the compile
2190 error. Exim's code has been fixed.
2192 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2195 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2197 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2199 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2200 because it turns out to be too big or there is a timeout) there is no
2201 "Completed" line in the log. When some messages of this type were
2202 selected by exigrep, they were listed as "not completed". Others were
2203 picked up by some special patterns. I have improved the selection
2204 criteria to be more general.
2206 PH/38 The host_find_failed option in the manualroute router can now be set
2207 to "ignore", to completely ignore a host whose IP address cannot be
2208 found. If all hosts are ignored, the behaviour is controlled by the new
2209 host_all_ignored option.
2211 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2212 homing or because of multiple MX records with /mx) generated more than
2213 one IP address, and the following item turned out to be the local host,
2214 all the secondary addresses of the first item were incorrectly removed
2215 from the list, along with the local host and any following hosts (which
2216 is what is supposed to happen).
2218 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2219 whoever called Exim into the -H file. In the case of the daemon it was
2220 behaving confusingly. When first started, it used values for whoever
2221 started the daemon, but after a SIGHUP it used the Exim user (because it
2222 calls itself on a restart). I have changed the code so that it now always
2225 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2226 message are rejected with the same error (e.g. no authentication or bad
2227 sender address), and a DATA command is nevertheless sent (as can happen
2228 with PIPELINING or a stupid MUA), the error message that was given to the
2229 RCPT commands is included in the rejection of the DATA command. This is
2230 intended to be helpful for MUAs that show only the final error to their
2233 PH/42 Another patch from the Sieve maintainer.
2235 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2236 Eximstats - Fixed some broken HTML links and added missing column headers
2238 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2239 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2241 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2243 PH/43 Yet another patch from the Sieve maintainer.
2245 PH/44 I found a way to check for a TCP/IP connection going away before sending
2246 the response to the final '.' that terminates a message, but only in the
2247 case where the client has not sent further data following the '.'
2248 (unfortunately, this is allowed). However, in many cases there won't be
2249 any further data because there won't be any more messages to send. A call
2250 to select() can be used: if it shows that the input is "ready", there is
2251 either input waiting, or the socket has been closed. An attempt to read
2252 the next input character can distinguish the two cases. Previously, Exim
2253 would have sent an OK response which the client would never have see.
2254 This could lead to message repetition. This fix should cure that, at
2255 least in a lot of common cases.
2257 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2258 advertised in response to EHLO.
2264 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2265 fixed by 4.65/MH/01 (is this a record?) are fixed:
2267 (i) An empty string was always treated as zero by the numeric comparison
2268 operators. This behaviour has been restored.
2270 (ii) It is documented that the numeric comparison operators always treat
2271 their arguments as decimal numbers. This was broken in that numbers
2272 starting with 0 were being interpreted as octal.
2274 While fixing these problems I realized that there was another issue that
2275 hadn't been noticed. Values of message_size_limit (both the global option
2276 and the transport option) were treated as octal if they started with 0.
2277 The documentation was vague. These values are now always treated as
2278 decimal, and I will make that clear in the documentation.
2284 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2285 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2288 MH/01 Don't check that the operands of numeric comparison operators are
2289 integers when their expansion is in "skipping" mode (fixes bug
2290 introduced by 4.64-PH/07).
2292 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2293 child addresses, Exim now panics and dies. Previously, because the count
2294 is held in a short int, deliveries were likely to be lost. As such a
2295 large number of recipients for a single message is ridiculous
2296 (performance will be very, very poor), I have chosen to impose a limit
2297 rather than extend the field.
2303 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2304 leftover -K file (the existence of which was triggered by #402).
2305 While we were at it, introduced process PID as part of the -K
2306 filename. This should rule out race conditions when creating
2309 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2310 processing considerably. Previous code took too long for large mails,
2311 triggering a timeout which in turn triggers #401.
2313 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2314 in the DK code in transports.c. sendfile() is not really portable,
2315 hence the _LINUX specificness.
2317 TF/01 In the add_headers option to the mail command in an Exim filter,
2318 there was a bug that Exim would claim a syntax error in any
2319 header after the first one which had an odd number of characters
2322 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2323 callout verification, Exim cached a "reject" for the entire domain. This
2324 is correct for most verifications, but it is not correct for a recipient
2325 verification with use_sender or use_postmaster set, because in that case
2326 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2327 case of MAIL FROM:<> rejection from other early rejections (e.g.
2328 rejection of HELO). When verifying a recipient using a non-null MAIL
2329 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2330 Whatever the result of the callout, the value of the domain cache is
2331 left unchanged (for any other kind of callout, getting as far as trying
2332 RCPT means that the domain itself is ok).
2334 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2337 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2338 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2339 session, the daemon ignores ECONNECT errors and logs others; it now
2340 ignores EPIPE as well.
2342 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2343 (quoted-printable decoding).
2345 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2346 later the small subsequent patch to fix an introduced bug.
2348 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2350 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2352 PH/08 An error is now given if message_size_limit is specified negative.
2354 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2355 to be given (somewhat) arbitrary names.
2357 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2360 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2361 miscellaneous code fixes
2363 PH/10 Added the log_reject_target ACL modifier to specify where to log
2366 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2367 hostname. This is wrong, because it relates to the incoming message (and
2368 probably the interface on which it is arriving) and not to the outgoing
2369 callout (which could be using a different interface). This has been
2370 changed to use the value of the helo_data option from the smtp transport
2371 instead - this is what is used when a message is actually being sent. If
2372 there is no remote transport (possible with a router that sets up host
2373 addresses), $smtp_active_hostname is used.
2375 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2376 tweaks were necessary in order to get it to work (see also 21 below):
2377 (a) The code assumed that strncpy() returns a negative number on buffer
2378 overflow, which isn't the case. Replaced with Exim's string_format()
2380 (b) There were several signed/unsigned issues. I just did the minimum
2381 hacking in of casts. There is scope for a larger refactoring.
2382 (c) The code used strcasecmp() which is not a standard C function.
2383 Replaced with Exim's strcmpic() function.
2384 (d) The code set only $1; it now sets $auth1 as well.
2385 (e) A simple test gave the error "authentication client didn't specify
2386 service in request". It would seem that Dovecot has changed its
2387 interface. Fortunately there's a specification; I followed it and
2388 changed what the client sends and it appears to be working now.
2390 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2393 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2394 address A is aliased to B and C, where B exists and C does not. Without
2395 -v the output is "A verified" because verification stops after a
2396 successful redirection if more than one address is generated. However,
2397 with -v the child addresses are also verified. Exim was outputting "A
2398 failed to verify" and then showing the successful verification for C,
2399 with its parentage. It now outputs "B failed to verify", showing B's
2400 parentage before showing the successful verification of C.
2402 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2403 look up a TXT record in a specific list after matching in a combined
2406 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2407 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2408 they consult the DNS. I had assumed they would set it the way they
2409 wanted; and indeed my experiments on Linux seem to show that in some
2410 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2411 To be on the safe side, however, I have now made the interface to
2412 host_find_byname() similar to host_find_bydns(), with an argument
2413 containing the DNS resolver options. The host_find_byname() function now
2414 sets these options at its start, just as host_find_bydns() does. The smtp
2415 transport options dns_qualify_single and dns_search_parents are passed to
2416 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2417 of host_find_byname() use the default settings of RES_DEFNAMES
2418 (qualify_single) but not RES_DNSRCH (search_parents).
2420 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2421 spool_read_header() do less string testing, by means of a preliminary
2422 switch on the second character of optional "-foo" lines. (This is
2423 overdue, caused by the large number of possibilities that now exist.
2424 Originally there were few.) While I was there, I also converted the
2425 str(n)cmp tests so they don't re-test the leading "-" and the first
2426 character, in the hope this might squeeze out yet more improvement.
2428 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2429 flag allowing group syntax was set by the header_syntax check but not
2430 turned off, possible causing trouble later; (2) The flag was not being
2431 set at all for the header_verify test, causing "group"-style headers to
2432 be rejected. I have now set it in this case, and also caused header_
2433 verify to ignore an empty address taken from a group. While doing this, I
2434 came across some other cases where the code for allowing group syntax
2435 while scanning a header line wasn't quite right (mostly, not resetting
2436 the flag correctly in the right place). These bugs could have caused
2437 trouble for malformed header lines. I hope it is now all correct.
2439 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2440 with the "reply" argument non-NULL. The code, however (which originally
2441 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2442 but it didn't always do it. This confused somebody who was copying the
2443 code for some other use. I have removed all the tests.
2445 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2446 feature that was used to support insecure browsers during the U.S. crypto
2447 embargo. It requires special client support, and Exim is probably the
2448 only MTA that supported it -- and would never use it because real RSA is
2449 always available. This code has been removed, because it had the bad
2450 effect of slowing Exim down by computing (never used) parameters for the
2451 RSA_EXPORT functionality.
2453 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2454 authenticator to fail if there's a tab character in the incoming data
2455 (there should never be unless someone is messing about, as it's supposed
2456 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2457 if the connection is using TLS or if the remote IP is the same as the
2458 local IP, and the "valid-client-cert option" if a client certificate has
2461 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2462 authenticators. This can be used for authorization after authentication
2463 succeeds. (In the case of plaintext, it servers for both authentication
2466 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2467 if any retry times were supplied.
2469 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2470 connection, where there is no client IP address to check. In this
2471 situation, the verify now always succeeds.
2473 PH/25 Applied John Jetmore's -Mset patch.
2475 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2477 PH/27 In a string expansion for a processed (not raw) header when multiple
2478 headers of the same name were present, leading whitespace was being
2479 removed from all of them, but trailing whitespace was being removed only
2480 from the last one. Now trailing whitespace is removed from each header
2481 before concatenation. Completely empty headers in a concatenation (as
2482 before) are ignored.
2484 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2485 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2487 PH/29 [Removed. This was a change that I later backed out, and forgot to
2488 correct the ChangeLog entry (that I had efficiently created) before
2489 committing the later change.]
2491 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2492 address errors (4xx response to RCPT) over the same connection as other
2493 messages routed to the same hosts. Such deliveries are always "forced",
2494 so retry times are not inspected. This resulted in far too many retries
2495 for the affected addresses. The effect occurred only when there were more
2496 hosts than the hosts_max_try setting in the smtp transport when it had
2497 the 4xx errors. Those hosts that it had tried were not added to the list
2498 of hosts for which the message was waiting, so if all were tried, there
2499 was no problem. Two fixes have been applied:
2501 (i) If there are any address or message errors in an SMTP delivery, none
2502 of the hosts (tried or untried) are now added to the list of hosts
2503 for which the message is waiting, so the message should not be a
2504 candidate for sending over the same connection that was used for a
2505 successful delivery of some other message. This seems entirely
2506 reasonable: after all the message is NOT "waiting for some host".
2507 This is so "obvious" that I'm not sure why it wasn't done
2508 previously. Hope I haven't missed anything, but it can't do any
2509 harm, as the worst effect is to miss an optimization.
2511 (ii) If, despite (i), such a delivery is accidentally attempted, the
2512 routing retry time is respected, so at least it doesn't keep
2513 hammering the server.
2515 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2516 in ${readsocket because some servers need this prod.
2518 PH/32 Added some extra debug output when updating a wait-xxx database.
2520 PH/33 The hint "could be header name not terminated by colon", which has been
2521 given for certain expansion errors for a long time, was not being given
2522 for the ${if def:h_colon_omitted{... case.
2524 PH/34 The spec says: "With one important exception, whenever a domain list is
2525 being scanned, $domain contains the subject domain." There was at least
2526 one case where this was not true.
2528 PH/35 The error "getsockname() failed: connection reset by peer" was being
2529 written to the panic log as well as the main log, but it isn't really
2530 panic-worthy as it just means the connection died rather early on. I have
2531 removed the panic log writing for the ECONNRESET error when getsockname()
2534 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2535 runs only) independently of the message's sender address. This meant
2536 that, if the 4xx error was in fact related to the sender, a different
2537 message to the same recipient with a different sender could confuse
2538 things. In particular, this can happen when sending to a greylisting
2539 server, but other circumstances could also provoke similar problems.
2540 I have changed the default so that the retry time for these errors is now
2541 based a combination of the sender and recipient addresses. This change
2542 can be overridden by setting address_retry_include_sender=false in the
2545 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2546 remote server are returned as part of bounce messages. This was not
2547 happening for LMTP over a pipe (the lmtp transport), but now it is the
2548 same for both kinds of LMTP.
2550 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2551 in header lines that were in fact CNAMEs. This is no longer the case.
2553 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2554 and queue runs started by the daemon processed all messages. This has
2555 been fixed so that -R and -S can now usefully be given with -q<time>.
2557 PH/40 Import PCRE release 6.7 (fixes some bugs).
2559 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2561 PH/42 Give an error if -q is specified more than once.
2563 PH/43 Renamed the variables $interface_address and $interface_port as
2564 $received_ip_address and $received_port, to make it clear that these
2565 values apply to message reception, and not to the outgoing interface when
2566 a message is delivered. (The old names remain recognized, of course.)
2568 PH/44 There was no timeout on the connect() call when using a Unix domain
2569 socket in the ${readsocket expansion. There now is.
2571 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2572 be meaningful with "accept".
2574 SC/01 Eximstats V1.43
2575 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2577 SC/02 Eximstats V1.44
2578 Use a glob alias rather than an array ref in the generated
2579 parser. This improves both readability and performance.
2581 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2582 Collect SpamAssassin and rejection statistics.
2583 Don't display local sender or destination tables unless
2584 there is data to show.
2585 Added average volumes into the top table text output.
2587 SC/04 Eximstats V1.46
2588 Collect data on the number of addresses (recipients)
2589 as well as the number of messages.
2591 SC/05 Eximstats V1.47
2592 Added 'Message too big' to the list of mail rejection
2593 reasons (thanks to Marco Gaiarin).
2595 SC/06 Eximstats V1.48
2596 Mainlog lines which have GMT offsets and are too short to
2597 have a flag are now skipped.
2599 SC/07 Eximstats V1.49 (Alain Williams)
2600 Added the -emptyok flag.
2602 SC/08 Eximstats V1.50
2603 Fixes for obtaining the IP address from reject messages.
2605 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2606 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2607 whitespace changes from 4.64-PH/27
2609 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2612 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2613 are found, allow negative numbers in numeric criteria)
2615 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2617 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2618 to match changes made in 4.64-PH/43
2620 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2622 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2623 "domain" is actually an IP address, and force a failure. This locks out
2624 those revolvers/nameservers that support "A-for-A" lookups, in
2625 contravention of the specifications.
2627 PH/48 When a host name was looked up from an IP address, and the subsequent
2628 forward lookup of the name timed out, the host name was left in
2629 $sender_host_name, contrary to the specification.
2631 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2632 restricted to single-key lookups, Exim was not diagnosing an error if
2633 * or *@ was used with a query-style lookup.
2635 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2637 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2638 long ago, but noone interested enough thought of it. Let's just say that
2639 the "1.1" means that there are some new functions that weren't there at
2640 some point in the past.
2642 PH/51 Error processing for expansion failure of helo_data from an smtp
2643 transport during callout processing was broken.
2645 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2646 tested/used via the -bh/-bhc/-bs options.
2648 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2649 bug, fixed in subsequent PCRE releases).
2651 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2652 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2654 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2660 SC/01 Use a glob alias rather than an array ref in eximstats generated
2661 parser. This improves both readability and performance.
2663 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2664 Don't display local sender or destination tables in eximstats unless
2665 there is data to show.
2666 Added average volumes into the eximstats top table text output.
2668 SC/03 Collect data on the number of addresses (recipients) as well
2669 as the number of messages in eximstats.
2671 TF/01 Correct an error in the documentation for the redirect router. Exim
2672 does (usually) call initgroups() when daemonizing.
2674 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2675 with consistent privilege compared to when running as a daemon.
2677 TF/03 Note in the spec that $authenticated_id is not set for local
2678 submissions from trusted users.
2680 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2681 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2683 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2684 by adding some example configuration directives to the default
2685 configuration file. A little bit of work is required to uncomment the
2686 directives and define how usernames and passwords are checked, but
2687 there is now a framework to start from.
2689 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2690 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2691 without this. I don't know how relevant this is to other LDAP libraries.
2693 PH/02 Add the verb name to the "unknown ACL verb" error.
2695 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2697 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2699 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2700 directory not expanded when it should be if an expanded home directory
2701 was set for the address (which is overridden by the transport).
2703 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2706 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2707 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2708 because it is too late at that time, and has no effect.
2710 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2711 security issue with \' (bugzilla #107). I could not use the
2712 PQescapeStringConn() function, because it needs a PGconn value as one of
2715 PH/08 When testing addresses using -bt, indicate those final addresses that
2716 are duplicates that would not cause an additional delivery. At least one
2717 person was confused, thinking that -bt output corresponded to deliveries.
2718 (Suppressing duplicates isn't a good idea as you lose the information
2719 about possibly different redirections that led to the duplicates.)
2721 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2722 systems where poll() doesn't work, in particular OS X.
2724 PH/10 Added more information to debugging output for retry time not reached.
2726 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2727 operations in malware.c.
2729 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2732 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2733 syslog_duplication set false, log lines that would normally be written
2734 both the the main log and to the reject log were not written to syslog at
2737 PH/14 In the default configuration, change the use of "message" in ACL warn
2738 statements to "add_header".
2740 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2741 not followed by a command (e.g. "seen endif").
2743 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2744 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2747 PH/17 Added extra conditions to the default value of delay_warning_condition
2751 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2752 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2753 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2756 The Auto-Submitted: and various List- headers are standardised, whereas I
2757 don't think Precedence: ever was.
2759 PH/18 Refactored debugging code in route_finduser() to show more information,
2760 in particular, the error code if getpwnam() issues one.
2762 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2763 This is apparently needed in addition to the PH/07 change above to avoid
2764 any possible encoding problems.
2766 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2767 but not after initializing Perl.
2769 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2770 output them only if debugging. By default they are written stderr,
2771 apparently, which is not desirable.
2773 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2776 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2779 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2781 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2782 authenticated or an ident call has been made. Suppress the default
2783 values for $authenticated_id and $authenticated_sender (but permit -oMai
2784 and -oMas) when testing with -bh.
2786 PH/24 Re-jigged the order of the tests in the default configuration so that the
2787 tests for valid domains and recipients precede the DNS black list and CSA
2788 tests, on the grounds that those ones are more expensive.
2790 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2791 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2792 This bug exists in every version of Exim that I still have, right back to
2795 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2796 However, an attempt to turn on case-sensitivity in a regex key by
2797 including (?-i) didn't work because the subject string was already
2798 lowercased, and the effects were non-intuitive. It turns out that a
2799 one-line patch can be used to allow (?-i) to work as expected.
2805 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2806 other effects) broke the use of negated acl sub-conditions.
2808 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2811 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2812 "Deny" causes Exim to reject the incoming connection with a 554 error.
2813 Unfortunately, if there is a major crisis, such as a disk failure,
2814 tcp-wrappers gives "deny", whereas what one would like would be some
2815 kind of temporary error. A kludge has been added to help with this.
2816 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2817 554 error is used if errno is still zero or contains ENOENT (which occurs
2818 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2821 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2823 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2824 errors. Otherwise a message that provokes a temporary error (when other
2825 messages do not) can cause a whole host to time out.
2827 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2828 addresses were routed directly to files or pipes from a redirect router.
2829 File deliveries just didn't batch; pipe deliveries might have suffered
2832 PH/06 A failure to get a lock for a hints database would erroneously always say
2833 "Failed to get write lock", even when it was really a read lock.
2835 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2836 of 0600. This has been changed to use the value of the lockfile_mode
2837 option (which defaults to 0600).
2839 PH/08 Applied small patch from the Sieve maintainer.
2841 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2842 folder from quota calculations, a direct delivery into this folder messed
2843 up the contents of the maildirsize file. This was because the regex was
2844 used only to exclude .Trash (or whatever) when the size of the mailbox
2845 was calculated. There was no check that a delivery was happening into an
2846 excluded directory. This bug has been fixed by ignoring all quota
2847 processing for deliveries into excluded directories.
2849 PH/10 Added the maildirfolder_create_regex option to appendfile.
2855 PH/01 The code for finding all the local interface addresses on a FreeBSD
2856 system running IPv6 was broken. This may well have applied to all BSD
2857 systems, as well as to others that have similar system calls. The broken
2858 code found IPv4 interfaces correctly, but gave incorrect values for the
2859 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2860 that it would not match correctly against @[] and not recognize the IPv6
2863 PH/02 The ipliteral router was not recognizing addresses of the form user@
2864 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2866 PH/03 Added disable_ipv6.
2868 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2869 decoded form, because it is most often used to construct To: headers
2870 lines in autoreplies, and the decoded form may well be syntactically
2871 invalid. However, $reply_address has leading white space removed, and all
2872 newlines turned into spaces so that the autoreply transport does not
2875 PH/05 If group was specified without a user on a router, and no group or user
2876 was specified on a transport, the group from the router was ignored.
2878 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2879 for visible compile-time settings that can be used to change these
2880 numbers, for those that want even more. Backwards compatibility with old
2881 spool files has been maintained. However, going back to a previous Exim
2882 release will lost any variables that are in spool files.
2884 PH/07 Two small changes when running in the test harness: increase delay when
2885 passing a TCP/IP connection to a new process, in case the original
2886 process has to generate a bounce, and remove special handling of
2887 127.0.0.2 (sic), which is no longer necessary.
2889 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2890 be the same on different OS.
2892 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2895 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2896 whether --show-vars was specified or not
2898 JJ/02 exipick: Added support for new ACL variable spool format introduced
2901 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2902 syntactically invalid From: or Reply-to: line, and a filter used this to
2903 generate an autoreply, and therefore failed to obtain an address for the
2904 autoreply, Exim could try to deliver to a non-existent relative file
2905 name, causing unrelated and misleading errors. What now happens is that
2906 it logs this as a hard delivery error, but does not attempt to create a
2909 PH/11 The exinext utility has a -C option for testing purposes, but although
2910 the given file was scanned by exinext itself; it wasn't being passed on
2911 when Exim was called.
2913 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2914 an end-of-file indication when reading a command response.
2916 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2917 compiled. In many other places in Exim, IPv6 addresses are always
2918 recognized, so I have changed this. It also means that IPv4 domain
2919 literals of the form [IPV4:n.n.n.n] are now always recognized.
2921 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2922 used if the router is not running as root, for example, when verifying at
2923 ACL time, or when using -bh. The debugging output from this situation was
2924 non-existent - all you got was a failure to exec. I have made two
2927 (a) Failures to set uid/gid, the current directory, or a process leader
2928 in a subprocess such as that created by queryprogram now generate
2929 suitable debugging output when -d is set.
2931 (b) The queryprogram router detects when it is not running as root,
2932 outputs suitable debugging information if -d is set, and then runs
2933 the subprocess without attempting to change uid/gid.
2935 PH/15 Minor change to Makefile for building test_host (undocumented testing
2938 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2939 additional section of a DNS packet that returns MX or SRV records.
2940 Instead, it always explicitly searches for A/AAAA records. This avoids
2941 major problems that occur when a DNS server includes only records of one
2942 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2943 fixed another bug: if SRV records were looked up and the corresponding
2944 address records were *not* found in the additional section, the port
2945 values from the SRV records were lost.
2947 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2948 using the correct key (the original address) when searching the retry
2949 rules in order to find which one to use for generating the retry hint.
2951 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2952 adding the default one. Similarly, if it contains a Reply-To: header, the
2953 errors_reply_to option, if set, is not used.
2955 PH/19 When calculating a retry time, Exim used to measure the "time since
2956 failure" by looking at the "first failed" field in the retry record. Now
2957 it does not use this if it is later than than the arrival time of the
2958 message. Instead it uses the arrival time. This makes for better
2959 behaviour in cases where some deliveries succeed, thus re-setting the
2960 "first failed" field. An example is a quota failure for a huge message
2961 when small messages continue to be delivered. Without this change, the
2962 "time since failure" will always be short, possible causing more frequent
2963 delivery attempts for the huge message than are intended.
2964 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2966 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2967 $1, $2, $3) because the numerical variables can be reset during some
2968 expansion items (e.g. "match"), thereby losing the authentication data.
2970 PH/21 Make -bV show the size of off_t variables so that the test suite can
2971 decide whether to run tests for quotas > 2G.
2973 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2974 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2975 filecount value is greater than 2G or if a quota value is greater than 2G
2976 on a system where the size of off_t is not greater than 4, a panic error
2979 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2980 never match. The debug and -bh output now contains an explicit error
2981 message indicating a malformed IPv4 address or mask.
2983 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2984 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2985 PH/23 above applies.
2987 PH/25 Do not write to syslog when running in the test harness. The only
2988 occasion when this arises is a failure to open the main or panic logs
2989 (for which there is an explicit test).
2991 PH/26 Added the /no_tell option to "control=freeze".
2993 PH/27 If a host name lookup failed very early in a connection, for example, if
2994 the IP address matched host_lookup and the reverse lookup yielded a name
2995 that did not have a forward lookup, an error message of the form "no IP
2996 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2997 could be logged. Now it outputs the IP address instead of "NULL".
2999 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3000 allows the sender and the authenticated sender to be set when
3001 submitting a message from within Exim. Since child_open_exim() is
3002 documented for local_scan(), the new function should be too.
3004 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3005 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3006 results in an empty string is now treated as unset.
3008 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3010 PH/31 Added sender_verify_fail logging option.
3012 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3013 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3014 tidied the source and removed it altogether.
3016 PH/33 When a queue run was abandoned because the load average was too high, a
3017 log line was always written; now it is written only if the queue_run log
3018 selector is set. In addition, the log line for abandonment now contains
3019 information about the queue run such as the pid. This is always present
3020 in "start" and "stop" lines but was omitted from the "abandon" line.
3022 PH/34 Omit spaces between a header name and the colon in the error message that
3023 is given when verify = headers_syntax fails (if there are lots of them,
3024 the message gets confusing).
3026 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3027 names, as there are now some PTR records that contain slashes. This check
3028 is only to protect against broken name servers that fall over on strange
3029 characters, so the fact that it applies to all lookups doesn't matter.
3031 PH/36 Now that the new test suite is complete, we can remove some of the
3032 special code in Exim that was needed for the old test suite. For example,
3033 sorting DNS records because real resolvers return them in an arbitrary
3034 order. The new test suite's fake resolver always returns records in the
3037 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3038 bounces) except when queue_only is set, to avoid logging races between
3039 the different processes.
3041 PH/38 Panic-die if .include specifies a non-absolute path.
3043 PH/39 A tweak to the "H" retry rule from its user.
3045 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3046 a label. They prevented compilation on older perls.
3048 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3049 a warning to be raised on newish perls.
3051 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3052 on queue. Changes to match documented behaviour of showing count of
3053 messages matching specified criteria.
3055 PH/40 Changed the default ident timeout from 30s to 5s.
3057 PH/41 Added support for the use of login_cap features, on those BSD systems
3058 that have them, for controlling the resources used by pipe deliveries.
3060 PH/42 The content-scanning code uses fopen() to create files in which to put
3061 message data. Previously it was not paying any attention to the mode of
3062 the files. Exim runs with umask(0) because the rest of the code creates
3063 files with open(), and sets the required mode explicitly. Thus, these
3064 files were ending up world-writeable. This was not a big issue, because,
3065 being within the spool directory, they were not world-accessible. I have
3066 created a function called modefopen, which takes an additional mode
3067 argument. It sets umask(777), creates the file, chmods it to the required
3068 mode, then resets the umask. All the relevant calls to fopen() in the
3069 content scanning code have been changed to use this function.
3071 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3072 to 24 hours. This avoids potential overflow problems when processing G
3073 and H retry rules. I suspect nobody ever tinkers with this value.
3075 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3077 PH/45 When the plaintext authenticator is running as a client, the server's
3078 challenges are checked to ensure they are valid base64 strings. By
3079 default, the authentication attempt is cancelled if an invalid string is
3080 received. Setting client_ignore_invalid_base64 true ignores these errors.
3081 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3082 they are received. Thus, the responses can be made to depend on the
3083 challenges. If an invalid string is ignored, an empty string is placed in
3086 PH/46 Messages that are created by the autoreply transport now contains a
3087 References: header, in accordance with RFCs 2822 and 3834.
3089 PH/47 Added authenticated_sender_force to the smtp transport.
3091 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3093 PH/49 Installed latest patch from the Sieve maintainer.
3095 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3096 was also set, the appendfile transport was unnecessarily scanning a
3097 directory of message files (e.g. for maildir delivery) to find the count
3098 of files (along with the size), even though it did not need this
3099 information. It now does the scan only if it needs to find either the
3100 size of the count of files.
3102 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3104 PH/52 Two bugs concerned with error handling when the smtp transport is
3107 (i) Exim was not creating retry information for temporary errors given
3108 for individual recipients after the DATA command when the smtp transport
3109 was used in LMTP mode. This meant that they could be retried too
3110 frequently, and not timed out correctly.
3112 (ii) Exim was setting the flag that allows error details to be returned
3113 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3114 recipients that were returned after the DATA command.
3116 PH/53 This is related to PH/52, but is more general: for any failing address,
3117 when detailed error information was permitted to be returned to the
3118 sender, but the error was temporary, then after the final timeout, only
3119 "retry timeout exceeded" was returned. Now it returns the full error as
3120 well as "retry timeout exceeded".
3122 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3123 do this, and (what is worse) MTAs that accept it.
3125 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3126 will now be deprecated.
3128 PH/56 New os.c-cygwin from the Cygwin maintainer.
3130 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3131 formats (previously only available in exim formats via -bpr, -bpru,
3132 and -bpra. Now also available in native and exiqgrep formats)
3134 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3135 with very large, slow to parse queues
3137 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3139 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3141 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3142 responses to authentication challenges, though it was showing the
3143 challenges; (ii) I've removed the CR characters from the debug output for
3146 PH/58 Allow for the insertion of a newline as well as a space when a string
3147 is turned into more than one encoded-word during RFC 2047 encoding. The
3148 Sieve code now uses this.
3150 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3151 data_4xx, lost_connection, tls_required.
3153 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3154 message was being sent as an SMTP response.
3156 PH/61 Add -l and -k options to exicyclog.
3158 PH/62 When verifying, if an address was redirected to one new address, so that
3159 verification continued, and the new address failed or deferred after
3160 having set something in $address_data, the value of $address_data was not
3161 passed back to the ACL. This was different to the case when no
3162 redirection occurred. The value is now passed back in both cases.
3164 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3165 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3166 use login_cap.h, so on its own it isn't the distinguishing feature. The
3167 new name refers directly to the setclassresources() function.
3169 PH/65 Added configuration files for NetBSD3.
3171 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3173 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3174 is preferred over IPv4.
3176 PH/68 The bounce_return_message and bounce_return_body options were not being
3177 honoured for bounces generated during the reception of non-SMTP messages.
3178 In particular, this applied to messages rejected by the ACL. This bug has
3179 been fixed. However, if bounce_return_message is true and bounce_return_
3180 body is false, the headers that are returned for a non-SMTP message
3181 include only those that have been read before the error was detected.
3182 (In the case of an ACL rejection, they have all been read.)
3184 PH/69 The HTML version of the specification is now built in a directory called
3185 spec_html instead of spec.html, because the latter looks like a path with
3186 a MIME-type, and this confuses some software.
3188 PH/70 Catch two compiler warnings in sieve.c.
3190 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3191 function verify_get_ident() calls ip_connect() to connect a socket, but
3192 if the "connect()" function timed out, ip_connect() used to close the
3193 socket. However, verify_get_ident() also closes the socket later, and in
3194 between Exim writes to the log, which may get opened at this point. When
3195 the socket was closed in ip_connect(), the log could get the same file
3196 descriptor number as the socket. This naturally causes chaos. The fix is
3197 not to close the socket in ip_connect(); the socket should be closed by
3198 the function that creates it. There was only one place in the code where
3199 this was missing, in the iplookup router, which I don't think anybody now
3200 uses, but I've fixed it anyway.
3202 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3203 well as to direct DNS lookups. Otherwise the handling of names in host
3204 lists is inconsistent and therefore confusing.
3210 PH/01 Two changes to the default runtime configuration:
3212 (1) Move the checks for relay_from_hosts and authenticated clients from
3213 after to before the (commented out) DNS black list checks.
3215 (2) Add control=submission to the relay_from_hosts and authenticated
3216 clients checks, on the grounds that messages accepted by these
3217 statements are most likely to be submissions.
3219 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3221 (1) Generate an error if the third argument for the ${prvs expansion is
3224 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3227 (3) Reset the variables that are obtained from the first argument of
3228 ${prvscheck and used in the second argument before leaving the code,
3229 because their memory is reclaimed, so using them afterwards may do
3232 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3233 one (it's much easier than Tom thought :-).
3235 (5) Because of (4), we can now allow for the use of $prvscheck_result
3236 inside the third argument.
3238 PH/03 For some reason, the default setting of PATH when running a command from
3239 a pipe transport was just "/usr/bin". I have changed it to
3242 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3243 anything to be listed in the output from -bV.
3245 PH/05 When a filter generated an autoreply, the entire To: header line was
3246 quoted in the delivery log line, like this:
3248 => >A.N.Other <ano@some.domain> <original@ddress> ...
3250 This has been changed so that it extracts the operative address. There
3251 may be more than one such address. If so, they are comma-separated, like
3254 => >ano@some.domain,ona@other.domain <original@ddress> ...
3256 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3257 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3258 not being looked up in the rDNS to get a host name, Exim was showing the
3259 IP address twice in Received: lines, even though the IP addresses were
3260 identical. For example:
3262 Received: from [1.2.3.4] (helo=[1.2.3.4])
3264 However, if the real host name was known, it was omitting the HELO data
3265 if it matched the actual IP address. This has been tidied up so that it
3266 doesn't show the same IP address twice.
3268 PH/07 When both +timestamp and +memory debugging was on, the value given by
3269 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3270 called by the debug printing, thereby overwriting the timestamp buffer.
3271 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3273 PH/08 When the original message was included in an autoreply transport, it
3274 always said "this is a copy of the message, including all the headers",
3275 even if body_only or headers_only was set. It now gives an appropriate
3278 PH/09 Applied a patch from the Sieve maintainer which:
3280 o fixes some comments
3281 o adds the (disabled) notify extension core
3282 o adds some debug output for the result of if/elsif tests
3283 o points to the current vacation draft in the documentation
3284 and documents the missing references header update
3288 o fixes a bug in processing the envelope test (when testing
3289 multiple envelope elements, the last element determined the
3292 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3293 Electronic Mail") by including:
3295 Auto-submitted: auto-generated
3297 in the messages that it generates (bounce messages and others, such as
3298 warnings). In the case of bounce messages for non-SMTP messages, there was
3299 also a typo: it was using "Auto_submitted" (underscore instead of
3300 hyphen). Since every message generated by Exim is necessarily in response
3301 to another message, thes have all been changed to:
3303 Auto-Submitted: auto-replied
3305 in accordance with these statements in the RFC:
3307 The auto-replied keyword:
3309 - SHOULD be used on messages sent in direct response to another
3310 message by an automatic process,
3312 - MUST NOT be used on manually-generated messages,
3314 - MAY be used on Delivery Status Notifications (DSNs) and Message
3315 Disposition Notifications (MDNs),
3317 - MUST NOT be used on messages generated by automatic or periodic
3318 processes, except for messages which are automatic responses to
3321 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3322 to the default Received: header definition.
3324 PH/12 Added log selector acl_warn_skipped (default on).
3326 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3327 variables because (a) they are in the wrong storage pool and (b) even if
3328 they were copied, it wouldn't work properly because of the caching.
3330 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3331 checking when decoding. Apparently there are clients that generate
3332 overlong encoded strings. Why am I not surprised?
3334 PH/15 If the first argument of "${if match_address" was not empty, but did not
3335 contain an "@" character, Exim crashed. Now it writes a panic log message
3336 and treats the condition as false.
3338 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3340 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3341 extension "envelope-auth". The code is finished and in agreement with
3342 other implementations, but there is no documentation so far and in fact,
3343 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3344 not changing the active code.
3346 Print executed "if" and "elsif" statements when debugging is used. This
3347 helps a great deal to understand what a filter does.
3349 Document more things not specified clearly in RFC3028. I had all this
3350 sorted out, when out of a sudden new issues came to my mind. Oops."
3352 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3355 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3356 canonical form (as documented). However, after a host name lookup from
3357 the IP address, check_host() was doing a simple string comparison with
3358 addresses acquired from the DNS when checking that the found name did
3359 have the original IP as one of its addresses. Since any found IPv6
3360 addresses are likely to be in abbreviated form, the comparison could
3361 fail. Luckily, there already exists a function for doing the comparison
3362 by converting both addresses to binary, so now that is used instead of
3363 the text comparison.
3365 PH/20 There was another similar case to PH/19, when a complete host name was
3366 given in a host list; looking up its IP address could give an abbreviated
3367 form, whereas the current host's name might or might not be abbreviated.
3368 The same fix has been applied.
3374 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3375 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3378 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3379 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3381 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3383 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3384 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3385 Also, for "Auto-Submitted", treat anything other than "no" as
3386 non-personal, in accordance with RFC 3834. (Previously it treated
3387 anything starting "auto-" as non-personal.)
3389 TF/01 The control=submission/name=... option had a problem with syntax
3390 errors if the name included a slash character. The /name= option
3391 now slurps the rest of the string, so it can include any characters
3392 but it must come last in the list of options (after /sender_retain
3395 PH/05 Some modifications to the interface to the fake nameserver for the new
3403 TK/01 Added the "success_on_redirect" address verification option. See
3404 NewStuff for rationale and an example.
3406 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3408 PH/02 Patch to exigrep to allow it to work on syslog lines.
3410 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3411 fread() to skip over the body file's header line, because in Cygwin the
3412 header line is locked and is inaccessible.
3414 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3415 co-exist for some time) to make it clear that it is the Exim ID that is
3416 referenced, not the Message-ID: header line.
3418 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3419 string_format() function, because snprintf() does not exist on all
3422 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3423 ${stat: expansion item.
3425 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3426 protocol synchronization error", to keep the pedants happy.
3428 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3429 well as for IRIX systems, when gcc is being used. See the host.c source
3432 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3434 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3437 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3438 transport and to the smtp transport in LMTP mode.
3440 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3442 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3443 run a filter in a subprocess. This could lead to confusion in subsequent
3444 lookups in the parent process. There should also be a search_tidyup() at
3445 the end of the subprocess.
3447 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3448 only if the host matched helo_try_verify_hosts, which caused the
3449 verification to occur when the EHLO/HELO command was issued. The ACL just
3450 tested the remembered result. Now, if a previous verification attempt has
3451 not happened, "verify = helo" does it there and then.
3453 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3455 TK/03 Fix log output including CR from clamd.
3457 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3458 exist provoked a memory error which could cause a segfault.
3460 PH/15 Installed PCRE 6.2
3462 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3464 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3465 of the problem. Specifically, suggested +O2 rather than +O1 for the
3468 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3470 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3471 record, the port was not correctly logged when the outgoing_port log
3472 selector was set (it logged the transort's default port).
3474 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3475 fallback_hosts, and "hosts" in the smtp transport.
3477 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3478 host errors such as "Connection refused".
3480 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3481 authentication with radiusclient 0.4.9:
3483 - Error returned from rc_read_config was caught wrongly
3484 - Username/password not passed on to radius server due to wrong length.
3486 The presumption is that some radiusclient API changes for 4.51/PH/17
3487 were not taken care of correctly. The code is still untested by me (my
3488 Linux distribution still has 0.3.2 of radiusclient), but it was
3489 contributed by a Radius user.
3491 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3492 expanding the "port" option of the smtp transport.
3494 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3495 while reading a MIME header. Thanks to Tom Hughes for a patch.
3497 PH/24 Include config.h inside local_scan.h so that configuration settings are
3500 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3501 that in an ACL for RCPT (for example), you can examine exactly what was
3504 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3505 commands, but it was not correctly comparing the address with the actual
3506 client host address. Thus, it would show the EHLO address in Received:
3507 header lines when this was not necessary.
3509 PH/27 Added the % operator to ${eval:}.
3511 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3512 it should be ignoring failures (because with -C, for example, it has lost
3513 privilege). It wasn't ignoring creation failures other than "already
3516 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3519 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3520 because an input line was too long, either on its own, or by virtue of
3521 too many continuations, the temporary file was not being removed, and the
3522 return code was incorrect.
3524 PH/31 Missing "BOOL" in function definition in filtertest.c.
3526 PH/32 Applied Sieve patches from the maintainer.
3528 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3530 PH/33 Added "verify = not_blind".
3532 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3533 Local/Makefile (with some defaults set). These are used in built scripts
3534 such as exicyclog, but they have never been used in the exim_install
3535 script (though there are many overriding facilities there). I have
3536 arranged that the exim_install script now takes note of these two
3539 PH/35 Installed configuration files for Dragonfly.
3541 PH/36 When a locally submitted message by a trusted user did not contain a
3542 From: header, and the sender address was obtained from -f or from an SMTP
3543 MAIL command, and the trusted user did not use -F to supply a sender
3544 name, $originator_name was incorrectly used when constructing a From:
3545 header. Furthermore, $originator_name was used for submission mode
3546 messages from external hosts without From: headers in a similar way,
3547 which is clearly wrong.
3549 PH/37 Added control=suppress_local_fixups.
3551 PH/38 When log_selector = +received_sender was set, and the addition of the
3552 sender made the log line's construction buffer exactly full, or one byte
3553 less than full, an overflow happened when the terminating "\n" was
3556 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3557 when the result of a list match is failure because a DNS lookup failed.
3559 PH/40 RM_COMMAND is now used in the building process.
3561 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3562 the "build-* directories that it finds.
3564 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3565 address in a domain literal was a prefix of an interface address.
3567 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3568 when verifying a sender address, unless rewrite_headers is false.
3570 PH/44 Wrote a long comment about why errors_to addresses are verified as
3571 recipients, not senders.
3573 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3574 the ratelimit ACL was added.
3576 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3578 PH/46 Added extra information about PostgreSQL errors to the error string.
3580 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3581 avoiding the need to install special zones in a real server. This is
3582 backwards compatible; if it can't find the fake resolver, it drops back.
3583 Thus, both old and new test suites can be run.
3585 TF/02 Added util/ratelimit.pl
3587 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3588 clock is set back in time.
3590 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3591 Candler <B.Candler@pobox.com>.
3593 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3594 OK for addresses that are the result of redirections.
3596 PH/48 A number of further additions for the benefit of the new test suite,
3597 including a fake gethostbyname() that interfaces to the fake DNS resolver
3600 TF/06 The fix for widen_domains has also been applied to qualify_single and
3601 search_parents which are the other dnslookup options that can cause
3604 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3607 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3609 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3610 service" if it encountered an MX record with an empty target hostname.
3611 The message is now "an MX or SRV record indicated no SMTP service".
3613 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3614 if the DNS of the sending site is misconfigured. This is quite a
3615 common situation. This change restores the behaviour of treating a
3616 helo verification defer as a failure.
3618 PH/51 If self=fail was set on a router, the bounce message did not include the
3619 actual error message.
3625 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3627 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3628 command itself ended in error, the underlying message about the transport
3629 filter timeout was being overwritten with the pipe command error. Now the
3630 underlying error message should be appended to the second error message.
3632 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3634 PH/02 Reduce the amount of output that "make" produces by default. Full output
3635 can still be requested.
3637 PH/03 The warning log line about a condition test deferring for a "warn" verb
3638 was being output only once per connection, rather than after each
3639 occurrence (because it was using the same function as for successful
3640 "warn" verbs). This seems wrong, so I have changed it.
3642 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3643 it should not have, which might have caused a crash in the right
3644 circumstances, but probably never did.
3646 PH/04 Installed a modified version of Tony Finch's patch to make submission
3647 mode fix the return path as well as the Sender: header line, and to
3648 add a /name= option so that you can make the user's friendly name appear
3651 TF/03 Added the control = fakedefer ACL modifier.
3653 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3654 Mark Lowes for thorough testing.
3656 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3658 TK/03 Merged latest SRS patch from Miles Wilton.
3660 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3661 with the definition in sysexits.h (which is #included earlier).
3662 Fortunately, Exim does not actually use EX_OK. The code used to try to
3663 preserve the sysexits.h value, by assuming that macro definitions were
3664 scanned for macro replacements. I have been disabused of this notion,
3665 so now the code just undefines EX_OK before #including unistd.h.
3667 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3668 in the smtp transport. When a block could not be written in a single
3669 write() function, the timeout was being re-applied to each part-write.
3670 This seems wrong - if the receiver was accepting one byte at a time it
3671 would take for ever. The timeout is now adjusted when this happens. It
3672 doesn't have to be particularly precise.
3674 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3675 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3677 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3678 without a domain if the check to <postmaster@domain> fails.
3680 SC/01 Eximstats: added -xls and the ability to specify output files
3681 (patch written by Frank Heydlauf).
3683 SC/02 Eximstats: use FileHandles for outputting results.
3685 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3687 SC/04 Eximstats: fixed display of large numbers with -nvr option
3689 SC/05 Eximstats: fixed merging of reports with empty tables.
3691 SC/06 Eximstats: added the -include_original_destination flag
3693 SC/07 Eximstats: removed tabs and trailing whitespace.
3695 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3697 TK/06 MBOX spool code: Add real "From " MBOX separator line
3698 so the .eml file is really in mbox format (even though
3699 most programs do not really care). Patch from Alex Miller.
3701 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3702 The latter is generated from $received_to and is only set if the
3703 message has one envelope recipient. SA can use these headers,
3704 obviously out-of-the-box. Patch from Alex Miller.
3706 PH/08 The ${def test on a variable was returning false if the variable's
3707 value was "0", contrary to what the specification has always said!
3708 The result should be true unless the variable is empty.
3710 PH/09 The syntax error of a character other than { following "${if
3711 def:variable_name" (after optional whitespace) was not being diagnosed.
3712 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3713 accidental colon was present, for example, could give incorrect results.
3715 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3716 result is used (not including appendfile, where other changes are about
3719 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3720 This involved changing a lot of size variables from int to off_t. It
3721 should work with maildirs and everything.
3723 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3724 spamd dying while we are connected to it.
3726 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3729 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3730 function for BDB 4.3.
3732 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3734 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3735 the sources for PCRE have been re-arranged and more files are now
3738 PH/15 The code I had for printing potentially long long variables in PH/11
3739 above was not the best (it lost precision). The length of off_t variables
3740 is now inspected at build time, and an appropriate printing format (%ld
3741 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3742 to be "long long int" or "long int". This is needed for the internal
3743 formatting function string_vformat().
3745 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3746 the configuration file to be ":syslog", then the script "guesses" where
3747 the logs files are, rather than using the compiled in default. In our
3748 case the guess is not the same as the compiled default, so the script
3749 suddenly stopped working when I started to use syslog. The patch checks
3750 to see if log_file_path is "". If so, it attempts to read it from exim
3751 with no configuration file to get the compiled in version, before it
3752 falls back to the previous guessing code."
3754 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3755 implementing BATV in an Exim configuration. See NewStuff for the gory
3758 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3759 Makefile that are specific to HP-UX.
3761 PH/18 If the "use_postmaster" option was set for a recipient callout together
3762 with the "random" option, the postmaster address was used as the MAIL
3763 FROM address for the random test, but not for the subsequent recipient
3764 test. It is now used for both.
3766 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3767 patch removes a few documentation additions to RFC 3028, because the
3768 latest draft now contains them. It adds the new en;ascii-case comparator
3769 and a new error check for 8bit text in MIME parts. Comparator and
3770 require names are now matched exactly. I enabled the subaddress
3771 extension, but it is not well tested yet (read: it works for me)."
3773 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3774 rework some of the code of TK/09 above to avoid the hardwired use of
3775 "%lld" and "long long". Replaced the call to snprintf() with a call to
3778 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3779 records point to non-existent hosts", "retry timeout exceeded", and
3780 "retry time not reached for any host after a long failure period".
3782 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3783 experimental DomainKeys support:
3785 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3786 (2) On an error such as an illegally used "control", the wrong name for
3787 the control was given.
3789 These problems did NOT occur unless DomainKeys support was compiled.
3791 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3793 PH/24 Added ${if match_ip condition.
3795 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3796 databases so that it will be absolutely obvious if a crash occurs in the
3797 DB library. This is a regular occurrence (often caused by mis-matched
3800 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3801 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3802 picked up on a user's system that detects such things. There doesn't seem
3803 to be a gcc warning option for this - only an attribute that has to be
3804 put on the function's prototype. It seems that in Fedora Core 4 they have
3805 set this on a number of new functions. No doubt there will be more in due
3808 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3809 specify a transport. However, if an address that was verified by such a
3810 router was the subject of a callout, Exim crashed because it tried to
3811 read the rcpt_include_affixes from the non-existent transport. Now it
3812 just assumes that the setting of that option is false. This bug was
3813 introduced by 4.51/PH/31.
3815 PH/28 Changed -d+all to exclude +memory, because that information is very
3816 rarely of interest, but it makes the output a lot bigger. People tend to
3817 do -d+all out of habit.
3819 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3820 code in os-type was giving problems when libc.so lives in lib64, like on
3823 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3824 aren't the modern standard, and it seems that some systems' include files
3825 don't always have them. Exim was already checking for some of the newer
3826 ones like T_AAAA, and defining it itself. I've added checks for all the
3827 record types that Exim uses.
3829 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3830 not automatically generating a new one, as it is supposed to. This
3831 prevented TLS from working. If the file did exist, but contained invalid
3832 data, a new version was generated, as expected. It was only the case of a
3833 non-existent file that was broken.
3835 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3836 with a change in libdomainkeys > 0.64.
3838 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3839 from DNS. If the selector record carries the flag, it now has
3840 precedence over the domain-wide flag.
3842 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3844 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3845 the use of a port name that isn't defined in /etc/services) occurred, the
3846 message was deferred as in a normal delivery, and thus remained on the
3847 spool, instead of being failed because of the mua_wrapper setting. This
3848 is now fixed, and I tidied up some of the mua_wrapper messages at the
3851 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3852 the messages in a hash of arrays rather than using individual hashes.
3853 This is a bit cleaner and results in dramatic memory savings, albeit
3854 at a slight CPU cost.
3856 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3857 as requested by Marc Sherman.
3859 SC/10 Eximstats: added histograms for user specified patterns as requested
3862 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3864 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3865 fopen() in the content-scanning modules that did not already have it.
3871 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3872 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3874 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3876 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3878 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3879 to test Sieve filters that use "vacation".
3881 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3882 that changes the way the GnuTLS parameters are stored in the cache file.
3883 The new format can be generated externally. For backward compatibility,
3884 if the data in the cache doesn't make sense, Exim assumes it has read an
3885 old-format file, and it generates new data and writes a new file. This
3886 means that you can't go back to an older release without removing the
3889 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3890 work if there are any delivery delays because "one_time" forces the
3891 parent to be marked "delivered", so its unseen clone is never tried
3892 again. For this reason, Exim now forbids the simultaneous setting of
3895 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3896 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3897 entry if you want to know the details. The fix, however, neglected to
3898 consider the case when local delivery batching is involved. The test for
3899 "previously delivered" was not happening when checking to see if an
3900 address could be batched with a previous (undelivered) one; under
3901 certain circumstances this could lead to multiple deliveries to the same
3904 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3905 in its include files, and this causes problems building Exim.
3907 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3908 header_syntax) but Exim was just ignoring anything given after a slash.
3909 In particular, this caused confusion with an attempt to use "verify =
3910 reverse_host_lookup/defer_ok". An error is now given when options are
3911 supplied for verify items that do not have them. (Maybe reverse_host_
3912 lookup should have a defer_ok option, but that's a different point.)
3914 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3915 defined by RFC 821) to 2048, because there were problems with some AUTH
3916 commands, and RFC 1869 says the size should be increased for extended
3917 SMTP commands that take arguments.
3919 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3922 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3923 "unknown" error; now it says that the functionality isn't in the binary.
3925 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3926 an address' error message when a string expansion fails (syntax or
3927 whatever). Otherwise the password may appear in the log. Following change
3928 PH/42 below, there is no longer a chance of it appearing in a bounce
3931 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3933 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3934 of its addresses was ever tried. (Bugzilla bug #2.)
3936 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3937 the result incorrectly in the debug output. (It correctly added a newline
3938 to what was transported.)
3940 TF/01 Added $received_time.
3942 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3943 commented out examples of how to interface to a virus scanner and to
3944 SpamAssassin. Also added commented examples of av_scanner and
3945 spamd_address settings.
3947 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3948 and controls are allowed in which ACLs. There were a couple of minor
3949 errors. Some of the entries in the conditions table (which is a table of
3950 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3951 negation of where the condition IS allowed.
3953 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3955 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3956 header file does not have a version number, so I've had to invent a new
3957 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3958 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3959 radiusclient), but it was contributed by a Radius user.
3961 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3962 files or directories, for OpenSSL.
3964 PH/19 When an Exim process that is running as root has to create an Exim log
3965 file, it does so in a subprocess that runs as exim:exim so as to get the
3966 ownership right at creation (otherwise, other Exim processes might see
3967 the file with the wrong ownership). There was no test for failure of this
3968 fork() call, which would lead to the process getting stuck as it waited
3969 for a non-existent subprocess. Forks do occasionally fail when resources
3970 run out. I reviewed all the other calls to fork(); they all seem to check
3973 PH/20 When checking for unexpected SMTP input at connect time (before writing
3974 the banner), Exim was not dealing correctly with a non-positive return
3975 from the read() function. If the client had disconnected by this time,
3976 the result was a log entry for a synchronization error with an empty
3977 string after "input=" when read() returned zero. If read() returned -1
3978 (an event I could not check), uninitialized data bytes were printed.
3979 There were reports of junk text (parts of files, etc) appearing after
3982 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3984 PH/22 Added support for macro redefinition, and (re)definition in between
3985 driver and ACL definitions.
3987 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3988 forgetting to use the resulting value; it was using the unexpanded value.
3990 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3991 hadn't been configured. The fix is from Juergen Kreileder, who
3992 understands it better than I do:
3994 "Here's what I see happening with three configured cyrus_sasl
3995 authenticators configured (plain, login, cram-md5):
3997 On startup auth_cyrus_sasl_init() gets called for each of these.
3998 This means three calls to sasl_listmech() without a specified mech_list.
3999 => SASL tests which mechs of all available mechs actually work
4000 => three warnings about OTP not working
4001 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4003 With the patch, sasl_listmech() also gets called three times. But now
4004 SASL's mech_list option is set to the server_mech specified in the the
4005 authenticator. Or in other words, the answer from sasl_listmech()
4006 gets limited to just the mech you're testing for (which is different
4008 => the return list contains just 'plain' or 'login', 'cram-md5' or
4009 nothing depending on the value of ob->server_mech.
4011 I've just tested the patch: Authentication still works fine,
4012 unavailable mechs specified in the exim configuration are still
4013 caught, and the auth.log warnings about OTP are gone."
4015 PH/25 When debugging is enabled, the contents of the command line are added
4016 to the debugging output, even when log_selector=+arguments is not
4019 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4020 answer is "GNU", and only if the return is "GNU/something" is the answer
4023 PH/27 $acl_verify_message is now set immediately after the failure of a
4024 verification in an ACL, and so is available in subsequent modifiers. In
4025 particular, the message can be preserved by coding like this:
4027 warn !verify = sender
4028 set acl_m0 = $acl_verify_message
4030 Previously, $acl_verify_message was set only while expanding "message"
4031 and "log_message" when a very denied access.
4033 PH/28 Modified OS/os.c-Linux with
4035 -#ifndef OS_LOAD_AVERAGE
4036 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4038 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4039 nomenclature these days.)
4041 PH/29 Installed patch from the Sieve maintainer that adds the options
4042 sieve_useraddress and sieve_subaddress to the redirect router.
4044 PH/30 In these circumstances:
4045 . Two addresses routed to the same list of hosts;
4046 . First host does not offer TLS;
4047 . First host accepts first address;
4048 . First host gives temporary error to second address;
4049 . Second host offers TLS and a TLS session is established;
4050 . Second host accepts second address.
4051 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4052 and peerdn, if requested) that were in fact used only for the second
4055 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4056 attention to any local part prefix or suffix that was matched by the
4057 router that accepted the address. It now behaves in the same way as it
4058 does for delivery: the affixes are removed from the local part unless
4059 rcpt_include_affixes is set on the transport.
4061 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4062 timeout during the DATA phase of an incoming message.
4064 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4065 applied a patch sanctioned by the Sieve maintainer.
4067 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4068 the uid or gid is negative. A case of a negative gid caused this to be
4069 noticed. The fix allows for either to be negative.
4071 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4072 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4075 PH/36 The change PH/12 above was broken. Fixed it.
4077 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4078 the grounds that routing the same address twice would always produce the
4079 same answer. This might have been true once, but it is certainly no
4080 longer true now. Routing a child address may depend on the previous
4081 routing that produced that child. Some complicated redirection strategies
4082 went wrong when messages had multiple recipients, and made Exim's
4083 behaviour dependent on the order in which the addresses were given.
4085 I have moved the duplicate checking until after the routing is complete.
4086 Exim scans the addresses that are assigned to local and remote
4087 transports, and removes any duplicates. This means that more work will be
4088 done, as duplicates will always all be routed, but duplicates are
4089 presumably rare, so I don't expect this is of any significance.
4091 For deliveries to pipes, files, and autoreplies, the duplicate checking
4092 still happens during the routing process, since they are not going to be
4095 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4096 It corrects a timeout issue with spamd. This is Ian's comment: "The
4097 background is that sometimes spamd either never reads data from a
4098 connection it has accepted, or it never writes response data. The exiscan
4099 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4100 blindly assumes that writes won't block so it may never time out."
4102 PH/39 Allow G after quota size as well as K and M.
4104 PH/40 The value set for $authenticated_id in an authenticator may not contain
4105 binary zeroes or newlines because the value is written to log lines and
4106 to spool files. There was no check on this. Now the value is run through
4107 the string_printing() function so that such characters are converted to
4108 printable escape sequences.
4110 PH/41 $message_linecount is a new variable that contains the total number of
4111 lines in the message. Compare $body_linecount, which is the count for the
4114 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4115 bounce and delay warning messages, except in certain special cases, which
4118 (a) An SMTP error message from a remote host;
4119 (b) A message specified in a :fail: redirection;
4120 (c) A message specified in a "fail" command in a system filter;
4121 (d) A message specified in a FAIL return from the queryprogram router;
4122 (e) A message specified by the cannot_route_message router option.
4124 In these cases only, Exim does include the error details in bounce and
4125 warning messages. There are also a few cases where bland messages such
4126 as "unrouteable address" or "local delivery error" are given.
4128 PH/43 $value is now also set for the "else" part of a ${run expansion.
4130 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4131 being worked on, but at least Exim now implements the latest version to
4134 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4135 process to complete was treated as a delivery failure, a timeout while
4136 writing the message to the pipe was logged, but erroneously treated as a
4137 successful delivery. Such timeouts include transport filter timeouts. For
4138 consistency with the overall process timeout, these timeouts are now
4139 treated as errors, giving rise to delivery failures by default. However,
4140 there is now a new Boolean option for the pipe transport called
4141 timeout_defer, which, if set TRUE, converts the failures into defers for
4142 both kinds of timeout. A transport filter timeout is now identified in
4145 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4146 systems where "make" and "gmake" are different, calling "gmake" at top
4147 level broke things. I've arranged for the value of $(MAKE) to be passed
4148 from the Makefile to this script so that it can call the same version of
4152 A note about Exim versions 4.44 and 4.50
4153 ----------------------------------------
4155 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4156 changes of various kinds. As a consequence, a big documentation update was
4157 needed. This delayed the release for rather longer than seemed good, especially
4158 in the light of a couple of (minor) security issues. Therefore, the changes
4159 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4160 release. So 4.44 and 4.50 are in effect two different branches that both start
4163 I have left the 4.50 change log unchanged; it contains all the changes since
4164 4.43. The change log for 4.44 is below; many of its items are identical to
4165 those for 4.50. This seems to be the most sensible way to preserve the
4166 historical information.
4172 1. Minor wording change to the doc/README.SIEVE file.
4174 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4175 computation of the current number of files was incorrect.
4177 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4178 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4181 4. Give more explanation in the error message when the command for a transport
4182 filter fails to execute.
4184 5. There are several places where Exim runs a non-Exim command in a
4185 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4186 was being done only for the command run by the queryprogram router. It is
4187 now done for all such subprocesses. The other cases are: ${run, transport
4188 filters, and the commands run by the lmtp and pipe transports.
4190 6. Added CONFIGURE_GROUP build-time option.
4192 7. Some older OS have a limit of 256 on the maximum number of file
4193 descriptors. Exim was using setrlimit() to set 1000 as a large value
4194 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4195 systems. I've change it so that if it can't get 1000, it tries for 256.
4197 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4198 was an oversight, and furthermore, ever since the addition of extra
4199 controls (e.g. 4.43/32), the checks on when to allow different forms of
4200 "control" were broken. There should now be diagnostics for all cases when a
4201 control that does not make sense is encountered.
4203 9. Added the /retain_sender option to "control=submission".
4205 10. $recipients is now available in the predata ACL (oversight).
4207 11. Tidy the search cache before the fork to do a delivery from a message
4208 received from the command line. Otherwise the child will trigger a lookup
4209 failure and thereby defer the delivery if it tries to use (for example) a
4210 cached ldap connection that the parent has called unbind on.
4212 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4213 of $address_data from the recipient verification was clobbered by the
4214 sender verification.
4216 13. The value of address_data from a sender verification is now available in
4217 $sender_address_data in subsequent conditions in the ACL statement.
4219 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4221 15. Added a new option "connect=<time>" to callout options, to set a different
4224 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4225 was its contents. (It was OK if the option was not defined at all.)
4227 17. A "Completed" log line is now written for messages that are removed from
4228 the spool by the -Mrm option.
4230 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4231 information about exactly what failed.
4233 19. Added -dd to debug only the daemon process.
4235 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4236 handles timeouts, both on the server side and network timeouts. Renamed the
4237 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4239 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4240 It is now set to "smtps".
4242 22. $host_address is now set to the target address during the checking of
4243 ignore_target_hosts.
4245 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4246 being passed; this would have caused $sender_host_name to have been used if
4247 matching the list had actually called for a host name (not very likely,
4248 since this list is usually IP addresses). A host name is now passed as
4251 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4252 code that specifies a non-restarting handler (typically sigaction() in
4253 modern systems) in an attempt to fix a rare and obscure crash bug.
4255 25. Narrowed the window for a race in the daemon that could cause it to ignore
4256 SIGCHLD signals. This is not a major problem, because they are used only to
4257 wake it up if nothing else does.
4259 26. A malformed maildirsize file could cause Exim to calculate negative values
4260 for the mailbox size or file count. Odd effects could occur as a result.
4261 The maildirsize information is now recalculated if the size or filecount
4264 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4265 support for a long time. Removed HAVE_SYS_VFS_H.
4267 28. Installed the latest version of exipick from John Jetmore.
4269 29. In an address list, if the pattern was not a regular expression, an empty
4270 subject address (from a bounce message) matched only if the pattern was an
4271 empty string. Non-empty patterns were not even tested. This was the wrong
4272 because it is perfectly reasonable to use an empty address as part of a
4273 database query. An empty address is now tested by patterns that are
4274 lookups. However, all the other forms of pattern expect the subject to
4275 contain a local part and a domain, and therefore, for them, an empty
4276 address still always fails if the pattern is not itself empty.
4278 30. Exim went into a mad DNS loop when attempting to do a callout where the
4279 host was specified on an smtp transport, and looking it up yielded more
4280 than one IP address.
4282 31. Re-factored the code for checking spool and log partition space into a
4283 function that finds that data and another that does the check. The former
4284 is then used to implement four new variables: $spool_space, $log_space,
4285 $spool_inodes, and $log_inodes.
4287 32. The RFC2047 encoding function was originally intended for short strings
4288 such as real names; it was not keeping to the 75-character limit for
4289 encoded words that the RFC imposes. It now respects the limit, and
4290 generates multiple encoded words if necessary. To be on the safe side, I
4291 have increased the buffer size for the ${rfc2047: expansion operator from
4294 33. It is now permitted to omit both strings after an "if" condition; if the
4295 condition is true, the result is "true". As before, when the second string
4296 is omitted, a false condition yields an empty string. This makes it less
4297 cumbersome to write custom ACL and router conditions.
4299 34. Failure to deliver a bounce message always caused it to be frozen, even if
4300 there was an errors_to setting on the router. The errors_to setting is now
4303 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4304 canonical form (fully expanded) before being placed in
4305 $sender_host_address.
4307 36. The table in the code that translates DNS record types into text (T_A to
4308 "A" for instance) was missing entries for NS and CNAME. It is just possible
4309 that this could have caused confusion if both these types were looked up
4310 for the same domain, because the text type is used as part of Exim's
4311 per-process caching. But the chance of anyone hitting this buglet seems
4314 37. The dnsdb lookup has been extended in a number of ways.
4316 (1) There is a new type, "zns", which walks up the domain tree until it
4317 finds some nameserver records. It should be used with care.
4319 (2) There is a new type, "mxh", which is like "mx" except that it returns
4320 just the host names, not the priorities.
4322 (3) It is now possible to give a list of domains (or IP addresses) to be
4323 looked up. The behaviour when one of the lookups defers can be
4324 controlled by a keyword.
4326 (4) It is now possible to specify the separator character for use when
4327 multiple records are returned.
4329 38. The dnslists ACL condition has been extended: it is now possible to supply
4330 a list of IP addresses and/or domains to be looked up in a particular DNS
4333 39. Added log_selector=+queue_time_overall.
4335 40. When running the queue in the test harness, wait just a tad after forking a
4336 delivery process, to get repeatability of debugging output.
4338 41. Include certificate and key file names in error message when GnuTLS fails
4339 to set them up, because the GnuTLS error message doesn't include the name
4340 of the failing file when there is a problem reading it.
4342 42. Allow both -bf and -bF in the same test run.
4344 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4346 44. The "Exiscan patch" is now merged into the mainline Exim source.
4348 45. Sometimes the final signoff response after QUIT could fail to get
4349 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4350 < 0 before doing a fflush(). This bug looks as though it goes back to the
4351 introduction of TLS in release 3.20, but "sometimes" must have been rare
4352 because the tests only now provoked it.
4354 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4355 (this can affect the format of dates).
4357 47. exim_tidydb, when checking for the continued existence of a message for
4358 which it has found a message-specific retry record, was not finding
4359 messages that were in split spool directories. Consequently, it was
4360 deleting retry records that should have stayed in existence.
4362 48. Steve fixed some bugs in eximstats.
4364 49. The SPA authentication driver was not abandoning authentication and moving
4365 on to the next authenticator when an expansion was forced to fail,
4366 contradicting the general specification for all authenticators. Instead it
4367 was generating a temporary error. It now behaves as specified.
4369 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4370 (the order specifies the preference for clients). The order is now AES256,
4371 AES128, 3DES, ARCFOUR128.
4373 51. Small patch to Sieve code - explicitly set From: when generating an
4376 52. Exim crashed if a remote delivery caused a very long error message to be
4377 recorded - for instance if somebody sent an entire SpamAssassin report back
4378 as a large number of 550 error lines. This bug was coincidentally fixed by
4379 increasing the size of one of Exim's internal buffers (big_buffer) that
4380 happened as part of the Exiscan merge. However, to be on the safe side, I
4381 have made the code more robust (and fixed the comments that describe what
4384 53. Now that there can be additional text after "Completed" in log lines (if
4385 the queue_time_overall log selector is set), a one-byte patch to exigrep
4386 was needed to allow it to recognize "Completed" as not the last thing in
4389 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4390 patch that reportedly fixes this has been added. I am not expert enough to
4391 create a test for it. This is what the patch creator wrote:
4393 "I found a little strange behaviour of ldap code when working with
4394 Windows 2003 AD Domain, where users was placed in more than one
4395 Organization Units. When I tried to give exim partial DN, the exit code
4396 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4397 But simultaneously result of request was absolutely normal ldap result,
4398 so I produce this patch..."
4400 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4401 REFERENCE, so I have modified the code to exclude the patch when that macro
4404 55. Some experimental protocols are using DNS PTR records for new purposes. The
4405 keys for these records are domain names, not reversed IP addresses. The
4406 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4407 leaves it alone. Component reversal etc. now happens only for IP addresses.
4410 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4412 57. Double the size of the debug message buffer (to 2048) so that more of very
4413 long debug lines gets shown.
4415 58. The exicyclog utility now does better if the number of log files to keep
4416 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4418 59. Two changes related to the smtp_active_hostname option:
4420 (1) $smtp_active_hostname is now available as a variable.
4421 (2) The default for smtp_banner uses $smtp_active_hostname instead
4422 of $primary_hostname.
4424 60. The host_aton() function is supposed to be passed a string that is known
4425 to be a valid IP address. However, in the case of IPv6 addresses, it was
4426 not checking this. This is a hostage to fortune. Exim now panics and dies
4427 if the condition is not met. A case was found where this could be provoked
4428 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4429 components; fortuitously, this particular loophole had already been fixed
4430 by change 4.50/55 above.
4432 If there are any other similar loopholes, the new check in host_aton()
4433 itself should stop them being exploited. The report I received stated that
4434 data on the command line could provoke the exploit when Exim was running as
4435 exim, but did not say which command line option was involved. All I could
4436 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4437 running as the user.
4440 61. There was a buffer overflow vulnerability in the SPA authentication code
4441 (which came originally from the Samba project). I have added a test to the
4442 spa_base64_to_bits() function which I hope fixes it.
4445 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4446 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4448 63. The daemon start-up calls getloadavg() while still root for those OS that
4449 need the first call to be done as root, but it missed one case: when
4450 deliver_queue_load_max is set with deliver_drop_privilege. This is
4451 necessary for the benefit of the queue runner, because there is no re-exec
4452 when deliver_drop_privilege is set.
4454 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4455 This has been fixed.
4457 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4458 was in use, was not putting the data itself into the right store pool;
4459 consequently, it could be overwritten for a subsequent message in the same
4460 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4463 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4465 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4466 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4467 the function were treating the return as a boolean value, which happened to
4468 work because 0=false and not-0=true, but is not correct code.
4470 68. The host_aton() function was not handling scoped IPv6 addresses (those
4471 with, for example, "%eth0" on the end) correctly.
4473 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4474 negated items (that is, ~something) in unsigned ints. Some compilers
4475 apparently mutter when there is no cast.
4477 70. If an address verification called from an ACL failed, and did not produce a
4478 user-specific message (i.e. there was only a "system" message), nothing was
4479 put in $acl_verify_message. In this situation, it now puts the system
4482 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4483 session; change 4.31/43 added the unwanted input to the log line - except
4484 that it did not do this in the start of session case. It now does.
4486 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4487 This is wrong and can cause the other end to generate a synchronization
4488 error if it is another Exim or anything else that does the synchronization
4489 check. A QUIT command is no longer sent after a timeout.
4491 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4492 during host lookups.
4494 74. The defer_ok option of callout verification was not working if it was used
4495 when verifying addresses in header lines, that is, for this case:
4497 verify = header_sender/callout=defer_ok
4499 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4500 those file descriptors could be used for SMTP connections. If anything
4501 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4502 could be sent to the SMTP client, causing chaos. The daemon now opens
4503 stdin, stdout, and stderr to /dev/null when it puts itself into the
4506 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4507 log by default. The user can override this with suitable Perl magic.
4509 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4510 the log message when discard triggers, was not working for the DATA ACL or
4511 for the non-SMTP ACL.
4513 78. Error message wording change in sieve.c.
4515 79. If smtp_accept_max_per_host was set, the number of connections could be
4516 restricted to fewer than expected, because the daemon was trying to set up
4517 a new connection before checking whether the processes handling previous
4518 connections had finished. The check for completed processes is now done
4519 earlier. On busy systems, this bug wouldn't be noticed because something
4520 else would have woken the daemon, and it would have reaped the completed
4523 80. If a message was submitted locally by a user whose login name contained one
4524 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4525 It caused a spool format error. I have fixed the spool reading code. A
4526 related problem was that the "from" clause in the Received: line became
4527 illegal because of the space(s). It is now covered by ${quote_local_part.
4529 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4532 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4534 83. Patch from Sieve maintainer to fix unterminated string problem in
4535 "vacation" handling.
4537 84. Some minor changes to the Linux configuration files to help with other
4538 OS variants using glibc.
4540 85. One more patch for Sieve to update vacation handling to latest spec.
4543 ----------------------------------------------------
4544 See the note above about the 4.44 and 4.50 releases.
4545 ----------------------------------------------------
4551 1. Change 4.43/35 introduced a bug that caused file counts to be
4552 incorrectly computed when quota_filecount was set in an appendfile
4555 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4556 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4559 3. Give more explanation in the error message when the command for a transport
4560 filter fails to execute.
4562 4. There are several places where Exim runs a non-Exim command in a
4563 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4564 was being done only for the command run by the queryprogram router. It is
4565 now done for all such subprocesses. The other cases are: ${run, transport
4566 filters, and the commands run by the lmtp and pipe transports.
4568 5. Some older OS have a limit of 256 on the maximum number of file
4569 descriptors. Exim was using setrlimit() to set 1000 as a large value
4570 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4571 systems. I've change it so that if it can't get 1000, it tries for 256.
4573 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4574 was an oversight, and furthermore, ever since the addition of extra
4575 controls (e.g. 4.43/32), the checks on when to allow different forms of
4576 "control" were broken. There should now be diagnostics for all cases when a
4577 control that does not make sense is encountered.
4579 7. $recipients is now available in the predata ACL (oversight).
4581 8. Tidy the search cache before the fork to do a delivery from a message
4582 received from the command line. Otherwise the child will trigger a lookup
4583 failure and thereby defer the delivery if it tries to use (for example) a
4584 cached ldap connection that the parent has called unbind on.
4586 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4587 of $address_data from the recipient verification was clobbered by the
4588 sender verification.
4590 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4591 was its contents. (It was OK if the option was not defined at all.)
4593 11. A "Completed" log line is now written for messages that are removed from
4594 the spool by the -Mrm option.
4596 12. $host_address is now set to the target address during the checking of
4597 ignore_target_hosts.
4599 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4600 being passed; this would have caused $sender_host_name to have been used if
4601 matching the list had actually called for a host name (not very likely,
4602 since this list is usually IP addresses). A host name is now passed as
4605 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4606 code that specifies a non-restarting handler (typically sigaction() in
4607 modern systems) in an attempt to fix a rare and obscure crash bug.
4609 15. Narrowed the window for a race in the daemon that could cause it to ignore
4610 SIGCHLD signals. This is not a major problem, because they are used only to
4611 wake it up if nothing else does.
4613 16. A malformed maildirsize file could cause Exim to calculate negative values
4614 for the mailbox size or file count. Odd effects could occur as a result.
4615 The maildirsize information is now recalculated if the size or filecount
4618 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4619 support for a long time. Removed HAVE_SYS_VFS_H.
4621 18. Updated exipick to current release from John Jetmore.
4623 19. Allow an empty sender to be matched against a lookup in an address list.
4624 Previously the only cases considered were a regular expression, or an
4627 20. Exim went into a mad DNS lookup loop when doing a callout where the
4628 host was specified on the transport, if the DNS lookup yielded more than
4631 21. The RFC2047 encoding function was originally intended for short strings
4632 such as real names; it was not keeping to the 75-character limit for
4633 encoded words that the RFC imposes. It now respects the limit, and
4634 generates multiple encoded words if necessary. To be on the safe side, I
4635 have increased the buffer size for the ${rfc2047: expansion operator from
4638 22. Failure to deliver a bounce message always caused it to be frozen, even if
4639 there was an errors_to setting on the router. The errors_to setting is now
4642 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4643 canonical form (fully expanded) before being placed in
4644 $sender_host_address.
4646 24. Updated eximstats to version 1.33
4648 25. Include certificate and key file names in error message when GnuTLS fails
4649 to set them up, because the GnuTLS error message doesn't include the name
4650 of the failing file when there is a problem reading it.
4652 26. Expand error message when OpenSSL has problems setting up cert/key files.
4655 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4656 (this can affect the format of dates).
4658 28. exim_tidydb, when checking for the continued existence of a message for
4659 which it has found a message-specific retry record, was not finding
4660 messages that were in split spool directories. Consequently, it was
4661 deleting retry records that should have stayed in existence.
4663 29. eximstats updated to version 1.35
4664 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4665 1.35 - bugfix such that pie charts by volume are generated correctly
4667 30. The SPA authentication driver was not abandoning authentication and moving
4668 on to the next authenticator when an expansion was forced to fail,
4669 contradicting the general specification for all authenticators. Instead it
4670 was generating a temporary error. It now behaves as specified.
4672 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4673 (the order specifies the preference for clients). The order is now AES256,
4674 AES128, 3DES, ARCFOUR128.
4676 31. Small patch to Sieve code - explicitly set From: when generating an
4679 32. Exim crashed if a remote delivery caused a very long error message to be
4680 recorded - for instance if somebody sent an entire SpamAssassin report back
4681 as a large number of 550 error lines. This bug was coincidentally fixed by
4682 increasing the size of one of Exim's internal buffers (big_buffer) that
4683 happened as part of the Exiscan merge. However, to be on the safe side, I
4684 have made the code more robust (and fixed the comments that describe what
4687 33. Some experimental protocols are using DNS PTR records for new purposes. The
4688 keys for these records are domain names, not reversed IP addresses. The
4689 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4690 leaves it alone. Component reversal etc. now happens only for IP addresses.
4693 34. The host_aton() function is supposed to be passed a string that is known
4694 to be a valid IP address. However, in the case of IPv6 addresses, it was
4695 not checking this. This is a hostage to fortune. Exim now panics and dies
4696 if the condition is not met. A case was found where this could be provoked
4697 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4698 components; fortuitously, this particular loophole had already been fixed
4699 by change 4.50/55 or 4.44/33 above.
4701 If there are any other similar loopholes, the new check in host_aton()
4702 itself should stop them being exploited. The report I received stated that
4703 data on the command line could provoke the exploit when Exim was running as
4704 exim, but did not say which command line option was involved. All I could
4705 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4706 running as the user.
4709 35. There was a buffer overflow vulnerability in the SPA authentication code
4710 (which came originally from the Samba project). I have added a test to the
4711 spa_base64_to_bits() function which I hope fixes it.
4714 36. The daemon start-up calls getloadavg() while still root for those OS that
4715 need the first call to be done as root, but it missed one case: when
4716 deliver_queue_load_max is set with deliver_drop_privilege. This is
4717 necessary for the benefit of the queue runner, because there is no re-exec
4718 when deliver_drop_privilege is set.
4720 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4721 was in use, was not putting the data itself into the right store pool;
4722 consequently, it could be overwritten for a subsequent message in the same
4723 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4726 38. Sometimes the final signoff response after QUIT could fail to get
4727 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4728 < 0 before doing a fflush(). This bug looks as though it goes back to the
4729 introduction of TLS in release 3.20, but "sometimes" must have been rare
4730 because the tests only now provoked it.
4736 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4737 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4738 Now it returns an integer. A number of calls were still expecting a T/F
4739 return. Fortuitously, in all cases, the tests worked in OK situations,
4740 which is the norm. However, things would have gone wrong on any write
4741 failures on the smtp file descriptor. This function is used when sending
4742 messages over SMTP and also when doing verify callouts.
4744 2. When Exim is called to do synchronous delivery of a locally submitted
4745 message (the -odf or -odi options), it no longer closes stderr before doing
4748 3. Implemented the mua_wrapper option.
4750 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4752 5. Implemented the functions header_remove(), header_testname(),
4753 header_add_at_position(), and receive_remove_recipient(), and exported them
4756 6. If an ACL "warn" statement specified the addition of headers, Exim already
4757 inserted X-ACL-Warn: at the start if there was no header name. However, it
4758 was not making this test for the second and subsequent header lines if
4759 there were newlines in the string. This meant that an invalid header could
4760 be inserted if Exim was badly configured.
4762 7. Allow an ACL "warn" statement to add header lines at the start or after all
4763 the Received: headers, as well as at the end.
4765 8. Added the rcpt_4xx retry error code.
4767 9. Added postmaster_mailfrom=xxx to callout verification option.
4769 10. Added mailfrom=xxxx to the callout verification option, for verify=
4772 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4773 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4775 12. Inserted some casts to stop certain compilers warning when using pointer
4776 differences as field lengths or precisions in printf-type calls (mostly
4777 affecting debugging statements).
4779 13. Added optional readline() support for -be (dynamically loaded).
4781 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4782 same clock tick as a message's arrival, so that its received time was the
4783 same as the "first fail" time on the retry record, and that message
4784 remained on the queue past the ultimate address timeout, every queue runner
4785 would try a delivery (because it was past the ultimate address timeout) but
4786 after another failure, the ultimate address timeout, which should have then
4787 bounced the address, did not kick in. This was a "< instead of <=" error;
4788 in most cases the first failure would have been in the next clock tick
4789 after the received time, and all would be well.
4791 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4792 being recognized when the domain list was tested by the match_domain
4793 condition in an expansion string.
4795 16. Added the ${str2b64: operator.
4797 17. Exim was always calling setrlimit() to set a large limit for the number of
4798 processes, without checking whether the existing limit was already
4799 adequate. (It did check for the limit on file descriptors.) Furthermore,
4800 errors from getrlimit() and setrlimit() were being ignored. Now they are
4801 logged to the main and panic logs, but Exim does carry on, to try to do its
4802 job under whatever limits there are.
4804 18. Imported PCRE 5.0.
4806 19. Trivial typo in log message " temporarily refused connection" (the leading
4809 20. If the log selector return_path_on_delivery was set and an address was
4810 redirected to /dev/null, the delivery process crashed because it assumed
4811 that a return path would always be set for a "successful" delivery. In this
4812 case, the whole delivery is bypassed as an optimization, and therefore no
4815 21. Internal re-arrangement: the function for sending a challenge and reading
4816 a response while authentication was assuming a zero-terminated challenge
4817 string. It's now changed to take a pointer and a length, to allow for
4818 binary data in such strings.
4820 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4822 23. Exim was not respecting finduser_retries when seeking the login of the
4823 uid under which it was called; it was always trying 10 times. (The default
4824 setting of finduser_retries is zero.) Also, it was sleeping after the final
4825 failure, which is pointless.
4827 24. Implemented tls_on_connect_ports.
4829 25. Implemented acl_smtp_predata.
4831 26. If the domain in control=submission is set empty, Exim assumes that the
4832 authenticated id is a complete email address when it generates From: or
4833 Sender: header lines.
4835 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4836 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4837 chown and chgrp in /bin and hostname in /usr/bin.
4839 28. Exim was keeping the "process log" file open after each use, just as it
4840 does for the main log. This opens the possibility of it remaining open for
4841 long periods when the USR1 signal hits a daemon. Occasional processlog
4842 errors were reported, that could have been caused by this. Anyway, it seems
4843 much more sensible not to leave this file open at all, so that is what now
4846 29. The long-running daemon process does not normally write to the log once it
4847 has entered its main loop, and it closes the log before doing so. This is
4848 so that log files can straightforwardly be renamed and moved. However,
4849 there are a couple of unusual error situations where the daemon does write
4850 log entries, and I had neglected to close the log afterwards.
4852 30. The text of an SMTP error response that was received during a remote
4853 delivery was being truncated at 512 bytes. This is too short for some of
4854 the long messages that one sometimes sees. I've increased the limit to
4857 31. It is now possible to make retry rules that apply only when a message has a
4858 specific sender, in particular, an empty sender.
4860 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4861 it possible to be selective about when SMTP synchronization is enforced.
4863 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4865 32. Implemented hosts_connection_nolog.
4867 33. Added an ACL for QUIT.
4869 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4872 35. Added mailbox_size and mailbox_filecount to appendfile.
4874 36. Added control = no_multiline_responses to ACLs.
4876 37. There was a bug in the logic of the code that waits for the clock to tick
4877 in the case where the clock went backwards by a substantial amount such
4878 that the microsecond fraction of "now" was more than the microsecond
4879 fraction of "then" (but the whole seconds number was less).
4881 38. Added support for the libradius Radius client library this is found on
4882 FreeBSD (previously only the radiusclient library was supported).
4888 1. When certain lookups returned multiple values in the form name=value, the
4889 quoting of the values was not always being done properly. Specifically:
4890 (a) If the value started with a double quote, but contained no whitespace,
4892 (b) If the value contained whitespace other than a space character (i.e.
4893 tabs or newlines or carriage returns) it was not quoted.
4894 This fix has been applied to the mysql and pgsql lookups by writing a
4895 separate quoting function and calling it from the lookup code. The fix
4896 should probably also be applied to nisplus, ibase and oracle lookups, but
4897 since I cannot test any of those, I have not disturbed their existing code.
4899 2. A hit in the callout cache for a specific address caused a log line with no
4900 reason for rejecting RCPT. Now it says "Previous (cached) callout
4901 verification failure".
4903 3. There was an off-by-one bug in the queryprogram router. An over-long
4904 return line was truncated at 256 instead of 255 characters, thereby
4905 overflowing its buffer with the terminating zero. As well as fixing this, I
4906 have increased the buffer size to 1024 (and made a note to document this).
4908 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4909 when Exim is waiting for an SMTP response from a remote server, Exim
4910 restarts its select() call on the socket, thereby resetting its timeout.
4911 This is not a problem when such interrupts are rare. Somebody set up a cron
4912 job to run exiwhat every 2 minutes, which is less than the normal select()
4913 timeout (5 or 10 minutes). This meant that the select() timeout never
4914 kicked in because it was always reset. I have fixed this by comparing the
4915 time when an interrupt arrives with the time at the start of the first call
4916 to select(). If more time than the timeout has elapsed, the interrupt is
4917 treated as a timeout.
4919 5. Some internal re-factoring in preparation for the addition of Sieve
4920 extensions (by MH). In particular, the "personal" test is moved to a
4921 separate function, and given an option for scanning Cc: and Bcc: (which is
4922 not set for Exim filters).
4924 6. When Exim created an email address using the login of the caller as the
4925 local part (e.g. when creating a From: or Sender: header line), it was not
4926 quoting the local part when it contained special characters such as @.
4928 7. Installed new OpenBSD configuration files.
4930 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4931 try to make them clearer.
4933 9. Callout options, other than the timeout value, were being ignored when
4934 verifying sender addresses in header lines. For example, when using
4936 verify = header_sender/callout=no_cache
4938 the cache was (incorrectly) being used.
4940 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4941 only the Cygwin environment.
4943 11. When return_path_on_delivery was set as a log selector, if different remote
4944 addresses in the same message used different return paths and parallel
4945 remote delivery occurred, the wrong values would sometimes be logged.
4946 (Whenever a remote delivery process finished, the return path value from
4947 the most recently started remote delivery process was logged.)
4949 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4950 lines when AUTH and/or TLS are in use. This is the "received protocol"
4951 field. Exim used to use "asmtp" for authenticated SMTP, without any
4952 indication (in the protocol name) for TLS use. Now it follows the RFC and
4953 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4954 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4955 names appear in log lines as well as in Received: header lines.
4957 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4958 extensions, and comparison tests, and to fix some bugs.
4960 14. Changes to the "personal" filter test:
4962 (1) The test was buggy in that it was just doing the equivalent of
4963 "contains" tests on header lines. For example, if a user's address was
4964 anne@some.where, the "personal" test would incorrectly be true for
4966 To: susanne@some.where
4968 This test is now done by extracting each address from the header in turn,
4969 and checking the entire address. Other tests that are part of "personal"
4970 are now done using regular expressions (for example, to check local parts
4971 of addresses in From: header lines).
4973 (2) The list of non-personal local parts in From: addresses has been
4974 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4975 taken from the Sieve specification recommendations.
4977 (3) If the message contains any header line starting with "List-" it is
4978 treated as non-personal.
4980 (4) The test for "circular" in the Subject: header line has been removed
4981 because it now seems ill-conceived.
4983 15. Minor typos in src/EDITME comments corrected.
4985 16. Installed latest exipick from John Jetmore.
4987 17. If headers_add on a router specified a text string that was too long for
4988 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4989 of string_sprintf() is now avoided.
4991 18. $message_body_size was not set (it was always zero) when running the DATA
4992 ACL and the local_scan() function.
4994 19. For the "mail" command in an Exim filter, no default was being set for
4995 the once_repeat time, causing a random time value to be used if "once" was
4996 specified. (If the value happened to be <= 0, no repeat happened.) The
4997 default is now 0s, meaning "never repeat". The "vacation" command was OK
4998 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4999 (I found it when inspecting the code).
5001 20. There is now an overall timeout for performing a callout verification. It
5002 defaults to 4 times the callout timeout, which applies to individual SMTP
5003 commands during the callout. The overall timeout applies when there is more
5004 than one host that can be tried. The timeout is checked before trying the
5005 next host. This prevents very long delays if there are a large number of
5006 hosts and all are timing out (e.g. when the network connections are timing
5007 out). The value of the overall timeout can be changed by specifying an
5008 additional sub-option for "callout", called "maxwait". For example:
5010 verify = sender/callout=5s,maxwait=20s
5012 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5013 to the end before writing, but this should make it even safer).
5015 22. Exim was forgetting that it had advertised PIPELINING for the second and
5016 subsequent messages on an SMTP connection. It was also not resetting its
5017 memory on STARTTLS and an internal HELO.
5019 23. When Exim logs an SMTP synchronization error within a session, it now
5020 records whether PIPELINING has been advertised or not.
5022 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5023 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5024 rather than long int.
5026 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5028 26. Added the never_mail option to autoreply.
5034 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5035 crash if the getsockname() call failed; this can happen if a connection is
5036 closed very soon after it is established. The problem was simply in the
5037 order in which certain operations were done, causing Exim to try to write
5038 to the SMTP stream before it had set up the file descriptor. The bug has
5039 been fixed by making things happen in the correct order.
5045 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5046 before the connection was closed, thus losing the rejection response.
5048 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5049 some early Solaris releases, but causes trouble in current releases where
5050 socklen_t is defined.
5052 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5055 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5058 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5059 that releasing the top of it at the end releases what was used for sub-
5060 expansions (unless the block got too big). However, discard this block if
5061 the first thing is a variable or header, so that we can use its block when
5062 it is dynamic (useful for very large $message_headers, for example).
5064 6. Lookups now cache *every* query, not just the most recent. A new, separate
5065 store pool is used for this. It can be recovered when all lookup caches are
5066 flushed. Lookups now release memory at the end of their result strings.
5067 This has involved some general refactoring of the lookup sources.
5069 7. Some code has been added to the store_xxx() functions to reduce the amount
5070 of flapping under certain conditions.
5072 8. log_incoming_interface used to affect only the <= reception log lines. Now
5073 it causes the local interface and port to be added to several more SMTP log
5074 lines, for example "SMTP connection from", and rejection lines.
5076 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5078 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5080 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5081 could be overwritten at the end of the current message (or the start of a
5082 new message if it was set in a HELO ACL). The value is now preserved for
5083 the duration of the SMTP connection.
5085 12. If a transport had a headers_rewrite setting, and a matching header line
5086 contained an unqualified address, that address was qualified, even if it
5087 did not match any rewriting rules. The underlying bug was that the values
5088 of the flags that permit the existence of unqualified sender and recipient
5089 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5090 non-local messages, and by -bnq for local messages) were not being
5091 preserved with the message after it was received.
5093 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5094 "next input=" as part of the text comprising the host identity instead of
5095 the correct text. The code was using the same buffer for two different
5096 strings. However, depending on which order the printing function evaluated
5097 its arguments, the bug did not always show up. Under Linux, for example, my
5098 test suite worked just fine.
5100 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5101 doesn't work with some older versions of Perl. It has been changed to "my",
5102 which in any case is probably the better facility to use.
5104 15. A really picky compiler found some instances of statements for creating
5105 error messages that either had too many or two few arguments for the format
5108 16. The size of the buffer for calls to the DNS resolver has been increased
5109 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5110 for addresses that have a lot of PTR records. This alleviates a problem; it
5111 does not fully solve it.
5113 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5114 buffer now truncates the list and logs the incident, which is the same
5115 action as happens when Exim is looking up a host name and its aliases.
5116 Previously in this situation something unpredictable would happen;
5117 sometimes it was "internal error: store_reset failed".
5119 18. If a server dropped the connection unexpectedly when an Exim client was
5120 using GnuTLS and trying to read a response, the client delivery process
5121 crashed while trying to generate an error log message.
5123 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5124 string, for example:
5126 warn message = H1: something\nH2: something
5128 the text was added as a single header line from Exim's point of view
5129 though it ended up OK in the delivered message. However, searching for the
5130 second and subsequent header lines using $h_h2: did not work. This has been
5131 fixed. Similarly, if a system filter added multiple headers in this way,
5132 the routers could not see them.
5134 20. Expanded the error message when iplsearch is called with an invalid key to
5135 suggest using net-iplsearch in a host list.
5137 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5138 ACLs are no longer actually imposed (and a message to that effect is
5141 22. If a "gecos" field in a passwd entry contained escaped characters, in
5142 particular, if it contained a \" sequence, Exim got it wrong when building
5143 a From: or a Sender: header from that name. A second bug also caused
5144 incorrect handling when an unquoted " was present following a character
5145 that needed quoting.
5147 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5148 was not being matched caselessly.
5150 24. Arranged for all hyphens in the exim.8 source to be escaped with
5153 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5154 properly. Recipient callout cache records were still being keyed to include
5155 the sender, even when use_sender was set false. This led to far more
5156 callouts that were necessary. The sender is no longer included in the key
5157 when use_sender is false.
5159 26. Added "control = submission" modifier to ACLs.
5161 27. Added the ${base62d: operator to decode base 62 numbers.
5163 28. dnsdb lookups can now access SRV records.
5165 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5166 the configuration file.
5168 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5169 (-v) mode. This makes the output for a verbose queue run more intelligible.
5171 31. Added a use_postmaster feature to recipient callouts.
5173 32. Added the $body_zerocount variable, containing the number of binary zero
5174 bytes in the message body.
5176 33. The time of last modification of the "new" subdirectory is now used as the
5177 "mailbox time last read" when there is a quota error for a maildir
5180 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5182 35. Added +ignore_unknown as a special item in host lists.
5184 36. Code for decoding IPv6 addresses in host lists is now included, even if
5185 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5186 address was recognized as an IP address, but was then not correctly decoded
5187 into binary, causing unexpected and incorrect effects when compared with
5194 1. Very minor rewording of debugging text in manualroute to say "list of
5195 hosts" instead of "hostlist".
5197 2. If verify=header_syntax was set, and a header line with an unqualified
5198 address (no domain) and a large number of spaces between the end of the
5199 name and the colon was received, the reception process suffered a buffer
5200 overflow, and (when I tested it) crashed. This was caused by some obsolete
5201 code that should have been removed. The fix is to remove it!
5203 3. When running in the test harness, delay a bit after writing a bounce
5204 message to get a bit more predictability in the log output.
5206 4. Added a call to search_tidyup() just before forking a reception process. In
5207 theory, someone could use a lookup in the expansion of smtp_accept_max_
5208 per_host which, without the tidyup, could leave open a database connection.
5210 5. Added the variables $recipient_data and $sender_data which get set from a
5211 lookup success in an ACL "recipients" or "senders" condition, or a router
5212 "senders" option, similar to $domain_data and $local_part_data.
5214 6. Moved the writing of debug_print from before to after the "senders" test
5217 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5218 problems for message scanning, either using a data ACL, or using
5219 local_scan() because the Received: header was not generated till after they
5220 were called (in order to set the time as the time of reception completion).
5221 I have revised the way this works. The header is now generated after the
5222 body is received, but before the ACL or local_scan() are called. After they
5223 are run, the timestamp in the header is updated.
5229 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5230 before starting a queue runner without re-exec. This happened only when
5231 deliver_drop_privilege was set or when the Exim user was set to root. The
5232 effect of the bug was that timeouts during subsequent deliveries caused
5233 crashes instead of being properly handled. The handler is now left at its
5234 default (and expected) setting.
5236 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5237 message, again when deliver_drop_privilege is set or Exim is run as root.
5238 The bug described in (1) was not present in this case, but the tidying up
5239 of the other signals was missing. I have made the two cases consistent.
5241 3. The ignore_target_hosts setting on a manualroute router was being ignored
5242 for hosts that were looked up using the /MX notation.
5244 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5247 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5248 operated on the sender address. After changing the $sender_address to <>
5249 for the sender address verify, Exim was re-instated it as the original
5250 (before rewriting) address, but remembering that it had rewritten it, so it
5251 wasn't rewriting it again. This bug also had the effect of breaking the
5252 sender address verification caching when the sender address was rewritten.
5254 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5255 This has been changed so that if the ip literal address matches
5256 ignore_target_hosts, the router declines.
5258 7. Added expansion conditions match_domain, match_address, and match_local_
5259 part (NOT match_host).
5261 8. The placeholder for the Received: header didn't have a length field set.
5263 9. Added code to Exim itself and to exim_lock to test for a specific race
5264 condition that could lead to file corruption when using MBX delivery. The
5265 issue is with the lockfile that is created in /tmp. If this file is removed
5266 after a process has opened it but before that process has acquired a lock,
5267 there is the potential for a second process to recreate the file and also
5268 acquire a lock. This could lead to two Exim processes writing to the file
5269 at the same time. The added code performs the same test as UW imapd; it
5270 checks after acquiring the lock that its file descriptor still refers to
5271 the same named file.
5273 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5274 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5277 11. Added the smtp_active_hostname option. If used, this will typically be made
5278 to depend on the incoming interface address. Because $interface_address is
5279 not set up until the daemon has forked a reception process, error responses
5280 that can happen earlier (such as "too many connections") no longer contain
5283 12. If an expansion in a condition on a "warn" statement fails because a lookup
5284 defers, the "warn" statement is abandoned, and the next ACL statement is
5285 processed. Previously this caused the whole ACL to be aborted.
5287 13. Added the iplsearch lookup type.
5289 14. Added ident_timeout as a log selector.
5291 15. Added tls_certificate_verified as a log selector.
5293 16. Added a global option tls_require_ciphers (compare the smtp transport
5294 option of the same name). This controls incoming TLS connections.
5296 17. I finally figured out how to make tls_require_ciphers do a similar thing
5297 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5298 before starting the TLS session.
5300 18. Tabs are now shown as \t in -bP output.
5302 19. If the log selector return_path_on_delivery was set, Exim crashed when
5303 bouncing a message because it had too many Received: header lines.
5305 20. If two routers both had headers_remove settings, and the first one included
5306 a superfluous trailing colon, the final name in the first list and the
5307 first name in the second list were incorrectly joined into one item (with a
5308 colon in the middle).
5314 1. Added -C and -D options to the exinext utility, mainly to make it easier
5315 to include in the automated testing, but these could be helpful when
5316 multiple configurations are in use.
5318 2. The exinext utility was not formatting the output nicely when there was
5319 an alternate port involved in the retry record key, nor when there was a
5320 message id as well (for retries that were specific to a specific message
5321 and a specific host). It was also confused by IPv6 addresses, because of
5322 the additional colons they contain. I have fixed the IPv4 problem, and
5323 patched it up to do a reasonable job for IPv6.
5325 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5326 delivery, the log line now contains "pipelined" if PIPELINING was used.
5328 4. An SMTP transport process used to panic and die if the bind() call to set
5329 an explicit outgoing interface failed. This has been changed; it is now
5330 treated in the same way as a connect() failure.
5332 5. A reference to $sender_host_name in the part of a conditional expansion
5333 that was being skipped was still causing a DNS lookup. This no longer
5336 6. The def: expansion condition was not recognizing references to header lines
5337 that used bh_ and bheader_.
5339 7. Added the _cache feature to named lists.
5341 8. The code for checking quota_filecount in the appendfile transport was
5342 allowing one more file than it should have been.
5344 9. For compatibility with Sendmail, the command line option
5352 and sets the incoming protocol and host name (for trusted callers). The
5353 host name and its colon can be omitted when only the protocol is to be set.
5354 Note the Exim already has two private options, -pd and -ps, that refer to
5355 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5356 "s", but I don't think that's a major issue.
5358 10. A number of refactoring changes to the code, none of which should affect
5361 (a) The number of logging options was getting close to filling up the
5362 32-bit word that was used as a bit map. I have split them into two classes:
5363 those that are passed in the argument to log_write(), and those that are
5364 only ever tested independently outside of that function. These are now in
5365 separate 32-bit words, so there is plenty of room for expansion again.
5366 There is no change in the user interface or the logging behaviour.
5368 (b) When building, for example, log lines, the code previously used a
5369 macro that called string_cat() twice, in order to add two strings. This is
5370 not really sufficiently general. Furthermore, there was one instance where
5371 it was actually wrong because one of the argument was used twice, and in
5372 one call a function was used. (As it happened, calling the function twice
5373 did not affect the overall behaviour.) The macro has been replaced by a
5374 function that can join an arbitrary number of extra strings onto a growing
5377 (c) The code for expansion conditions now uses a table and a binary chop
5378 instead of a serial search (which was left over from when there were very
5379 few conditions). Also, it now recognizes conditions like "pam" even when
5380 the relevant support is not compiled in: a suitably worded error message is
5381 given if an attempt is made to use such a condition.
5383 11. Added ${time_interval:xxxxx}.
5385 12. A bug was causing one of the ddress fields not to be passed back correctly
5386 from remote delivery subprocesses. The field in question was not being
5387 subsequently used, so this caused to problems in practice.
5389 13. Added new log selectors queue_time and deliver_time.
5391 14. Might have fixed a bug in maildirsizefile handling that threw up
5392 "unexpected character" debug warnings, and recalculated the data
5393 unnecessarily. In any case, I expanded the warning message to give more
5396 15. Added the message "Restricted characters in address" to the statements in
5397 the default ACL that block characters like @ and % in local parts.
5399 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5400 Three changes have been made:
5402 (a) There was a serious bug; a negative response to MAIL caused the whole
5403 recipient domain to be cached as invalid, thereby blocking all messages
5404 to all local parts at the same domain, from all senders. This bug has
5405 been fixed. The domain is no longer cached after a negative response to
5406 MAIL if the sender used is not empty.
5408 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5411 (c) A new callout option, "use_sender" has been added for people who want
5412 the modified behaviour.
5418 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5421 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5422 indeed breaks things for older releases.
5424 3. Added additional logging to the case where there is a problem reading data
5425 from a filter that is running in a subprocess using a pipe, in order to
5426 try to track down a specific problem.
5428 4. Testing facility fudge: when running in the test harness and attempting
5429 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5430 getting "No route to host". Convert this to a timeout.
5432 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5435 6. Some OS don't have socklen_t but use size_t instead. This affects the
5436 fifth argument of getsockopt() amongst other things. This is now
5437 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5438 can be set for individual OS. I have set it for SunOS5, OSF1, and
5439 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5440 some earlier ones do not.
5442 7. Change 4.30/15 was not doing the test caselessly.
5444 8. The standard form for an IPv6 address literal was being rejected by address
5445 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5446 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5447 this, as well as the form without the "IPv6" on the front (but only when
5448 address literals are enabled, of course).
5450 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5452 10. Exim crashed if a message with an empty sender address specified by -f
5453 encountered a router with an errors_to setting. This could be provoked only
5454 by a command such as
5458 where an empty string was supplied; "<>" did not hit this bug.
5460 11. Installed PCRE release 4.5.
5462 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5463 remained set. It is now erased.
5465 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5466 times from message ids (which are base 36 rather than the normal 62).
5468 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5469 were being counted as actual protocol errors, and logged if the log
5470 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5471 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5472 and DATA following a set of rejected RCPTs do not count as protocol errors.
5473 In other words, Exim assumes they were pipelined, though this may not
5474 actually be the case. Of course, in all cases the client gets an
5475 appropriate error code.
5477 15. If a lookup fails in an ACL condition, a message about the failure may
5478 be available; it is used if testing the ACL cannot continue, because most
5479 such messages specify what the cause of the deferral is. However, some
5480 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5481 that caused an old message to be retained and used if a later statement
5482 caused a defer, replacing the real cause of the deferral.
5484 16. If an IP address had so many PTR records that the DNS lookup buffer
5485 was not large enough to hold them, Exim could crash while trying to process
5486 the truncated data. It now detects and logs this case.
5488 17. Further to 4.21/58, another change has been made: if (and only if) the
5489 first line of a message (the first header line) ends with CRLF, a bare LF
5490 in a subsequent header line has a space inserted after it, so as not to
5491 terminate the header.
5493 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5494 unnecessarily, used atoi() instead of strtol(), and didn't check the
5495 termination when getting file sizes from file names by regex.
5497 19. Completely re-implemented the support for maildirsize files, in the light
5498 of a number of problems with the previous contributed implementation
5499 (4.30/29). In particular:
5501 . If the quota is zero, the maildirsize file is maintained, but no quota is
5504 . If the maildir directory does not exist, it is created before any attempt
5505 to write a maildirsize file.
5507 . The quota value in the file is just a cache; if the quota is changed in
5508 the transport, the new value overrides.
5510 . A regular expression is available for excluding directories from the
5513 20. The autoreply transport checks the characters in options that define the
5514 message's headers; it allows continued headers, but it was checking with
5515 isspace() after an embedded newline instead of explicitly looking for a
5518 21. If all the "regular" hosts to which an address was routed had passed their
5519 expiry times, and had not reached their retry times, the address was
5520 bounced, even if fallback hosts were defined. Now Exim should go on to try
5523 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5524 equivalent code in the SMTP transport. Some hosts send humungous responses
5525 to HELO/EHLO, more than 1024 it seems.
5527 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5528 gives compiler warnings in some environments. I've now done it "properly",
5531 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5532 (because of problems with the built-in one) was declared to return uschar *
5533 instead of char *, causing compiler failure.
5535 25. Fixed a file descriptor leak when processing alias/forward files.
5537 26. Fixed a minor format string issue in dbfn.c.
5539 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5541 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5542 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5545 29. When a sender address is verified, it is cached, to save repeating the test
5546 when there is more than one recipient in a message. However, when the
5547 verification involves a callout, it is possible for different callout
5548 options to be set for different recipients. It is too complicated to keep
5549 track of this in the cache, so now Exim always runs a verification when a
5550 callout is required, relying on the callout cache for the optimization.
5551 The overhead is duplication of the address routing, but this should not be
5554 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5555 to be verified with callout=postmaster, and the main callout worked but the
5556 postmaster check failed, the verification correctly failed. However, if a
5557 subsequent RCPT command asked for sender verification *without* the
5558 postmaster check, incorrect caching caused this verification also to fail,
5561 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5562 it was not caching the DNS options (qualify_single, search_parents) that
5563 were used when the lookup failed. A subsequent lookup with different
5564 options therefore always gave the same answer, though there were cases
5565 where it should not have. (Example: a "domains = !$mx_any" option on a
5566 dnslookup router: the "domains" option is always processed without any
5567 widening, but the router might have qualify_single set.) Now Exim uses the
5568 cached value only when the same options are set.
5570 32. Added John Jetmore's "exipick" utility to the distribution.
5572 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5573 than a timeout (e.g. a certificate is required, and is not provided), an
5574 Exim server now closes the connection immediately. Previously it waited for
5575 the client to close - but if the client is SSL, it seems that they each
5576 wait for each other, leading to a delay before one of them times out.
5578 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5579 maintained 0.8.x compatibility because I don't think many are using it, and
5580 it is clearly obsolete.
5582 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5585 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5586 client certificate was expired. A simple patch fixes this, though I don't
5587 understand the full logic of why the verify callback is called multiple
5590 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5591 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5592 which causes problems with some clients (such as the Certicom SSL Plus
5593 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5594 disables the coutermeasure allowing Eudora to connect."
5596 38. Exim was not checking that a write() to a log file succeeded. This could
5597 lead to Bad Things if a log got too big, in particular if it hit a file
5598 size limit. Exim now panics and dies if it cannot write to a log file, just
5599 as it does if it cannot open a log file.
5601 39. Modified OS/Makefile-Linux so that it now contains
5603 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5605 The two -D definitions ensure that Exim is compiled with large file
5606 support, which makes it possible to handle log files that are bigger than
5609 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5610 instance) a domain was checked against a named list that involved a lookup,
5611 causing $domain_data to be set, then another domain was checked against the
5612 same list, then the first domain was re-checked, the value of $domain_data
5613 after the final check could be wrong. In particular, if the second check
5614 failed, it could be set empty. This bug probably also applied to
5617 41. The strip_trailing_dot option was not being applied to the address given
5618 with the -f command-line option.
5620 42. The code for reading a message's header from the spool was incrementing
5621 $received_count, but never initializing it. This meant that the value was
5622 incorrect (doubled) while delivering a message in the same process in which
5623 it was received. In the most common configuration of Exim, this never
5624 happens - a fresh exec is done - but it can happen when
5625 deliver_drop_privilege is set.
5627 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5628 it now includes up to 150 characters of the unexpected data in the log
5631 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5632 and building data strings. The size of both of these buffers was 10 000
5633 bytes - far larger than anybody would *ever* want, thought I. Needless to
5634 say, somebody hit the limit. I have increased the maximum line length to
5635 20 000 and the maximum data length of concatenated lines to 100 000. I have
5636 also fixed two bugs, because there was no checking on these buffers. Tsk,
5637 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5638 buffer is too small.
5640 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5641 lsearch lookups. Now it does.
5643 46. When parsing a route_list item in a manualroute router, a fixed-length
5644 buffer was used for the list of hosts. I made this 1024 bytes long,
5645 thinking that nobody would ever have a list of hosts that long. Wrong.
5646 Somebody had a whole pile of complicated expansion conditions, and the
5647 string was silently truncated, leading to an expansion error. It turns out
5648 that it is easier to change to an unlimited length (owing to other changes
5649 that have happened since this code was originally written) than to build
5650 structure for giving a limitation error. The length of the item that
5651 expands into the list of hosts is now unlimited.
5653 47. The lsearch lookup could not handle data where the length of text line was
5654 more than 4095 characters. Such lines were truncated, leading to shortened
5655 data being returned. It should now handle lines of any length.
5657 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5658 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5661 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5662 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5663 no longer happens in comment lines. A list of replacements is now placed
5664 at the head of all of the source files, except those whose only change is
5665 to replace PERL_COMMAND in the very first #! line.
5667 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5668 messages on the queue, with a bottom-up merge sort, using code contributed
5669 by Michael Haardt. This should make operations like -bp somewhat faster on
5670 large queues. It won't affect queue runners, except when queue_run_in_order
5673 51. Installed eximstats 1.31 in the distribution.
5675 52. Added support for SRV lookups to the dnslookup router.
5677 53. If an ACL referred to $message_body or $message_body_end, the value was not
5678 reset for any messages that followed in the same SMTP session.
5680 54. The store-handling optimization for building very long strings was not
5681 differentiating between the different store pools. I don't think this
5682 actually made any difference in practice, but I've tidied it.
5684 55. While running the routers to verify a sender address, $sender_address
5685 was still set to the sender address. This is wrong, because when routing to
5686 send a bounce to the sender, it would be empty. Therefore, I have changed
5687 it so that, while verifying a sender address, $sender_address is set to <>.
5688 (There is no change to what happens when verifying a recipient address.)
5690 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5691 target A or AAAA records (if not already returned) without resetting the
5692 qualify_single or search_parents options of the DNS resolver. These are
5693 inappropriate in this case because the targets of MX and SRV records must
5694 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5695 target that, when qualified, matched something in the local domain. These
5696 two options are now turned off when doing these lookups.
5698 57. It seems that at least some releases of Reiserfs (which does not have the
5699 concept of a fixed number of inodes) returns zero and not -1 for the
5700 number of available inodes. This interacted badly with check_spool_inodes,
5701 which assumed that -1 was the "no such thing" setting. What I have done is
5702 to check that the total number of inodes is greater than zero before doing
5703 the test of how many are available.
5705 58. When a "warn" ACL statement has a log_message modifier, the message is
5706 remembered, and not repeated. This is to avoid a lot of repetition when a
5707 message has many recipients that cause the same warning to be written.
5708 However, Exim was preserving the list of already written lines for an
5709 entire SMTP session, which doesn't seem right. The memory is now reset if a
5710 new message is started.
5712 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5713 debugging output unless log_rewrite was also set.
5715 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5716 of (int)(handle) when we know that handle contains (void *)(-1).
5718 61. The Exim daemon panic-logs an error return when it closes the incoming
5719 connection. However "connection reset by peer" seems to be common, and
5720 isn't really an error worthy of noting specially, so that particular error
5723 62. When Exim is trying to find all the local interfaces, it used to panic and
5724 die if the ioctl to get the interface flags failed. However, it seems that
5725 on at least one OS (Solaris 9) it is possible to have an interface that is
5726 included in the list of interfaces, but for which you get a failure error
5727 for this call. This happens when the interface is not "plumbed" into a
5728 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5729 failure of the "get flags" call assumes that the interface is down.
5731 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5732 makes life easier for people who are doing arithmetic on fields extracted
5733 from dates, where you often get leading zeros that should not be
5734 interpreted as octal.
5736 64. Added qualify_domain to the redirect router, to override the global
5739 65. If a pathologically long header line contained very many addresses (the
5740 report of this problem mentioned 10 000) and each of them was rewritten,
5741 Exim could use up a very large amount of memory. (It kept on making new
5742 copies of the header line as it rewrote, and never released the old ones.)
5743 At the expense of a bit more processing, the header rewriting function has
5744 been changed so that it no longer eats memory in this way.
5746 66. The generation of the Received: header has been moved from the time that a
5747 message starts to be received, to the time that it finishes. The timestamp
5748 in the Received: header should now be very close to that of the <= log
5749 line. There are two side-effects of this change:
5751 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5752 logged header lines no longer include the local Received: line, because
5753 it has not yet been created. The same applies to a copy of the message
5754 that is returned to a non-SMTP sender when a message is rejected.
5756 (b) When a filter file is tested using -bf, no additional Received: header
5757 is added to the test message. After some thought, I decided that this
5760 This change does not affect the value of $received_for. It is still set
5761 after address rewriting, but before local_scan() is called.
5763 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5765 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5766 gave an unhelpful panic error message, and a defer error. I have managed to
5767 change this behaviour so that it now rejects any supplied certificate,
5768 which seems right, as the list of acceptable certificates is empty.
5770 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5771 gave an unhelpful defer error. I have not managed to make this reject any
5772 supplied certificates, but the error message it gives is "no certificate
5773 supplied", which is not helpful.
5775 70. exigrep's output now also includes lines that are not associated with any
5776 message, but which match the given pattern. Implemented by a patch from
5777 Martin Sluka, which also tidied up the Perl a bit.
5779 71. Recipient callout verification, like sender verification, was using <> in
5780 the MAIL FROM command. This isn't really the right thing, since the actual
5781 sender may affect whether the remote host accepts the recipient or not. I
5782 have changed it to use the actual sender in the callout; this means that
5783 the cache record is now keyed on a recipient/sender pair, not just the
5784 recipient address. There doesn't seem to be a real danger of callout loops,
5785 since a callout by the remote host to check the sender would use <>.
5786 [SEE ABOVE: changed after hitting problems.]
5788 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5789 temporary errors. However, in the case of such a code being given after
5790 the end of a data transmission (i.e. after ".") Exim was failing to write
5791 a retry record for the message. (Yes, there was some broken host that was
5792 actually sending 8xx at this point.)
5794 73. An unknown lookup type in a host list could cause Exim to panic-die when
5795 the list was checked. (An example that provoked this was putting <; in the
5796 middle of a list instead of at the start.) If this happened during a DATA
5797 ACL check, a -D file could be left lying around. This kind of configuration
5798 error no longer causes Exim to die; instead it causes a defer error. The
5799 incident is still logged to the main and panic logs.
5801 74. Buglet left over from Exim 3 conversion. The message "too many messages
5802 in one connection" was written to the rejectlog but not the mainlog, except
5803 when address rewriting (yes!) was being logged.
5805 75. Added write_rejectlog option.
5807 76. When a system filter was run not as root (that is, when system_filter_user
5808 was set), the values of the $n variables were not being returned to the
5809 main process; thus, they were not subsequently available in the $sn
5812 77. Added +return_path_on_delivery log selector.
5814 78. A connection timeout was being treated differently from recipients deferred
5815 when testing hosts_max_try with a message that was older than the host's
5816 retry timeout. (The host should not be counted, thus allowing all hosts to
5817 be tried at least once before bouncing.) This may have been the cause of an
5818 occasionally reported bug whereby a message would remain on the queue
5819 longer than the retry timeout, but would be bounced if a delivery was
5820 forced. I say "may" because I never totally pinned down the problem;
5821 setting up timeout/retry tests is difficult. See also the next item.
5823 79. The ultimate address timeout was not being applied to errors that involved
5824 a combination of host plus message (for example, a timeout on a MAIL
5825 command). When an address resolved to a number of possible hosts, and they
5826 were not all tried for each delivery (e.g. because of hosts_max_try), a
5827 message could remain on the queue longer than the retry timeout.
5829 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5832 81. Fixed an obscure SMTP outgoing bug which required at least the following
5833 conditions: (a) there was another message waiting for the same server;
5834 (b) the server returned 5xx to all RCPT commands in the first message so
5835 that the message was not completed; (c) the server dropped the connection
5836 or gave a negative response to the RSET that Exim sends to abort the
5837 transaction. The observed case was a dropped connection after DATA that had
5838 been sent in pipelining mode. That is, the server had advertised PIPELINING
5839 but was not implementing it correctly. The effect of the bug was incorrect
5840 behaviour, such as trying another host, and this could lead to a crash.
5846 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5847 and daemon.c were passed as pointers to ints; they should have been
5848 pointers to socklen_t variables (which are typically unsigned ints).
5850 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5853 3. Fixed a really odd bug that affected only the testing scheme; patching a
5854 certain fixed string in the binary changed the value of another string that
5855 happened to be identical to the end of the original first string.
5857 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5858 name", it returns that address as the IP address. On some operating
5859 systems (e.g. Solaris), it also passes back the IP address string as the
5860 "host name". However, on others (e.g. Linux), it passes back an empty
5861 string. Exim wasn't checking for this, and was changing the host name to an
5862 empty string, assuming it had been canonicalized.
5864 5. Although rare, it is permitted to have more than one PTR record for a given
5865 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5866 all the names associated with an address, because they do in Solaris.
5867 However, it seems that they do not in Linux for data that comes from the
5868 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5869 I found this out when I moved to a new Linux workstation and tried to run
5870 the Exim test suite.
5872 To get round this problem I have changed the code so that it now does its
5873 own call to the DNS to look up PTR records when searching for a host name.
5874 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5875 addresses that are only in /etc/hosts are still found.
5877 This behaviour is, however, controlled by an option called host_lookup_
5878 order, which defaults to "bydns:byaddr". If people want to use the other
5879 order, or indeed, just use one or the other means of lookup, they can
5880 specify it in this variable.
5882 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5883 some operating systems, this comes back from gethostbyaddr() as an empty
5884 string, and this is what Exim used to test for. However, it seems that in
5885 other systems, "." is yielded. Exim now tests for this case too.
5887 7. The values of check_spool_space and check_log_space are now held internally
5888 as a number of kilobytes instead of an absolute number of bytes. If a
5889 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5890 kilobyte. This means that much larger values can be stored.
5892 8. Exim monitor: an attempt to get the action menu when not actually pointing
5893 at a message produces an empty menu entitled "No message selected". This
5894 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5895 no entries in it ("Shell widget menu has zero width and/or height"). So I
5896 have added a single, blank menu entry in this case.
5898 9. Added ${quote_local_part.
5900 10. MIME decoding is now applied to the contents of Subject: header lines when
5903 11. Now that a reference to $sender_host_address automatically causes a reverse
5904 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5905 host lookup before query-style lookups in lists that might use this
5906 variable. This has therefore been abolished, and the "net-" prefix is no
5907 longer necessary for query-style lookups.
5909 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5910 have been a typo for LFLAGS, so it has been changed.
5912 13. The install script calls Exim with "-C /dev/null" in order to find the
5913 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5914 to be output. However, since Exim outputs its version number before the
5915 error, it didn't break the script. It just looked ugly. I fixed this by
5916 always allowing "-C /dev/null" if the caller is root.
5918 14. Ignore overlarge ACL variable number when reading spool file - insurance
5919 against a later release with more variables having written the file.
5921 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5922 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5923 this, as well as the form without the "IPv6" on the front.
5925 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5926 OS/Makefile-Darwin file.
5928 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5929 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5931 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5932 variable. However, some OS use ints for inodes. I've added cast to long int
5933 to get rid of the compiler warning.
5935 19. I had forgotten to lock out "/../" in configuration file names when
5936 ALT_CONFIG_PREFIX was set.
5938 20. Routers used for verification do not need to specify transports. However,
5939 if such a router generated a host list, and callout was configured, Exim
5940 crashed, because it could not find a port number from the (non-existent)
5941 transport. It now assumes port 25 in this circumstance.
5943 21. Added the -t option to exigrep.
5945 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5946 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5947 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5948 them is set, LOOKUP_LSEARCH is forced.
5950 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5952 24. Added sender and host information to the "rejected by local_scan()" log
5953 line; previously there was no indication of these.
5955 25. Added .include_if_exists.
5957 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5958 Linux. It turns out that not all file systems support this. Apparently some
5959 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5960 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5961 not supported on the file descriptor, is now ignored when Exim is trying to
5962 sync a directory. This applies only to Linux.
5964 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5966 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5967 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5970 29. Added support for maildirsize files from supplied patch (modified a bit).
5972 30. The use of :fail: followed by an empty string could lead Exim to respond to
5973 sender verification failures with (e.g.):
5975 550 Verification failed for <xxx>
5976 550 Sender verify failed
5978 where the first response line was missing the '-' that indicates it is not
5979 the final line of the response.
5981 31. The loop for finding the name of the user that called Exim had a hardwired
5982 limit of 10; it now uses the value of finduser_retries, which is used for
5983 all other user lookups.
5985 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5988 33. Exim was neglecting to zero errno before one call of strtol() when
5989 expanding a string and expecting an integer value. On some systems this
5990 resulted in spurious "integer overflow" errors. Also, it was casting the
5991 result into an int without checking.
5993 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5994 did not work. The code looks as if it has *never* worked, though it appears
5995 to have been documented since at least release 1.62. I have made it work.
5997 35. The "timeout_DNS" error in retry rules, also documented since at least
5998 1.62, also never worked. As it isn't clear exactly what this means, and
5999 clearly it isn't a major issue, I have abolished the feature by treating it
6000 as "timeout", and writing a warning to the main and panic logs.
6002 36. The display of retry rules for -brt wasn't always showing the error code
6005 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6006 timeout_connect_A, timeout_connect_MX.
6008 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6009 to the empty sender.
6011 39. The daemon was not analysing the content of -oX till after it had closed
6012 stderr and disconnected from the controlling terminal. This meant that any
6013 syntax errors were only noted on the panic log, and the return code from
6014 the command was 0. By re-arranging the code a little, I've made the
6015 decoding happen first, so such errors now appear on stderr, and the return
6016 code is 1. However, the actual setting up of the sockets still happens in
6017 the disconnected process, so errors there are still only recorded on the
6020 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6021 connections (as happens on some IP stacks) was logged at start up time as
6022 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6023 it from "IPv6 and IPv4", which means that two separate sockets are being
6026 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6027 says whether AF_INET or AF_INET6 was passed as an argument.
6029 42. Exiwhat output was messed up when time zones were included in log
6032 43. Exiwhat now gives more information about the daemon's listening ports,
6033 and whether -tls-on-connect was used.
6035 44. The "port" option of the smtp transport is now expanded.
6037 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6038 silently ignored. Now an error message is written to the main and panic
6041 46. There's a new ACL modifier called "logwrite" which writes to a log file
6042 as soon as it is encountered.
6044 47. Added $local_user_uid and $local_user_gid at routing time.
6046 48. Exim crashed when trying to verify a sender address that was being
6049 49. Exim was recognizing only a space character after ".include". It now also
6050 recognizes a tab character.
6052 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6053 extracting the relevant information from the specification. The man page no
6054 longer contains scrambled data for the -d option, and I've added a section
6055 at the front about calling Exim under different names.
6057 51. Added "extra_headers" argument to the "mail" command in filter files.
6059 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6062 53. Installed eximstats 1.29.
6064 54. Added transport_filter_timeout as a generic transport option.
6066 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6067 Cc: header lines. This was required by RFC 822, but it not required by RFC
6070 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6071 incoming messages that did not have them. Now it does so only if the
6072 message originates locally, that is, if there is no associated remote host
6073 address. When Resent- header lines are present, this applies to the Resent-
6074 lines rather than the non-Resent- lines.
6076 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6077 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6079 58. Messages for configuration errors now include the name of the main
6080 configuration file - useful now that there may be more than one file in a
6081 list (.included file names were always shown).
6083 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6084 for those rare installations that do not start the daemon as root or run it
6085 setuid root. I've cut out the call to initgroups() if the daemon is not
6088 60. The Exim user and group can now be bound into the binary as text strings
6089 that are looked up at the start of Exim's processing.
6091 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6093 62. Added $mailstore_basename variable.
6095 63. Installed patch to sieve.c from Michael Haardt.
6097 64. When Exim failed to open the panic log after failing to open the main log,
6098 the original message it was trying to log was written to stderr and debug
6099 output, but if they were not available (the usual case in production), it
6100 was lost. Now it is written to syslog before the two lines that record the
6101 failures to open the logs.
6103 65. Users' Exim filters run in subprocesses under the user's uid. It is
6104 possible for a "deliver" command or an alias in a "personal" command to
6105 provoke an address rewrite. If logging of address rewriting is configured,
6106 this fails because the process is not running as root or exim. There may be
6107 a better way of dealing with this, but for the moment (because 4.30 needs
6108 to be released), I have disabled address rewrite logging when running a
6109 filter in a non-root, non-exim process.
6115 1. The buildconfig auxiliary program wasn't quoting the value set for
6116 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6117 not defined. This bug was masked in 4.22 by the effect that was fixed in
6120 2. Some messages that were rejected after a message id was allocated were
6121 shown as "incomplete" by exigrep. It no longer does this for messages that
6122 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6124 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6125 have allow_domain_literals set, the ID did not get logged in the <= line.
6126 Domain literals are now always recognized in Message-ID: header lines.
6128 4. The first argument for a ${extract expansion item is the key name or field
6129 number. Leading and trailing spaces in this item were not being ignored,
6130 causing some misleading effects.
6132 5. When deliver_drop_privilege was set, single queue runner processes started
6133 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6134 same command in the process it spins off) were not dropping privilege.
6136 6. When the daemon running as "exim" started a queue runner, it always
6137 re-executed Exim in the spun-off process. This is a waste of effort when
6138 deliver_drop_privilege is set. The new process now just calls the
6139 queue-runner function directly.
6145 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6148 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6149 ignored. Though the use of -oP was forcing the writing of a pid file, it
6150 was always written to the default place.
6152 3. If the message "no IP address found for host xxxx" is generated during
6153 incoming verification, it is now followed by identification of the incoming
6154 connection (so you can more easily find what provoked it).
6156 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6158 5. Added some features to "harden" Exim a bit more against certain attacks:
6160 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6161 be put in Local/Makefile. This is like the never_users runtime option,
6162 but it cannot be overridden. The default setting is "root".
6164 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6165 prefix string with which any file named in a -C command line option
6168 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6169 is retained for -C and -D only if the caller of Exim is root. Without
6170 it, the exim user may also use -C and -D and retain privilege.
6172 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6173 command line option is disabled.
6175 6. Macro names set by the -D option must start with an upper case letter, just
6176 like macro names defined in the configuration file.
6178 7. Added "dereference=" facility to LDAP.
6180 8. Two instances of the typo "uknown" in the source files are fixed.
6182 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6183 the Configure-Makefile script screwed up while processing it.
6185 10. Incorporated PCRE 4.4.
6187 11. The SMTP synchronization check was not operating right at the start of an
6188 SMTP session. For example, it could not catch a HELO sent before the client
6189 waited for the greeting. There is now a check for outstanding input at the
6190 point when the greeting is written. Because of the duplex, asynchronous
6191 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6192 way, but not yet received, when the check is performed.
6194 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6195 on TCP/IP sockets, because this apparently causes some broken clients to
6198 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6199 unchanged) from the Cygwin maintainer.
6201 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6202 was supported instead of testing for mbx. Effectively a typo.
6204 15. The spa authenticator server code was not checking that the input it
6205 received was valid base64.
6207 16. The debug output line for the "set" modifier in ACLs was not showing the
6208 name of the variable that was being set.
6210 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6212 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6213 lookup on its own. Something else was needed to trigger the lookup. For
6214 example, a match in host_lookup or the need for a host name in a host list.
6215 Now, if $sender_host_name is referenced and the host name has not yet been
6216 looked up, a lookup is performed. If the lookup fails, the variable remains
6217 empty, and $host_lookup_failed is set to "1".
6219 19. Added "eqi" as a case-independent comparison operator.
6221 20. The saslauthd authentication condition could segfault if neither service
6222 nor realm was specified.
6224 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6225 error message that was logged was misleading, and incoming SMTP
6226 connections were dropped. The message is now more accurate, and temporary
6227 errors are given to SMTP connections.
6229 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6230 (except RSET) with a 421 error, until QUIT is received. However, it was
6231 failing to send a response to QUIT.
6233 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6234 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6235 ACL is now run after the helo_try_verify_hosts code.
6237 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6238 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6239 case-independent, so other case variants are also recognized). Apparently
6240 some systems use these upper case variants.
6242 25. If more than two messages were waiting for the same host, and a transport
6243 filter was specified for the transport, Exim sent two messages over the
6244 same TCP/IP connection, and then failed with "socket operation on non-
6245 socket" when it tried to send the third.
6247 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6249 27. The extern definition of crypt16() in expand.c was not being excluded when
6250 the OS had its own crypt16() function.
6252 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6253 as a preferred synonym for return_size_limit, both as an option and as an
6256 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6258 30. Changed the default configuration ACL to relax the local part checking rule
6259 for addresses that are not in any local domains. For these addresses,
6260 slashes and pipe symbols are allowed within local parts, but the sequence
6261 /../ is explicitly forbidden.
6263 31. SPA server authentication was not clearing the challenge buffer before
6266 32. log_message in a "warn" ACL statement was writing to the reject log as
6267 well as to the main log, which contradicts the documentation and doesn't
6268 seem right (because no rejection is happening). So I have stopped it.
6270 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6271 However, I am unable to do any testing of this.
6273 34. Fixed an infelicity in the appendfile transport. When checking directories
6274 for a mailbox, to see if any needed to be created, it was accidentally
6275 using path names with one or more superfluous leading slashes; tracing
6276 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6278 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6279 contents are added to the log line that is written for every discarded
6280 recipient. (Previously a log_message setting was ignored.)
6282 36. The ${quote: operator now quotes the string if it is empty.
6284 37. The install script runs exim in order to find its version number. If for
6285 some reason other than non-existence or emptiness, which it checks, it
6286 could not run './exim', it was installing it with an empty version number,
6287 i.e. as "exim-". This error state is now caught, and the installation is
6290 38. An argument was missing from the function that creates an error message
6291 when Exim fails to connect to the socket for saslauthd authentication.
6292 This could cause Exim to crash, or give a corrupted message.
6294 39. Added isip, isip4, and isip6 to ${if conditions.
6296 40. The ACL variables $acl_xx are now saved with the message, and can be
6297 accessed later in routers, transports, and filters.
6299 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6300 strings in the file are not string-expanded.
6302 42. If a MAIL command specified a SIZE value that was too large to fit into an
6303 int variable, the check against message_size_limit failed. Such values are
6304 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6305 day this will have to be increased, but I don't think I want to be around
6306 when emails are that large.
6313 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6314 iconv() is not standard in FreeBSD.
6316 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6317 IPv6 enabled. The observed symptom was a segmentation fault on return from
6318 the function os_common_find_running_interfaces() in src/os.c.
6320 3. In the check_special_case() function in daemon.c I had used "errno" as an
6321 argument name, which causes warnings on some systems. This was basically a
6322 typo, since it was named "eno" in the comments!
6324 4. The code that waits for the clock to tick (at a resolution of some fraction
6325 of a second) so as to ensure message-id uniqueness was always waiting for
6326 at least one whole tick, when it could have waited for less. [This is
6327 almost certainly not relevant at current processor speeds, where it is
6328 unlikely to ever wait at all. But we try to future-proof.]
6330 5. The function that sleeps for a time interval that includes fractions of a
6331 second contained a race. It did not block SIGALRM between setting the
6332 timer, and suspending (a couple of lines later). If the interval was short
6333 and the sigsuspend() was delayed until after it had expired, the suspension
6334 never ended. On busy systems this could lead to processes getting stuck for
6337 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6338 process, before it forks any delivery processes. The open lookup caching
6339 mechanism meant that the open file or database connection was passed into
6340 the delivery process. The problem was that delivery processes always tidy
6341 up cached lookup data. This could cause a problem for the next delivery
6342 process started by the queue runner, because the external queue runner
6343 process does not know about the closure. So the next delivery process
6344 still has data in the lookup cache. In the case of a file lookup, there was
6345 no problem because closing a file descriptor in a subprocess doesn't affect
6346 the parent. However, if the lookup was caching a connection to a database,
6347 the connection was closed, and the second delivery process was likely to
6348 see errors such as "PGSQL: query failed: server closed the connection
6349 unexpectedly". The problem has been fixed by closing all cached lookups
6350 in a queue runner before running a delivery process.
6352 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6353 seem to have the "const" qualifier which it has on other OS. I've
6356 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6357 *of the same type* (client or server) with the same public name that an
6358 error should be diagnosed.
6360 9. When Exim looked up a host name for an IP address, but failed to find the
6361 original IP address when looking up the host name (a safety check), it
6362 output the message "<ip address> does not match any IP for NULL", which was
6363 confusing, to say the least. The bug was that the host name should have
6364 appeared instead of "NULL".
6366 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6367 user that is built into the binary, and the -C or -D options is used, root
6368 privilege is dropped before the configuration file is read. In addition,
6369 logging is switched to stderr instead of the normal log files. If the
6370 configuration then re-defines the Exim user, the unprivileged environment
6371 is probably not what is expected, so Exim logs a panic warning message (but
6374 However, if deliver_drop_privilege is set, the unprivileged state may well
6375 be exactly what is intended, so the warning has been cut out in that case,
6376 and Exim is allowed to try to write to its normal log files.
6382 1. smtp_return_error_details was not giving details for temporary sender
6383 or receiver verification errors.
6385 2. Diagnose a configuration error if two authenticators have the same public
6388 3. Exim used not to create the message log file for a message until the first
6389 delivery attempt. This could be confusing when incoming messages were held
6390 for policy or load reasons. The message log file is now created at the time
6391 the message is received, and an initial "Received" line is written to it.
6393 4. The automatically generated man page for command line options had a minor
6394 bug that caused no ill effects; however, a more serious problem was that
6395 the procedure for building the man page automatically didn't always
6396 operate. Consequently, release 4.20 contains an out-of-date version. This
6397 shouldn't happen again.
6399 5. When building Exim with embedded Perl support, the script that builds the
6400 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6401 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6403 6. The freeze_tell option was not being used for messages that were frozen on
6404 arrival, either by an ACL or by local_scan().
6406 7. Added the smtp_incomplete_transaction log selector.
6408 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6409 was accepting AUTH without a new EHLO.
6411 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6412 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6415 10. Logging of TCP/IP connections (when configured) now happens in the main
6416 daemon process instead of the child process, so that the TCP/IP connection
6417 count is more accurate (but it can never be perfect).
6419 11. The use of "drop" in a nested ACL was not being handled correctly in the
6420 outer ACL. Now, if condition failure induced by the nested "drop" causes
6421 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6422 or "require"), the connection is dropped.
6424 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6425 that yield "discard" can now be used with an "accept" or a "discard" verb,
6426 but an error is generated for any others (because I can't see a useful way
6427 to define what should happen).
6429 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6430 are now processed in the same way as lines in the Exim configuration file.
6431 In particular, continuation lines are supported.
6433 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6435 15. Added -ti meaning -t -i.
6437 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6438 domains, and warn by logging if others are found.
6440 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6441 returned from GIFCONF when Exim is trying to find the list of interfaces on
6442 a host. The code in os.c has been modified to copy each ifreq to an aligned
6443 structure in all cases.
6445 Also, in some cases, the returned ifreq's were being copied to a 'struct
6446 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6447 means the last couple of bytes of an IPv6 address could be chopped if the
6448 ifreq contained only a normal sockaddr (14 bytes storage).
6450 18. Named domain lists were not supported in the hosts_treat_as_local option.
6451 An entry such as +xxxx was not recognized, and was treated as a literal
6454 19. Ensure that header lines added by a DATA ACL are included in the reject log
6455 if the ACL subsequently rejects the message.
6457 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6458 MD5 (which is deprecated).
6460 21. When testing a filter file using -bf, Exim was writing a message when it
6461 took the sender from a "From " line in the message, but it was not doing so
6462 when it took $return_path from a Return-Path: header line. It now does.
6464 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6465 with a valid header line field name (a series of printing characters
6466 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6468 23. Changed "disc" in the source to "disk" to conform to the documentation and
6469 the book and for uniformity.
6471 24. Ignore Sendmail's -Ooption=value command line item.
6473 25. When execve() failed while trying to run a command in a pipe transport,
6474 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6475 could be confused with a return value of 69 from the command itself. This
6476 has been changed to 127, the value the shell returns if it is asked to run
6477 a non-existent command. The wording for the related log line suggests a
6478 non-existent command as the problem.
6480 26. If received_header_text expands to an empty string, do not add a Received:
6481 header line to the message. (Well, it adds a token one on the spool, but
6482 marks it "old" so that it doesn't get used or transmitted.)
6484 27. Installed eximstats 1.28 (addition of -nt option).
6486 28. There was no check for failure on the call to getsockname() in the daemon
6487 code. This can fail if there is a shortage of resources on the system, with
6488 ENOMEM, for example. A temporary error is now given on failure.
6490 29. Contrary to the C standard, it seems that in some environments, the
6491 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6492 program. Exim now does this explicitly; it affects the formatting of
6493 timestamps using strftime().
6495 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6496 complaints. I've now initialized all the variables, to avoid this.
6498 32. Header lines added by a system filter were not being "seen" during
6499 transport-time rewrites.
6501 33. The info_callback() function passed to OpenSSL is set up with type void
6502 (*)(SSL *, int, int), as described somewhere. However, when calling the
6503 function (actually a macro) that sets it up, the type void(*)() is
6504 expected. I've put in a cast to prevent warnings from picky compilers.
6506 34. If a DNS black list lookup found a CNAME record, but there were no A
6507 records associated with the domain it pointed at, Exim crashed.
6509 35. If a DNS black list lookup returned more than one A record, Exim ignored
6510 all but the first. It now scans all returned addresses if a particular IP
6511 value is being sought. In this situation, the contents of the
6512 $dnslist_value variable are a list of all the addresses, separated by a
6515 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6516 to accept a host name and all its aliases if the forward lookup for any of
6517 them yielded the IP address of the incoming connection. Now it accepts only
6518 those names whose forward lookup yields the correct IP address. Any other
6519 names are discarded. This closes a loophole whereby a rogue DNS
6520 administrator could create reverse DNS records to break through a
6521 wildcarded host restriction in an ACL.
6523 37. If a user filter or a system filter that ran in a subprocess used any of
6524 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6525 the wrong values were passed to the pipe command ($thisaddress had the
6526 value of $0, $0 had the value of $1, etc). This bug was introduced by
6527 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6529 38. Improved the line breaking for long SMTP error messages from ACLs.
6530 Previously, if there was no break point between 40 and 75 characters, Exim
6531 left the rest of the message alone. Two changes have been made: (a) I've
6532 reduced the minimum length to 35 characters; (b) if it can't find a break
6533 point between 35 and 75 characters, it looks ahead and uses the first one
6534 that it finds. This may give the occasional overlong line, but at least the
6535 remaining text gets split now.
6537 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6538 file descriptors might be low, and that setting 1000 would always raise it.
6539 It turns out that in some environments, the limit is already over 1000 and
6540 that lowering it causes trouble. So now Exim takes care not to decrease it.
6542 40. When delivering a message, the value of $return_path is set to $sender_
6543 address at the start of routing (routers may change the value). By an
6544 oversight, this default was not being set up when an address was tested by
6545 -bt or -bv, which affected the outcome if any router or filter referred to
6548 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6549 log only when "message" or "log_message" are set. However, if one of the
6550 conditions was an address verification, or a call to a nested ACL, the
6551 messages generated by the underlying test were being passed through. This
6552 no longer happens. The underlying message is available in $acl_verify_
6553 message for both "message" and "log_message" expansions, so it can be
6554 passed through if needed.
6556 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6557 new expansion $bh_ to give the encoded byte string without charset
6558 translation. Translation happens only if iconv() is available; HAVE_ICONV
6559 indicates this at build time. HEADERS_CHARSET gives the charset to
6560 translate to; headers_charset can change it in the configuration, and
6561 "headers charset" can change it in an individual filter file.
6563 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6564 that creates RFC 2047 encoded "words" labels them as that charset instead
6565 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6566 expansion operator; (ii) when Exim creates a From: line for a local
6567 message; (iii) when a header line is rewritten to include a "phrase" part.
6569 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6570 buggy, causing it to skip the first lines of messages whose message ID
6571 ended in 'D'. This would not have bitten before Exim release 4.14, because
6572 message IDs were unlikely to end in 'D' before then. The effect was to have
6573 incorrect size information for certain domains.
6575 45. #include "config.h" was missing at the start of the crypt16.c module. This
6576 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6579 46. If there was a timeout during a "random" callout check, Exim treated it as
6580 a failure of the random address, and carried on sending RSET and the real
6581 address. If the delay was just some slowness somewhere, the response to the
6582 original RCPT would be taken as a response to RSET and so on, causing
6583 mayhem of various kinds.
6585 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6586 when I implemented it. It didn't allow for the fact that some option values
6587 may legitimately be negative (e.g. size_addition), and it didn't even do
6588 the right test for positive values.
6590 48. Domain names in DNS records are case-independent. Exim always looks them up
6591 in lower case. Some resolvers return domain names in exactly the case they
6592 appear in the zone file, that is, they may contain uppercase letters. Not
6593 all resolvers do this - some return always lower case. Exim was treating a
6594 change of case by a resolver as a change of domain, similar to a widening
6595 of a domain abbreviation. This triggered its re-routing code and so it was
6596 trying to route what was effectively the same domain again. This normally
6597 caused routing to fail (because the router wouldn't handle the domain
6598 twice). Now Exim checks for this case specially, and just changes the
6599 casing of the domain that it ultimately uses when it transmits the message
6602 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6605 50. If a filter generated a file delivery with a non-absolute name (possible if
6606 no home directory exists for the router), the forbid_file option was not
6609 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6610 the existing equality matching.
6612 52. Exim was using ints instead of ino_t variables in some places where it was
6613 dealing with inode numbers.
6615 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6616 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6617 called TMPDIR, and if it finds it is different, it changes its value.
6619 54. The smtp_printf() function is now made available to local_scan() so
6620 additional output lines can be written before returning. There is also an
6621 smtp_fflush() function to enable the detection of a dropped connection.
6622 The variables smtp_input and smtp_batched_input are exported to
6625 55. Changed the default runtime configuration: the message "Unknown user"
6626 has been removed from the ACL, and instead placed on the localuser router,
6627 using the cannot_route_message feature. This means that any verification
6628 failures that generate their own messages won't get overridden. Similarly,
6629 the "Unrouteable address" message that was in the ACL for unverifiable
6630 relay addresses has also been removed.
6632 56. Added hosts_avoid_esmtp to the smtp transport.
6634 57. The exicyclog script was not checking for the esoteric option
6635 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6636 will work only if exicyclog is run under the appropriate euid.
6638 58. Following a discussion on the list, the rules by which Exim recognises line
6639 endings on incoming messages have been changed. The -dropcr and drop_cr
6640 options are now no-ops, retained only for backwards compatibility. The
6641 following line terminators are recognized: LF CRLF CR. However, special
6642 processing applies to CR:
6644 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6645 nor a local message in the state where . is a terminator.
6647 (ii) If a bare CR is encountered in a header line, an extra space is added
6648 after the line terminator so as not to end the header. The reasoning
6649 behind this is that bare CRs in header lines are most likely either
6650 to be mistakes, or people trying to play silly games.
6652 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6653 was being incorrectly given as 18 bytes larger than it should have been.
6654 This is a VOB (very old bug).
6656 60. This may never have affected anything current, but just in case it has:
6657 When the local host is found other than at the start of a list of hosts,
6658 the local host, those with the same MX, and any that follow, are discarded.
6659 When the list in question was part of a longer list of hosts, the following
6660 hosts (not currently being processed) were also being discarded. This no
6661 longer happens. I'm not sure if this situation could ever has previously
6664 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6667 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6668 header. This is something that is recommended in a new Internet Draft, and
6669 is something that is documented as being done by Sendmail. There are two
6670 possible values. For messages generated by the autoreply transport, Exim
6673 Auto-Submitted: auto-replied
6675 whereas for all other generated messages (e.g. bounces) it adds
6677 Auto-Submitted: auto-generated
6679 63. The "personal" condition in filters now includes a test for the
6680 Auto-Submitted: header. If it contains the string "auto-" the message it
6681 not considered personal.
6683 64. Added rcpt_include_affixes as a generic transport option.
6685 65. Added queue_only_override (default true).
6687 66. Added the syslog_duplication option.
6689 67. If what should have been the first header line of a message consisted of
6690 a space followed by a colon, Exim was mis-interpreting it as a header line.
6691 It isn't of course - it is syntactically invalid and should therefore be
6692 treated as the start of the message body. The misbehaviour could have
6693 caused a number of strange effects, including loss of data in subsequent
6694 header lines, and spool format errors.
6696 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6697 client host had authenticated. This control can now be exercised by an ACL
6698 for more flexibility.
6700 69. By default, callouts do not happen when testing with -bh. There is now a
6701 variant, -bhc, which does actually run the callout code, including
6702 consulting and updating the callout cache.
6704 70. Added support for saslauthd authentication, courtesy of Alexander
6707 71. If statvfs() failed on the spool or log directories while checking their
6708 size for availability, Exim confusingly gave the error "space shortage".
6709 Furthermore, in debugging mode it crashed with a floating point exception.
6710 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6711 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6712 serious problem, Exim now writes to the main and panic logs when this
6713 happens, with details of the failure. It then refuses to accept the
6714 incoming message, giving the message "spool directory problem" or "log
6715 directory problem" with a 421 code for SMTP messages.
6717 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6718 0, 1, and 2 exist, because some OS complain for execs without them (see
6719 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6720 descriptors. However, the code omitted to check that the open succeeded,
6721 causing mysterious errors if for some reason the permissions on /dev/null
6722 got screwed. Now Exim writes a message to the main and panic logs, and
6723 bombs out if it can't open /dev/null.
6725 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6726 interact so that it is all more flexible. It is supposed to remain
6727 backwards compatible. Also added extra_local_interfaces.
6729 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6730 to bomb out with an assertion failure - to the client this appears as a
6731 connection drop. This problem occurs in the part of the code that was taken
6732 from the Samba project. Fortunately, the assertion is in a very simple
6733 function, so I have fixed this by reproducing the function inline in the
6734 one place where it is called, and arranging for authentication to fail
6735 instead of killing the process with assert().
6737 75. The SPA client code was not working when the server requested OEM rather
6738 than Unicode encoding.
6740 76. Added code to make require_files with a specific uid setting more usable in
6741 the case where statting the file as root fails - usually a non-root-mounted
6742 NFS file system. When this happens and the failure is EACCES, Exim now
6743 forks a subprocess and does the per-uid checking as the relevant uid.
6745 77. Added process_log_path.
6747 78. If log_file_path was not explicitly set, a setting of check_log_space or
6748 check_log_inodes was ignored.
6750 79. If a space check for the spool or log partitions fails, the incident is now
6751 logged. Of course, in the latter case the data may get lost...
6753 80. Added the %p formatting code to string_format() so that it can be used to
6754 print addresses in debug_print(). Adjusted all the address printing in the
6755 debugging in store.c to use %p rather than %d.
6757 81. There was a concern that a line of code in smtp_in.c could overflow a
6758 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6759 initially expressed, the concern was not well-founded, because trailing
6760 spaces are removed early. However, if the trailing spaces were followed by
6761 a NULL, they did not get removed, so the overflow was possible. Two fixes
6764 (a) I re-wrote the offending code in a cleaner fashion.
6765 (b) If an incoming SMTP command contains a NULL character, it is rejected
6768 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6769 runs initgroups(), so that if the Exim user is in any additional groups,
6770 they will be used during message reception.
6776 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob