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 Bug 1922: Support IDNA2008. This has slightly different conversion rules
12 than -2003 did; needs libidn2 in addition to libidn.
14 JH/02 The path option on a pipe transport is now expanded before use.
16 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
17 Patch provided by "Björn", documentation fix added too.
19 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
20 missing a wire-to-host endian conversion.
22 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
23 close after a BDAT command line could be taken as a following command,
24 giving a synch failure. Fix by only checking for synch immediately
25 before acknowledging the chunk.
27 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
28 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
29 macro. Patches provided by Josh Soref.
31 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
32 Previously we did not; the RFC seems ambiguous and VRFY is not listed
33 by IANA as a service extension. However, John Klensin suggests that we
36 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
37 the dkim code may be unix-mode line endings rather than smtp wire-format
38 CRLF, so prepend a CR to any bare LF.
40 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
41 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
43 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
44 the first were themselves being wrongly included in the feed into dkim
45 processing; with most chunk sizes in use this resulted in an incorrect
46 body hash calculated value.
48 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
49 DKIM signature block, for verification. Although advised against by
50 standards it is specifically not ruled illegal.
52 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
54 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
55 missing a body hash (the bh= tag).
57 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
58 It seems that HAProxy sends the Proxy Protocol information in clear and
59 only then does a TLS startup, so do the same.
61 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
62 TCP connections (such as for Spamd) unless the daemon successfully set
63 Fast Open mode on its listening sockets. This fixes breakage seen on
64 too-old kernels or those not configured for Fast Open, at the cost of
65 requiring both directions being enabled for TFO, and TFO never being used
66 by non-daemon-related Exim processes.
68 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
69 endings, at least on the first header line. Try to canonify any that get
70 past that check, despite the cost.
72 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
73 now limited to an arbitrary five deep, while parsing addresses with the
74 strip_excess_angle_brackets option enabled.
76 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
77 instead leave the unprompted TLS handshake in socket buffer for the
78 TLS library to consume.
80 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
82 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
84 JH/16 Drop variables when they go out of scope. Memory management drops a whole
85 region in one operation, for speed, and this leaves assigned pointers
86 dangling. Add checks run only under the testsuite which checks all
87 variables at a store-reset and panics on a dangling pointer; add code
88 explicitly nulling out all the variables discovered. Fixes one known
89 bug: a transport crash, where a dangling pointer for $sending_ip_address
90 originally assigned in a verify callout, is re-used.
92 PP/06 Drop '.' from @INC in various Perl scripts.
94 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
96 PP/08 Reduce a number of compilation warnings under clang; building with
97 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
98 should be warning-free.
100 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
102 HS/01 Fix portability problems introduced by PP/08 for platforms where
103 realloc(NULL) is not equivalent to malloc() [SunOS et al].
105 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
106 chunk. This allows us to accept broken chunked messages. We need a more
107 general solution here.
113 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
114 supports it and a size is available (ie. the sending peer gave us one).
116 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
117 years of being deprecated). The replacements are the ACLs
118 acl_smtp_mime and acl_not_smtp_mime.
120 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
121 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
122 or even an in-clear connection were permitted. Now, if the host lookup
123 was dnssec and dane was requested then the host is only used if the
124 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
125 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
126 if one fails this test.
127 This means that a poorly-configured remote DNS will make it incommunicado;
128 but it protects against a DNS-interception attack on it.
130 JH/04 Bug 1810: make continued-use of an open smtp transport connection
131 non-noisy when a race steals the message being considered.
133 JH/05 If main configuration option tls_certificate is unset, generate a
134 self-signed certificate for inbound TLS connections.
136 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
137 in rewrites and routers.
139 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
140 and logged a warning sing 4.83; now they are a configuration file error.
142 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
143 (lacking @domain). Apply the same qualification processing as RCPT.
145 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
147 JH/10 Support ${sha256:} applied to a string (as well as the previous
150 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
151 a cutthrough deliver is pending, as we always want to make a connection.
152 This also avoids re-routing the message when later placing the cutthrough
153 connection after a verify cache hit.
154 Do not update it with the verify result either.
156 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
157 when routing results in more than one destination address.
159 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
160 signing (which inhibits the cutthrough capability). Previously only
161 the presence of an option was tested; now an expansion evaluating as
162 empty is permissible (obviously it should depend only on data available
163 when the cutthrough connection is made).
165 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
166 the relevant preceding SMTP command did not note the pipelining mode.
168 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
169 Previously they were not counted.
171 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
172 as one having no matching records. Previously we deferred the message
173 that needed the lookup.
175 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
176 distinguished as "(=".
178 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
179 for missing MX records. Previously it only worked for missing A records.
181 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
183 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
184 after the data-go-ahead and data-ack. Patch from Jason Betts.
186 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
187 even for a "none" policy. Patch from Tony Meyer.
189 JH/22 Fix continued use of a connection for further deliveries. If a port was
190 specified by a router, it must also match for the delivery to be
193 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
194 When one of the recipients of a message was unsuitable for the connection
195 (has no matching addresses), we lost track of needing to mark it
196 deferred. As a result mail would be lost.
198 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
200 JH/25 Decoding ACL controls is now done using a binary search; the source code
201 takes up less space and should be simpler to maintain. Merge the ACL
202 condition decode tables also, with similar effect.
204 JH/26 Fix problem with one_time used on a redirect router which returned the
205 parent address unchanged. A retry would see the parent address marked as
206 delivered, so not attempt the (identical) child. As a result mail would
209 JH/27 Fix a possible security hole, wherein a process operating with the Exim
210 UID can gain a root shell. Credit to http://www.halfdog.net/ for
211 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
214 JH/28 Enable {spool,log} filesystem space and inode checks as default.
215 Main config options check_{log,spool}_{inodes,space} are now
216 100 inodes, 10MB unless set otherwise in the configuration.
218 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
219 Previously it only applied to the main-section connection policy
222 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
224 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
225 by me. Added RFC7919 DH primes as an alternative.
227 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
228 are not in the system include path.
230 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
231 GnuTLS, when a session startup failed (eg because the client disconnected)
232 Exim did stdio operations after fclose. This was exposed by a recent
233 change which nulled out the file handle after the fclose.
235 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
236 signed directly by the cert-signing cert, rather than an intermediate
237 OCSP-signing cert. This is the model used by LetsEncrypt.
239 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
241 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
242 an incoming connection.
244 HS/02 Bug 1802: Do not half-close the connection after sending a request
247 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
248 fallback to "prime256v1".
250 JH/34 SECURITY: Use proper copy of DATA command in error message.
251 Could leak key material. Remotely exploitable. CVE-2016-9963.
257 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
258 and 3.4.4 - once the server is enabled to respond to an OCSP request
259 it does even when not requested, resulting in a stapling non-aware
260 client dropping the TLS connection.
262 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
263 support variable-length bit vectors. No functional change.
265 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
266 The I= interface field on outgoing lines is now after the H= remote
267 host field, same as incoming lines. There is a separate
268 outgoing_interface log selector which allows you to disable the
271 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
272 If not running log_selector +smtp_connection the mainlog would be held
273 open indefinitely after a "too many connections" event, including to a
274 deleted file after a log rotate. Leave the per net connection logging
275 leaving it open for efficiency as that will be quickly detected by the
276 check on the next write.
278 HS/01 Bug 1671: Fix post transport crash.
279 Processing the wait-<transport> messages could crash the delivery
280 process if the message IDs didn't exist for some reason. When
281 using 'split_spool_directory=yes' the construction of the spool
282 file name failed already, exposing the same netto behaviour.
284 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
285 mime_regex ACL conditions.
287 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
288 to DSN fail messages (bounces): remote IP, remote greeting, remote response
289 to HELO, local diagnostic string.
291 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
292 log line to debug. Even when configured with a tls authenticator many
293 client connections are expected to not authenticate in this way, so
294 an authenticate fail is not an error.
296 HS/02 Add the Exim version string to the process info. This way exiwhat
297 gives some more detail about the running daemon.
299 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
300 matter for fast-change records such as DNSBLs.
302 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
303 retry record, even if constant. There may be multiple transports with
304 different interface settings and the retry behaviour needs to be kept
307 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
309 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
311 JH/10 Bug 840: fix log_defer_output option of pipe transport
313 JH/11 Bug 830: use same host for all RCPTS of a message, even under
314 hosts_randomize. This matters a lot when combined with mua_wrapper.
316 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
317 ${quote_pgsql:<string>} operator.
319 JH/13 Bug 1708: avoid misaligned access in cached lookup.
321 JH/14 Change header file name for freeradius-client. Relevant if compiling
322 with Radius support; from the Gentoo tree and checked under Fedora.
324 JH/15 Bug 1712: Introduce $prdr_requested flag variable
326 JH/16 Bug 1714: Permit an empty string as expansion result for transport
327 option transport_filter, meaning no filtering.
329 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
331 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
332 defaults to "*" (all hosts). The variable is now available when not built
333 with TLS, default unset, mainly to enable keeping the testsuite sane.
334 If a server certificate is not supplied (via tls_certificate) an error is
335 logged, and clients will find TLS connections fail on startup. Presumably
336 they will retry in-clear.
337 Packagers of Exim are strongly encouraged to create a server certificate
338 at installation time.
340 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
341 with the $config_file variable.
343 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
344 in transport context, after the attempt, and per-recipient. The latter type
345 is per host attempted. The event data is the error message, and the errno
346 information encodes the lookup type (A vs. MX) used for the (first) host,
347 and the trailing two digits of the smtp 4xx response.
349 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
350 to write to mainlog (or rejectlog, paniclog) in the window between file
351 creation and permissions/ownership being changed. Particularly affects
352 installations where exicyclog is run as root, rather than exim user;
353 result is that the running daemon panics and dies.
355 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
357 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
358 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
359 "pri" and "weight". Note that the previous implicit priority given by the
360 list order is no longer honoured.
362 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
365 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
366 by defining SUPPORT_SOCKS.
368 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
369 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
370 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
371 variables are renamed to proxy_{local,external}_{address,port}.
373 JH/27 Move Internationalisation support from Experimental to mainline, enabled
374 for a build by defining SUPPORT_I18N
376 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
377 of the query string, and make ${quote_redis:} do that quoting.
379 JH/29 Move Events support from Experimental to mainline, enabled by default
380 and removable for a build by defining DISABLE_EVENT.
382 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
384 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
385 cached by the daemon.
387 JH/32 Move Redis support from Experimental to mainline, enabled for a build
388 by defining LOOKUP_REDIS. The libhiredis library is required.
390 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
391 keys are given for lookup.
393 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
394 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
395 only supported when built with TLS support. The PolarSSL SHA routines
396 are still used when the TLS library is too old for convenient support.
398 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
399 openssl_options), for security. OpenSSL forces this from version 1.1.0
400 server-side so match that on older versions.
402 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
403 allocation for $value could be released as the expansion processing
404 concluded, but leaving the global pointer active for it.
406 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
407 and to use the domains and local_parts ACL conditions.
409 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
410 incorrectly not doubled on cutthrough transmission, hence seen as a
411 body-termination at the receiving system - resulting in truncated mails.
412 Commonly the sender saw a TCP-level error, and retransmitted the message
413 via the normal store-and-forward channel. This could result in duplicates
414 received - but deduplicating mailstores were liable to retain only the
415 initial truncated version.
417 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
419 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
421 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
422 we're in there, support oversigning also; bug 1309.
424 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
426 HS/04 Add support for keep_environment and add_environment options.
428 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
429 either intentional arithmetic overflow during PRNG, or testing config-
432 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
433 delivery resulted in actual delivery. Cancel cutthrough before DATA
436 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
437 rejects a recipient: pass the reject to the originator.
439 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
440 Many were false-positives and ignorable, but it's worth fixing the
443 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
444 for the new environment-manipulation done at startup. Move the routines
445 from being local to tls.c to being global via the os.c file.
447 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
448 an extract embedded as result-arg for a map, the first arg for extract
449 is unavailable so we cannot tell if this is a numbered or keyed
450 extraction. Accept either.
456 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
459 JH/02 The smtp transport option "multi_domain" is now expanded.
461 JH/03 The smtp transport now requests PRDR by default, if the server offers
464 JH/04 Certificate name checking on server certificates, when exim is a client,
465 is now done by default. The transport option tls_verify_cert_hostnames
466 can be used to disable this per-host. The build option
467 EXPERIMENTAL_CERTNAMES is withdrawn.
469 JH/05 The value of the tls_verify_certificates smtp transport and main options
470 default to the word "system" to access the system default CA bundle.
471 For GnuTLS, only version 3.0.20 or later.
473 JH/06 Verification of the server certificate for a TLS connection is now tried
474 (but not required) by default. The verification status is now logged by
475 default, for both outbound TLS and client-certificate supplying inbound
478 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
481 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
482 Status Notification (bounce) messages are now MIME format per RFC 3464.
483 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
484 under the control of the dsn_advertise_hosts option, and routers may
485 have a dsn_lasthop option.
487 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
488 default, modifiable by a malware= option. The list separator for
489 the options can now be changed in the usual way. Bug 68.
491 JH/10 The smtp_receive_timeout main option is now expanded before use.
493 JH/11 The incoming_interface log option now also enables logging of the
494 local interface on delivery outgoing connections.
496 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
497 if the interface and destination host and port all match.
499 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
502 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
503 Patch from Andrew Lewis.
505 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
506 now supports optional time-restrictions, weighting, and priority
507 modifiers per server. Patch originally by <rommer@active.by>.
509 JH/16 The spamd_address main option now supports a mixed list of local
510 and remote servers. Remote servers can be IPv6 addresses, and
511 specify a port-range.
513 JH/17 Bug 68: The spamd_address main option now supports an optional
514 timeout value per server.
516 JH/18 Bug 1581: Router and transport options headers_add/remove can
517 now have the list separator specified.
519 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
522 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
525 JH/21 Support for the A6 type of dns record is withdrawn.
527 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
528 rather than the verbs used.
530 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
531 from 255 to 1024 chars.
533 JH/24 Verification callouts now attempt to use TLS by default.
535 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
536 are generic router options now. The defaults didn't change.
538 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
539 Original patch from Alexander Shikoff, worked over by JH.
541 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
542 files if ZCAT_COMMAND is not executable.
544 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
546 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
548 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
549 Normally benign, it bites when the pair was led to by a CNAME;
550 modern usage is to not canonicalize the domain to a CNAME target
551 (and we were inconsistent anyway for A-only vs AAAA+A).
553 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
555 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
556 when evaluating $sender_host_dnssec.
558 JH/31 Check the HELO verification lookup for DNSSEC, adding new
559 $sender_helo_dnssec variable.
561 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
563 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
565 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
567 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
568 documented as working, but never had. Support all but $spam_report.
570 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
571 added for tls authenticator.
573 HS/03 Add perl_taintmode main config option
579 TL/01 When running the test suite, the README says that variables such as
580 no_msglog_check are global and can be placed anywhere in a specific
581 test's script, however it was observed that placement needed to be near
582 the beginning for it to behave that way. Changed the runtest perl
583 script to read through the entire script once to detect and set these
584 variables, reset to the beginning of the script, and then run through
585 the script parsing/test process like normal.
587 TL/02 The BSD's have an arc4random API. One of the functions to induce
588 adding randomness was arc4random_stir(), but it has been removed in
589 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
590 function when detected.
592 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
593 cause callback expansion.
595 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
596 syntax errors in an expansion can be treated as a string instead of
597 logging or causing an error, due to the internal use of bool_lax
598 instead of bool when processing it.
600 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
601 server certificates when making smtp deliveries.
603 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
605 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
607 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
609 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
610 Merged patch from Sebastian Wiedenroth.
612 JH/05 Fix results-pipe from transport process. Several recipients, combined
613 with certificate use, exposed issues where response data items split
614 over buffer boundaries were not parsed properly. This eventually
615 resulted in duplicates being sent. This issue only became common enough
616 to notice due to the introduction of connection certificate information,
617 the item size being so much larger. Found and fixed by Wolfgang Breyha.
619 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
620 size buffer was used, resulting in syntax errors when an expansion
623 JH/07 Add support for directories of certificates when compiled with a GnuTLS
624 version 3.3.6 or later.
626 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
627 is EXPERIMENTAL_EVENT, the main-configuration and transport options
628 both become "event_action", the variables become $event_name, $event_data
629 and $event_defer_errno. There is a new variable $verify_mode, usable in
630 routers, transports and related events. The tls:cert event is now also
631 raised for inbound connections, if the main configuration event_action
634 TL/06 In test suite, disable OCSP for old versions of openssl which contained
635 early OCSP support, but no stapling (appears to be less than 1.0.0).
637 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
638 server certificate names available under the smtp transport option
639 "tls_verify_cert_hostname" now do not permit multi-component wildcard
642 JH/10 Time-related extraction expansions from certificates now use the main
643 option "timezone" setting for output formatting, and are consistent
644 between OpenSSL and GnuTLS compilations. Bug 1541.
646 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
647 encoded parameter in the incoming message. Bug 1558.
649 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
650 include certificate info, eximon was claiming there were spoolfile
653 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
655 JH/14 Log delivery-related information more consistently, using the sequence
656 "H=<name> [<ip>]" wherever possible.
658 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
659 are problematic for Debian distribution, omit them from the release
662 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
664 JH/16 Fix string representation of time values on 64bit time_t architectures.
667 JH/17 Fix a null-indirection in certextract expansions when a nondefault
668 output list separator was used.
673 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
674 checkers that were complaining about end of non-void function with no
677 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
678 This was a regression introduced in 4.83 by another bugfix.
680 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
682 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
683 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
689 TF/01 Correctly close the server side of TLS when forking for delivery.
691 When a message was received over SMTP with TLS, Exim failed to clear up
692 the incoming connection properly after forking off the child process to
693 deliver the message. In some situations the subsequent outgoing
694 delivery connection happened to have the same fd number as the incoming
695 connection previously had. Exim would try to use TLS and fail, logging
696 a "Bad file descriptor" error.
698 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
699 utilities have not been installed.
701 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
702 temporary space as the ACL may create new global variables.
704 TL/01 LDAP support uses per connection or global context settings, depending
705 upon the detected version of the libraries at build time.
707 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
708 to extract and use the src ip:port in logging and expansions as if it
709 were a direct connection from the outside internet. PPv2 support was
710 updated based on HAProxy spec change in May 2014.
712 JH/02 Add ${listextract {number}{list}{success}{fail}}.
714 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
715 Properly escape header and check for NULL return.
717 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
720 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
722 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
723 characters in header names, implemented as a verify condition.
724 Contributed by Michael Fischer v. Mollard.
726 TL/05 Rename SPF condition results err_perm and err_temp to standardized
727 results permerror and temperror. Previous values are deprecated but
728 still accepted. In a future release, err_perm and err_temp will be
729 completely removed, which will be a backward incompatibility if the
730 ACL tests for either of these two old results. Patch contributed by
731 user bes-internal on the mailing list.
733 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
735 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
736 selectors, in both main and reject logs.
738 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
741 JH/07 Add malware type "sock" for talking to simple daemon.
743 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
745 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
746 routers/transports under cutthrough routing.
748 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
749 numbers. Touch up "bool" conditional to keep the same definition.
751 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
753 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
755 JH/12 Expand items in router/transport headers_add or headers_remove lists
756 individually rather than the list as a whole. Bug 1452.
758 Required for reasonable handling of multiple headers_ options when
759 they may be empty; requires that headers_remove items with embedded
760 colons must have them doubled (or the list-separator changed).
762 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
763 view the policy declared in the DMARC record. Currently, $dmarc_status
764 is a combined value of both the record presence and the result of the
767 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
769 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
770 dnslookup router and the smtp transport (applying to the forward
773 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
774 of ldap servers used for a specific lookup. Patch provided by Heiko
777 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
778 New variable $lookup_dnssec_authenticated for observability.
780 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
781 Patch submitted by Lars Timman.
783 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
785 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
786 Requires trusted mode and valid format message id, aborts otherwise.
787 Patch contributed by Heiko Schlichting.
789 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
790 certextract with support for various fields. Bug 1358.
792 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
793 is requested by default, modifiable by smtp transport option
796 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
797 operate on certificate variables to give certificate fingerprints
798 Also new ${sha256:cert_variable}.
800 JH/23 The PRDR feature is moved from being Experimental into the mainline.
802 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
803 Christian Aistleitner.
805 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
807 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
808 file. Patch from Wolfgang Breyha.
810 JH/25 Expand the coverage of the delivery $host and $host_address to
811 client authenticators run in verify callout. Bug 1476.
813 JH/26 Port service names are now accepted for tls_on_connect_ports, to
814 align with daemon_smtp_ports. Bug 72.
816 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
817 support and error reporting did not work properly.
819 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
820 and is readable. Patch from Andrew Colin Kissa.
822 TL/14 Enhance documentation of ${run expansion and how it parses the
823 commandline after expansion, particularly in the case when an
824 unquoted variable expansion results in an empty value.
826 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
828 PP/02 Fix internal collision of T_APL on systems which support RFC3123
829 by renaming away from it. Addresses GH issue 15, reported by
832 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
834 TL/15 SECURITY: prevent double expansion in math comparison functions
835 (can expand unsanitized data). Not remotely exploitable.
842 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
844 PP/02 Make -n do something, by making it not do something.
845 When combined with -bP, the name of an option is not output.
847 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
850 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
851 $sender_host_name and config options to manage this, and basic check
854 PP/05 DSCP support for outbound connections and control modifier for inbound.
856 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
857 (Only plugin which currently uses this is kerberos4, which nobody should
858 be using, but we should make it available and other future plugins might
859 conceivably use it, even though it would break NAT; stuff *should* be
860 using channel bindings instead).
862 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
863 name; added for Sendmail compatibility; requires admin caller.
864 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
865 just ignore it); requires trusted caller.
866 Also parse but ignore: -Ac -Am -X<logfile>
869 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
871 TL/02 Add +smtp_confirmation as a default logging option.
873 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
874 Patch by Magnus Holmgren from 2007-02-20.
876 TL/04 Bugzilla 1281 - Spec typo.
877 Bugzilla 1283 - Spec typo.
878 Bugzilla 1290 - Spec grammar fixes.
880 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
882 TL/06 Add Experimental DMARC support using libopendmarc libraries.
884 TL/07 Fix an out of order global option causing a segfault. Reported to dev
885 mailing list by by Dmitry Isaikin.
887 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
889 JH/02 Support "G" suffix to numbers in ${if comparisons.
891 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
893 NM/01 Bugzilla 1197 - Spec typo
894 Bugzilla 1196 - Spec examples corrections
896 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
898 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
899 gnutls_enable_pkcs11, but renamed to more accurately indicate its
902 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
903 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
905 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
906 "acl {{name}{arg}...}", and optional args on acl condition
909 JH/05 Permit multiple router/transport headers_add/remove lines.
911 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
913 JH/07 Avoid using a waiting database for a single-message-only transport.
914 Performance patch from Paul Fisher. Bugzilla 1262.
916 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
919 JH/09 Add $headers_added variable, with content from use of ACL modifier
920 add_header (but not yet added to the message). Bugzilla 199.
922 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
923 Pulled from Bugzilla 817 by Wolfgang Breyha.
925 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
927 (nb: this is the same fix as in Exim 4.80.1)
929 JH/11 Add A= logging on delivery lines, and a client_set_id option on
932 JH/12 Add optional authenticated_sender logging to A= and a log_selector
935 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
937 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
938 advertise SMTP AUTH mechanism to us, instead of a generic
939 protocol violation error. Also, make Exim more robust to bad
940 data from the Dovecot auth socket.
942 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
944 When a queue runner is handling a message, Exim first routes the
945 recipient addresses, during which it prunes them based on the retry
946 hints database. After that it attempts to deliver the message to
947 any remaining recipients. It then updates the hints database using
950 So if a recipient address works intermittently, it can get repeatedly
951 deferred at routing time. The retry hints record remains fresh so the
952 address never reaches the final cutoff time.
954 This is a fairly common occurrence when a user is bumping up against
955 their storage quota. Exim had some logic in its local delivery code
956 to deal with this. However it did not apply to per-recipient defers
957 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
959 This change adds a proper retry rule check during routing so that the
960 final cutoff time is checked against the message's age. We only do
961 this check if there is an address retry record and there is not a
962 domain retry record; this implies that previous attempts to handle
963 the address had the retry_use_local_parts option turned on. We use
964 this as an approximation for the destination being like a local
965 delivery, as in LMTP.
967 I suspect this new check makes the old local delivery cutoff check
968 redundant, but I have not verified this so I left the code in place.
970 TF/02 Correct gecos expansion when From: is a prefix of the username.
972 Test 0254 submits a message to Exim with the header
976 When I ran the test suite under the user fanf2, Exim expanded
977 the header to contain my full name, whereas it should have added
978 a Resent-Sender: header. It erroneously treats any prefix of the
979 username as equal to the username.
981 This change corrects that bug.
983 GF/01 DCC debug and logging tidyup
984 Error conditions log to paniclog rather than rejectlog.
985 Debug lines prefixed by "DCC: " to remove any ambiguity.
987 TF/03 Avoid unnecessary rebuilds of lookup-related code.
989 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
990 Bug spotted by Jeremy Harris; was flawed since initial commit.
991 Would have resulted in OCSP responses post-SNI triggering an Exim
992 NULL dereference and crash.
994 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
996 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
997 Bug detection, analysis and fix by Samuel Thibault.
998 Bugzilla 1331, Debian bug #698092.
1000 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1002 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1003 Server implementation by Todd Lyons, client by JH.
1004 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1005 config variable "prdr_enable" controls whether the server
1006 advertises the facility. If the client requests PRDR a new
1007 acl_data_smtp_prdr ACL is called once for each recipient, after
1008 the body content is received and before the acl_smtp_data ACL.
1009 The client is controlled by both of: a hosts_try_prdr option
1010 on the smtp transport, and the server advertisement.
1011 Default client logging of deliveries and rejections involving
1012 PRDR are flagged with the string "PRDR".
1014 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1015 fclose(). Diagnosis by Todd Lyons.
1017 PP/17 Update configure.default to handle IPv6 localhost better.
1018 Patch by Alain Williams (plus minor tweaks).
1021 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1022 This is now consistent with GnuTLS, and is now documented: the
1023 previous undocumented portable approach to treating the option as
1024 unset was to force an expansion failure. That still works, and
1025 an empty string is now equivalent.
1027 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1028 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1029 not performing validation itself.
1031 PP/20 Added force_command boolean option to pipe transport.
1032 Patch from Nick Koston, of cPanel Inc.
1034 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1037 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1039 PP/21 Fix eximon continuous updating with timestamped log-files.
1040 Broken in a format-string cleanup in 4.80, missed when I repaired the
1041 other false fix of the same issue.
1042 Report and fix from Heiko Schlichting.
1045 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1046 Report from Prashanth Katuri.
1048 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1049 It's SecureTransport, so affects any MacOS clients which use the
1050 system-integrated TLS libraries, including email clients.
1052 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1053 using a MIME ACL for non-SMTP local injection.
1054 Report and assistance in diagnosis by Warren Baker.
1056 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1058 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1060 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1061 last id that failed so it may be referenced in subsequent ACL's.
1063 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1066 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1067 ldap library initialization, allowing self-signed CA's to be
1068 used. Also properly sets require_cert option later in code by
1069 using NULL (global ldap config) instead of ldap handle (per
1070 session). Bug diagnosis and testing by alxgomz.
1072 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1073 the src/util/ subdirectory.
1075 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1076 renamed to Transport Post Delivery Action by Jeremy Harris, as
1079 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1080 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1081 redis_servers = needs to be configured which will be used by the redis
1082 lookup. Patch from Warren Baker, of The Packet Hub.
1084 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1086 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1087 hostname or reverse DNS when processing a host list. Used suggestions
1088 from multiple comments on this bug.
1090 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1092 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1093 Missed a few lines, added it to make the runtest require no keyboard
1096 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1097 contains upper case chars. Make router use caseful_local_part.
1099 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1100 support when GnuTLS has been built with p11-kit.
1106 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1108 This, or similar/improved, will also be change PP/11 of 4.82.
1114 PP/01 Handle short writes when writing local log-files.
1115 In practice, only affects FreeBSD (8 onwards).
1116 Bugzilla 1053, with thanks to Dmitry Isaikin.
1118 NM/01 Bugzilla 949 - Documentation tweak
1120 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1123 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1125 PP/02 Implemented gsasl authenticator.
1127 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1129 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1130 `pkg-config foo` for cflags/libs.
1132 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1133 with rest of GSASL and with heimdal_gssapi.
1135 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1136 `pkg-config foo` for cflags/libs for the TLS implementation.
1138 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1139 properties get this fed in as external SSF. A number of robustness
1140 and debugging improvements to the cyrus_sasl authenticator.
1142 PP/08 cyrus_sasl server now expands the server_realm option.
1144 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1145 Patch by Jeremy Harris.
1147 PP/10 Added dbmjz lookup type.
1149 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1151 PP/12 MAIL args handles TAB as well as SP, for better interop with
1152 non-compliant senders.
1153 Analysis and variant patch by Todd Lyons.
1155 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1156 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1157 Patch from Dirk Mueller <dmueller@suse.com>
1159 PP/13 tls_peerdn now print-escaped for spool files.
1160 Observed some $tls_peerdn in wild which contained \n, which resulted
1161 in spool file corruption.
1163 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1164 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1165 or write after TLS renegotiation, which otherwise led to messages
1168 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1169 as a tracking header (ie: a signed header comes before the signature).
1170 Patch from Wolfgang Breyha.
1172 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1173 comma-sep list; embedded commas doubled.
1175 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1177 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1179 Report and patch from Dmitry Banschikov.
1181 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1182 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1183 failures. We appear to now support TLS1.1+ with Exim.
1185 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1186 lets Exim select keys and certificates based upon TLS SNI from client.
1187 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1188 before an outbound SMTP session. New log_selector, +tls_sni.
1190 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1191 NULL dereference. Report and patch from Alun Jones.
1193 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1194 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1195 Not seeing resolver debug output on NetBSD, but suspect this is a
1196 resolver implementation change.
1198 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1199 Left warnings. Added "eximon gdb" invocation mode.
1201 PP/21 Defaulting "accept_8bitmime" to true, not false.
1203 PP/22 Added -bw for inetd wait mode support.
1205 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1206 locate the relevant includes and libraries. Made this the default.
1208 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1209 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1211 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1212 This may cause build issues on older platforms.
1214 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1215 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1216 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1217 Added SNI support via GnuTLS too.
1218 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1220 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1222 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1223 Applied second patch from Janne, implementing suggestion to default
1224 multiple-strings-in-record handling to match SPF spec.
1226 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1228 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1229 read-only, out of scope).
1230 Patch from Wolfgang Breyha, report from Stuart Northfield.
1232 PP/29 Fix three issues highlighted by clang analyser static analysis.
1233 Only crash-plausible issue would require the Cambridge-specific
1234 iplookup router and a misconfiguration.
1235 Report from Marcin Mirosław.
1237 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1239 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1240 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1241 As part of this, removing so much warning spew let me fix some minor
1242 real issues in debug logging.
1244 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1245 assignment on my part. Fixed.
1247 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1248 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1249 Janne Snabb (who went above and beyond: thank you).
1251 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1252 string otherwise requires a connection and a bunch more work and it's
1253 relatively easy to get wrong. Should also expose TLS library linkage
1256 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1257 64-bit ${eval} (JH/03).
1259 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1260 GNU libc to support some of the 64-bit stuff, should not lead to
1261 conflicts. Defined before os.h is pulled in, so if a given platform
1262 needs to override this, it can.
1264 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1265 protection layer was required, which is not implemented.
1266 Bugzilla 1254, patch from Wolfgang Breyha.
1268 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1269 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1270 tls_dhparam take prime identifiers. Also unbreak combination of
1271 OpenSSL+DH_params+TLSSNI.
1273 PP/39 Disable SSLv2 by default in OpenSSL support.
1279 PP/01 Solaris build fix for Oracle's LDAP libraries.
1280 Bugzilla 1109, patch from Stephen Usher.
1282 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1284 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1287 TF/02 Fix a couple more cases where we did not log the error message
1288 when unlink() failed. See also change 4.74-TF/03.
1290 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1291 lock up or crash if it happened to be inside a call to libc when it
1292 got a SIGUSR1 from exiwhat.
1294 The SIGUSR1 handler appends the current process status to the process
1295 log which is later printed by exiwhat. It used to use the general
1296 purpose logging code to do this, but several functions it calls are
1297 not safe for signals.
1299 The new output code in the SIGUSR1 handler is specific to the process
1300 log, and simple enough that it's easy to inspect for signal safety.
1301 Removing some special cases also simplifies the general logging code.
1302 Removing the spurious timestamps from the process log simplifies
1305 TF/04 Improved ratelimit ACL condition.
1307 The /noupdate option has been deprecated in favour of /readonly which
1308 has clearer semantics. The /leaky, /strict, and /readonly update modes
1309 are mutually exclusive. The update mode is no longer included in the
1310 database key; it just determines when the database is updated. (This
1311 means that when you upgrade Exim will forget old rate measurements.)
1313 Exim now checks that the per_* options are used with an update mode that
1314 makes sense for the current ACL. For example, when Exim is processing a
1315 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1316 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1317 must specify per_mail/readonly. If you omit the update mode it defaults to
1318 /leaky where that makes sense (as before) or /readonly where required.
1320 The /noupdate option is now undocumented but still supported for
1321 backwards compatibility. It is equivalent to /readonly except that in
1322 ACLs where /readonly is required you may specify /leaky/noupdate or
1323 /strict/noupdate which are treated the same as /readonly.
1325 A useful new feature is the /count= option. This is a generalization
1326 of the per_byte option, so that you can measure the throughput of other
1327 aggregate values. For example, the per_byte option is now equivalent
1328 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1330 The per_rcpt option has been generalized using the /count= mechanism
1331 (though it's more complicated than the per_byte equivalence). When it is
1332 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1333 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1334 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1335 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1336 Note that using per_rcpt with a non-readonly update mode in more than
1337 one ACL will cause the recipients to be double-counted. (The per_mail
1338 and per_byte options don't have this problem.)
1340 The handling of very low rates has changed slightly. If the computed rate
1341 is less than the event's count (usually one) then this event is the first
1342 after a long gap. In this case the rate is set to the same as this event's
1343 count, so that the first message of a spam run is counted properly.
1345 The major new feature is a mechanism for counting the rate of unique
1346 events. The new per_addr option counts the number of different
1347 recipients that someone has sent messages to in the last time period. It
1348 behaves like per_rcpt if all the recipient addresses are different, but
1349 duplicate recipient addresses do not increase the measured rate. Like
1350 the /count= option this is a general mechanism, so the per_addr option
1351 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1352 example, measure the rate that a client uses different sender addresses
1353 with the options per_mail/unique=$sender_address. There are further
1354 details in the main documentation.
1356 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1358 TF/06 Removed a few PCRE remnants.
1360 TF/07 Automatically extract Exim's version number from tags in the git
1361 repository when doing development or release builds.
1363 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1364 Bugzilla 879. Patch from Paul Fisher.
1366 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1367 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1370 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1372 PP/05 Variable $av_failed, true if the AV scanner deferred.
1373 Bugzilla 1078. Patch from John Horne.
1375 PP/06 Stop make process more reliably on build failure.
1376 Bugzilla 1087. Patch from Heiko Schlittermann.
1378 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1379 Bugzilla 1089. Patch from Heiko Schlittermann.
1381 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1382 Bugzilla 1131. Patch from Holger Weiß.
1384 PP/09 Handle IPv6 addresses with SPF.
1385 Bugzilla 860. Patch from Wolfgang Breyha.
1387 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1389 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1392 PP/11 match_* no longer expand right-hand-side by default.
1393 New compile-time build option, EXPAND_LISTMATCH_RHS.
1394 New expansion conditions, "inlist", "inlisti".
1396 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1398 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1400 PP/14 fix log_write() format string regression from TF/03.
1401 Bugzilla 1152. Patch from Dmitry Isaikin.
1407 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1409 PP/02 Harmonised TLS library version reporting; only show if debugging.
1410 Layout now matches that introduced for other libraries in 4.74 PP/03.
1412 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1414 PP/04 New "dns_use_edns0" global option.
1416 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1419 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1420 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1422 TK/01 Updated PolarSSL code to 0.14.2.
1423 Bugzilla 1097. Patch from Andreas Metzler.
1425 PP/07 Catch divide-by-zero in ${eval:...}.
1426 Fixes bugzilla 1102.
1428 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1431 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1432 format-string attack -- SECURITY: remote arbitrary code execution.
1434 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1435 time unintentionally subject to list matching rules, letting the header
1436 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1437 arbitrary string expansion). This allowed for information disclosure.
1439 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1440 INT_MIN/-1 -- value coerced to INT_MAX.
1446 NM/01 Workaround for PCRE version dependency in version reporting
1449 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1450 This fixes portability to compilers other than gcc, notably
1451 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1453 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1454 makefiles for portability to HP-UX and POSIX correctness.
1456 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1457 Also via indented variable definition in the Makefile.
1458 (Debugging by Oliver Heesakkers).
1460 PP/02 Restore caching of spamd results with expanded spamd_address.
1461 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1463 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1464 Improves build reliability. Fix from: Frank Elsner
1466 NM/02 Fix wide character breakage in the rfc2047 coding
1467 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1469 NM/03 Allow underscore in dnslist lookups
1470 Fixes bug 1026. Patch from Graeme Fowler
1472 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1473 Code patches from Adam Ciarcinski of NetBSD.
1475 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1478 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1479 is logged, to avoid truncation. Patch from John Horne.
1481 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1482 Patch from Jakob Hirsch.
1484 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1485 SQL string expansion failure details.
1486 Patch from Andrey Oktyabrski.
1488 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1489 Patch from Simon Arlott.
1491 PP/09 New lookups functionality failed to compile on old gcc which rejects
1492 extern declarations in function scope.
1493 Patch from Oliver Fleischmann
1495 PP/10 Use sig_atomic_t for flags set from signal handlers.
1496 Check getgroups() return and improve debugging.
1497 Fixed developed for diagnosis in bug 927 (which turned out to be
1500 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1501 Patch from Mark Zealey.
1503 PP/12 Bugzilla 1056: Improved spamd server selection.
1504 Patch from Mark Zealey.
1506 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1507 Based on patch from Heiko Schlittermann.
1509 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1510 Patch from Uwe Doering, sign-off by Michael Haardt.
1512 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1513 variable declaration deep within a block. Bug and patch from
1516 PP/15 lookups-Makefile IRIX compatibility coercion.
1518 PP/16 Make DISABLE_DKIM build knob functional.
1520 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1521 Patch by Simon Arlott
1523 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1524 variable argument macros. Our copy now differs from upstream.
1530 TF/01 Failure to get a lock on a hints database can have serious
1531 consequences so log it to the panic log.
1533 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1534 controlled using the smtp_confirmation log selector.
1536 TF/03 Include the error message when we fail to unlink a spool file.
1538 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1539 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1540 for maintaining out-of-tree patches for some time.
1542 PP/01 Bugzilla 139: Documentation and portability issues.
1543 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1544 Handle per-OS dynamic-module compilation flags.
1546 PP/02 Let /dev/null have normal permissions.
1547 The 4.73 fixes were a little too stringent and complained about the
1548 permissions on /dev/null. Exempt it from some checks.
1549 Reported by Andreas M. Kirchwitz.
1551 PP/03 Report version information for many libraries, including
1552 Exim version information for dynamically loaded libraries. Created
1553 version.h, now support a version extension string for distributors
1554 who patch heavily. Dynamic module ABI change.
1556 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1557 privilege escalation vulnerability whereby the Exim run-time user
1558 can cause root to append content of the attacker's choosing to
1561 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1564 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1565 If dropping privileges for untrusted macros, we disabled normal logging
1566 on the basis that it would fail; for the Exim run-time user, this is not
1567 the case, and it resulted in successful deliveries going unlogged.
1568 Fixed. Reported by Andreas Metzler.
1574 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1575 only prepend for the Resent-* case. Fixes regression introduced in
1576 Exim 4.70 by NM/22 for Bugzilla 607.
1578 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1579 increasing numbers of administrators be bitten by this.
1581 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1583 PP/03 Bugzilla 994: added openssl_options main configuration option.
1585 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1587 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1589 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1591 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1592 without a peer certificate, leading to a segfault because of an
1593 assumption that peers always have certificates. Be a little more
1594 paranoid. Problem reported by Martin Tscholak.
1596 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1597 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1598 NB: ClamAV planning to remove STREAM in "middle of 2010".
1599 CL also introduces -bmalware, various -d+acl logging additions and
1600 more caution in buffer sizes.
1602 PP/09 Implemented reverse_ip expansion operator.
1604 PP/10 Bugzilla 937: provide a "debug" ACL control.
1606 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1608 PP/12 Bugzilla 973: Implement --version.
1610 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1612 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1614 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1616 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1617 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1618 ignore trailing whitespace.
1620 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1622 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1625 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1626 Notification from Dr Andrew Aitchison.
1628 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1629 ExtendedDetectionInfo response format.
1630 Notification from John Horne.
1632 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1635 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1636 XSL and documented dependency on system catalogs, with examples of how
1639 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1642 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1643 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1644 configuration file which is writeable by the Exim user or group.
1646 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1647 of configuration files to cover files specified with the -C option if
1648 they are going to be used with root privileges, not just the default
1651 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1652 option (effectively making it always true).
1654 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1655 files to be used while preserving root privileges.
1657 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1658 that rogue child processes cannot use them.
1660 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1661 run-time user, instead of root.
1663 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1664 Exim run-time user without dropping privileges.
1666 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1667 result string, instead of calling string_vformat() twice with the same
1670 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1671 for other users. Others should always drop root privileges if they use
1672 -C on the command line, even for a whitelisted configure file.
1674 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1676 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1682 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1683 $data_path, and $header_path variables; fixed documentation bugs and
1686 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1687 exipick to access non-standard spools, including the "frozen" queue
1690 NM/01 Bugzilla 965: Support mysql stored procedures.
1691 Patch from Alain Williams
1693 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1695 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1696 Patch from Andreas Metzler
1698 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1699 Patch from Kirill Miazine
1701 NM/05 Bugzilla 671: Added umask to procmail example.
1703 JJ/03 installed exipick 20100323.0, fixing doc bug
1705 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1706 directory. Notification and patch from Dan Rosenberg.
1708 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1710 TK/02 Improve log output when DKIM signing operation fails.
1712 MH/01 Treat the transport option dkim_domain as a colon separated
1713 list, not as a single string, and sign the message with each element,
1714 omitting multiple occurences of the same signer.
1716 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1717 Bugzilla 985, 986. Patch by Simon Arlott
1719 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1720 Patch by Simon Arlott
1722 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1723 MBX locking. Notification from Dan Rosenberg.
1729 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1731 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1733 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1735 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1737 NM/04 Bugzilla 915: Flag broken perl installation during build.
1743 TK/01 Added patch by Johannes Berg that expands the main option
1744 "spamd_address" if it starts with a dollar sign.
1746 TK/02 Write list of recipients to X-Envelope-Sender header when building
1747 the mbox-format spool file for content scanning (suggested by Jakob
1750 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1751 (http://www.dcc-servers.net/) support via dccifd. Activated by
1752 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1754 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1755 by Mark Daniel Reidel <mr@df.eu>.
1757 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1758 When building exim an external PCRE library is now needed -
1759 PCRE is a system library on the majority of modern systems.
1760 See entry on PCRE_LIBS in EDITME file.
1762 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1763 conversation. Added nologin parameter to request.
1764 Patch contributed by Kirill Miazine.
1766 TF/01 Do not log submission mode rewrites if they do not change the address.
1768 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1770 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1771 log files in place. Contributed by Roberto Lima.
1773 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1775 TF/03 Bugzilla 615: When checking the local_parts router precondition
1776 after a local_part_suffix or local_part_prefix option, Exim now
1777 does not use the address's named list lookup cache, since this
1778 contains cached lookups for the whole local part.
1780 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1781 Robert Millan. Documentation is in experimental-spec.txt.
1783 TF/04 Bugzilla 668: Fix parallel build (make -j).
1785 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1787 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1788 Patch provided by Jan Srzednicki.
1790 TF/05 Leading white space used to be stripped from $spam_report which
1791 wrecked the formatting. Now it is preserved.
1793 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1794 that they are available at delivery time.
1796 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1798 TF/08 TLS error reporting now respects the incoming_interface and
1799 incoming_port log selectors.
1801 TF/09 Produce a more useful error message if an SMTP transport's hosts
1802 setting expands to an empty string.
1804 NM/06 Bugzilla 744: EXPN did not work under TLS.
1805 Patch provided by Phil Pennock.
1807 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1808 Patch provided by Richard Godbee.
1810 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1811 acl_smtp_notquit, added index entry.
1813 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1814 Patch provided by Eugene Bujak.
1816 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1817 accept(). Patch provided by Maxim Dounin.
1819 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1820 Patch provided by Phil Pennock.
1822 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1824 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1825 Patch provided by Brad "anomie" Jorsch.
1827 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1828 Patch provided by Dean Brooks.
1830 TK/05 Add native DKIM support (does not depend on external libraries).
1832 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1833 Patch provided by Graeme Fowler.
1835 NM/16 Bugzilla 851: Documentation example syntax fix.
1837 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1839 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1842 NM/19 Bugzilla 745: TLS version reporting.
1843 Patch provided by Phil Pennock.
1845 NM/20 Bugzilla 167: bool: condition support.
1846 Patch provided by Phil Pennock.
1848 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1849 clients. Patch provided by Phil Pennock.
1851 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1852 Patch provided by Brad "anomie" Jorsch.
1854 NM/23 Bugzilla 687: Fix misparses in eximstats.
1855 Patch provided by Heiko Schlittermann.
1857 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1858 Patch provided by Heiko Schlittermann.
1860 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1861 plus update to original patch.
1863 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1865 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1866 Patch provided by David Brownlee.
1868 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1870 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1872 NM/30 Bugzilla 888: TLS documentation bugfixes.
1874 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1876 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1877 Unlike the original bugzilla I have changed all shell scripts in src tree.
1879 NM/33 Bugzilla 898: Transport filter timeout fix.
1880 Patch by Todd Rinaldo.
1882 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1883 Patch by Serge Demonchaux.
1885 NM/35 Bugzilla 39: Base64 decode bug fixes.
1886 Patch by Jakob Hirsch.
1888 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1890 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1892 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1894 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1900 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1901 ALT-N's libdkim that I have put here:
1902 http://duncanthrax.net/exim-experimental/
1904 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1905 were called 'true' and it seems that C99 defines that as a reserved
1906 keyword to be used with 'bool' variable types. That means you could
1907 not include C99-style headers which use bools without triggering
1908 build errors in sieve.c.
1910 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1911 as mailq or other aliases. Changed the --help handling significantly
1912 to do whats expected. exim_usage() emits usage/help information.
1914 SC/01 Added the -bylocaldomain option to eximstats.
1916 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1918 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1920 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1926 PH/01 Another patch from the Sieve maintainer.
1928 PH/02 When an IPv6 address is converted to a string for single-key lookup
1929 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1930 dots are used instead of colons so that keys in lsearch files need not
1931 contain colons. This was done some time before quoting was made available
1932 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1933 (notated using the quote facility) so as to distinguish them from IPv4
1934 keys. This meant that lookups for IP addresses in host lists did not work
1935 for iplsearch lookups.
1937 This has been fixed by arranging for IPv6 addresses to be expressed with
1938 colons if the lookup type is iplsearch. This is not incompatible, because
1939 previously such lookups could never work.
1941 The situation is now rather anomalous, since one *can* have colons in
1942 ordinary lsearch keys. However, making the change in all cases is
1943 incompatible and would probably break a number of configurations.
1945 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1948 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1949 conversion specification without a maximum field width, thereby enabling
1950 a rogue spamd server to cause a buffer overflow. While nobody in their
1951 right mind would setup Exim to query an untrusted spamd server, an
1952 attacker that gains access to a server running spamd could potentially
1953 exploit this vulnerability to run arbitrary code as the Exim user.
1955 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1956 $primary_hostname instead of what libspf2 thinks the hosts name is.
1958 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1959 a directory entry by the name of the lookup key. Previously, if a
1960 symlink pointed to a non-existing file or a file in a directory that
1961 Exim lacked permissions to read, a lookup for a key matching that
1962 symlink would fail. Now it is enough that a matching directory entry
1963 exists, symlink or not. (Bugzilla 503.)
1965 PH/03 The body_linecount and body_zerocount variables are now exported in the
1968 PH/04 Added the $dnslist_matched variable.
1970 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1971 This means they are set thereafter only if the connection becomes
1974 PH/06 Added the client_condition to authenticators so that some can be skipped
1975 by clients under certain conditions.
1977 PH/07 The error message for a badly-placed control=no_multiline_responses left
1978 "_responses" off the end of the name.
1980 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1982 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1983 (without spaces) instead of just copying the configuration text.
1985 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1987 PH/11 Added $max_received_linelength.
1989 PH/12 Added +ignore_defer and +include_defer to host lists.
1991 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1992 way in which PCRE > 7.0 is built.
1994 PH/14 Implemented queue_only_load_latch.
1996 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1997 MAIL command. The effect was to mangle the value on 64-bit systems.
1999 PH/16 Another patch from the Sieve maintainer.
2001 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2003 PH/18 If a system quota error occurred while trying to create the file for
2004 a maildir delivery, the message "Mailbox is full" was not appended to the
2005 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2006 only to a quota excession during the actual writing of the file.
2008 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2009 characters?) which causes problems in log lines. The DN values are now
2010 passed through string_printing() before being added to log lines.
2012 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2013 and InterBase are left for another time.)
2015 PH/21 Added message_body_newlines option.
2017 PH/22 Guard against possible overflow in moan_check_errorcopy().
2019 PH/23 POSIX allows open() to be a macro; guard against that.
2021 PH/24 If the recipient of an error message contained an @ in the local part
2022 (suitably quoted, of course), incorrect values were put in $domain and
2023 $local_part during the evaluation of errors_copy.
2029 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2030 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2033 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2034 issue a MAIL command.
2036 PH/02 In an ACL statement such as
2038 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2040 if a client was not listed at all, or was listed with a value other than
2041 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2042 the condition was not true (as it should be), so access was not denied.
2043 The bug was that the ! inversion was incorrectly passed on to the second
2044 item. This has been fixed.
2046 PH/03 Added additional dnslists conditions == and =& which are different from
2047 = and & when the dns lookup returns more than one IP address.
2049 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2050 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2052 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2053 FSYNC, which compiles an option called disable_fsync that allows for
2054 bypassing fsync(). The documentation is heavily laced with warnings.
2056 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2058 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2059 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2060 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2061 including adding "make clean"; (3) Added -fPIC when compiling the test
2062 dynamically loaded module, to get rid of a warning.
2064 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2065 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2066 = 0s. The bug is otherwise harmless.
2068 PH/07 There was a bug in the dovecot authenticator such that the value of
2069 $auth1 could be overwritten, and so not correctly preserved, after a
2070 successful authentication. This usually meant that the value preserved by
2071 the server_setid option was incorrect.
2073 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2075 PH/09 Installed PCRE release 7.0.
2077 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2078 run for batched SMTP input. It is now run at the start of every message
2079 in the batch. While fixing this I discovered that the process information
2080 (output by running exiwhat) was not always getting set for -bs and -bS
2081 input. This is fixed, and it now also says "batched" for BSMTP.
2083 PH/11 Added control=no_pipelining.
2085 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2086 patch, slightly modified), and move the expansion of helo_data till after
2087 the connection is made in the smtp transport (so it can use these
2090 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2092 PH/14 Added log_selector = +pid.
2094 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2096 PH/16 Add ${if forany and ${if forall.
2098 PH/17 Added dsn_from option to vary the From: line in DSNs.
2100 PH/18 Flush SMTP output before performing a callout, unless control =
2101 no_callout_flush is set.
2103 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2104 was true (the default) a successful delivery failed to delete the retry
2105 item, thus causing premature timeout of the address. The bug is now
2108 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2110 PH/21 Long custom messages for fakedefer and fakereject are now split up
2111 into multiline responses in the same way that messages for "deny" and
2112 other ACL rejections are.
2114 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2115 with slight modification.
2117 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2118 draft, changing the syntax and factoring some duplicate code".
2120 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2121 for deliveries of the second and subsequent messages over the same SMTP
2124 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2125 ${reduce, with only minor "tidies".
2127 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2129 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2130 expansion side effects.
2132 PH/27 When a message times out after an over-quota error from an Exim-imposed
2133 quota, the bounce message says "mailbox is full". This message was not
2134 being given when it was a system quota that was exceeded. It now should
2137 MH/03 Made $recipients available in local_scan(). local_scan() already has
2138 better access to the recipient list through recipients_list[], but
2139 $recipients can be useful in postmaster-provided expansion strings.
2141 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2142 in the case of a MAIL command with additional options following the
2143 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2144 were accidentally chopped off.
2146 PH/29 SMTP synchronization checks are implemented when a command is read -
2147 there is a check that no more input is waiting when there shouldn't be
2148 any. However, for some commands, a delay in an ACL can mean that it is
2149 some time before the response is written. In this time, more input might
2150 arrive, invalidly. So now there are extra checks after an ACL has run for
2151 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2152 pipelining has not been advertised.
2154 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2156 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2157 returned in the SMTP response when smtp_return_error_details was false.
2158 This has been fixed.
2160 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2161 instead of the C I/O that was originally supplied, because problems were
2162 reported on Solaris.
2164 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2165 Exim which did not show up earlier: it was assuming that a call to
2166 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2167 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2168 was a macro that became an assignment, so it seemed to work. This has
2169 changed to a proper function call with a void return, hence the compile
2170 error. Exim's code has been fixed.
2172 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2175 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2177 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2179 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2180 because it turns out to be too big or there is a timeout) there is no
2181 "Completed" line in the log. When some messages of this type were
2182 selected by exigrep, they were listed as "not completed". Others were
2183 picked up by some special patterns. I have improved the selection
2184 criteria to be more general.
2186 PH/38 The host_find_failed option in the manualroute router can now be set
2187 to "ignore", to completely ignore a host whose IP address cannot be
2188 found. If all hosts are ignored, the behaviour is controlled by the new
2189 host_all_ignored option.
2191 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2192 homing or because of multiple MX records with /mx) generated more than
2193 one IP address, and the following item turned out to be the local host,
2194 all the secondary addresses of the first item were incorrectly removed
2195 from the list, along with the local host and any following hosts (which
2196 is what is supposed to happen).
2198 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2199 whoever called Exim into the -H file. In the case of the daemon it was
2200 behaving confusingly. When first started, it used values for whoever
2201 started the daemon, but after a SIGHUP it used the Exim user (because it
2202 calls itself on a restart). I have changed the code so that it now always
2205 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2206 message are rejected with the same error (e.g. no authentication or bad
2207 sender address), and a DATA command is nevertheless sent (as can happen
2208 with PIPELINING or a stupid MUA), the error message that was given to the
2209 RCPT commands is included in the rejection of the DATA command. This is
2210 intended to be helpful for MUAs that show only the final error to their
2213 PH/42 Another patch from the Sieve maintainer.
2215 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2216 Eximstats - Fixed some broken HTML links and added missing column headers
2218 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2219 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2221 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2223 PH/43 Yet another patch from the Sieve maintainer.
2225 PH/44 I found a way to check for a TCP/IP connection going away before sending
2226 the response to the final '.' that terminates a message, but only in the
2227 case where the client has not sent further data following the '.'
2228 (unfortunately, this is allowed). However, in many cases there won't be
2229 any further data because there won't be any more messages to send. A call
2230 to select() can be used: if it shows that the input is "ready", there is
2231 either input waiting, or the socket has been closed. An attempt to read
2232 the next input character can distinguish the two cases. Previously, Exim
2233 would have sent an OK response which the client would never have see.
2234 This could lead to message repetition. This fix should cure that, at
2235 least in a lot of common cases.
2237 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2238 advertised in response to EHLO.
2244 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2245 fixed by 4.65/MH/01 (is this a record?) are fixed:
2247 (i) An empty string was always treated as zero by the numeric comparison
2248 operators. This behaviour has been restored.
2250 (ii) It is documented that the numeric comparison operators always treat
2251 their arguments as decimal numbers. This was broken in that numbers
2252 starting with 0 were being interpreted as octal.
2254 While fixing these problems I realized that there was another issue that
2255 hadn't been noticed. Values of message_size_limit (both the global option
2256 and the transport option) were treated as octal if they started with 0.
2257 The documentation was vague. These values are now always treated as
2258 decimal, and I will make that clear in the documentation.
2264 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2265 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2268 MH/01 Don't check that the operands of numeric comparison operators are
2269 integers when their expansion is in "skipping" mode (fixes bug
2270 introduced by 4.64-PH/07).
2272 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2273 child addresses, Exim now panics and dies. Previously, because the count
2274 is held in a short int, deliveries were likely to be lost. As such a
2275 large number of recipients for a single message is ridiculous
2276 (performance will be very, very poor), I have chosen to impose a limit
2277 rather than extend the field.
2283 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2284 leftover -K file (the existence of which was triggered by #402).
2285 While we were at it, introduced process PID as part of the -K
2286 filename. This should rule out race conditions when creating
2289 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2290 processing considerably. Previous code took too long for large mails,
2291 triggering a timeout which in turn triggers #401.
2293 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2294 in the DK code in transports.c. sendfile() is not really portable,
2295 hence the _LINUX specificness.
2297 TF/01 In the add_headers option to the mail command in an Exim filter,
2298 there was a bug that Exim would claim a syntax error in any
2299 header after the first one which had an odd number of characters
2302 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2303 callout verification, Exim cached a "reject" for the entire domain. This
2304 is correct for most verifications, but it is not correct for a recipient
2305 verification with use_sender or use_postmaster set, because in that case
2306 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2307 case of MAIL FROM:<> rejection from other early rejections (e.g.
2308 rejection of HELO). When verifying a recipient using a non-null MAIL
2309 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2310 Whatever the result of the callout, the value of the domain cache is
2311 left unchanged (for any other kind of callout, getting as far as trying
2312 RCPT means that the domain itself is ok).
2314 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2317 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2318 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2319 session, the daemon ignores ECONNECT errors and logs others; it now
2320 ignores EPIPE as well.
2322 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2323 (quoted-printable decoding).
2325 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2326 later the small subsequent patch to fix an introduced bug.
2328 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2330 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2332 PH/08 An error is now given if message_size_limit is specified negative.
2334 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2335 to be given (somewhat) arbitrary names.
2337 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2340 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2341 miscellaneous code fixes
2343 PH/10 Added the log_reject_target ACL modifier to specify where to log
2346 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2347 hostname. This is wrong, because it relates to the incoming message (and
2348 probably the interface on which it is arriving) and not to the outgoing
2349 callout (which could be using a different interface). This has been
2350 changed to use the value of the helo_data option from the smtp transport
2351 instead - this is what is used when a message is actually being sent. If
2352 there is no remote transport (possible with a router that sets up host
2353 addresses), $smtp_active_hostname is used.
2355 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2356 tweaks were necessary in order to get it to work (see also 21 below):
2357 (a) The code assumed that strncpy() returns a negative number on buffer
2358 overflow, which isn't the case. Replaced with Exim's string_format()
2360 (b) There were several signed/unsigned issues. I just did the minimum
2361 hacking in of casts. There is scope for a larger refactoring.
2362 (c) The code used strcasecmp() which is not a standard C function.
2363 Replaced with Exim's strcmpic() function.
2364 (d) The code set only $1; it now sets $auth1 as well.
2365 (e) A simple test gave the error "authentication client didn't specify
2366 service in request". It would seem that Dovecot has changed its
2367 interface. Fortunately there's a specification; I followed it and
2368 changed what the client sends and it appears to be working now.
2370 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2373 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2374 address A is aliased to B and C, where B exists and C does not. Without
2375 -v the output is "A verified" because verification stops after a
2376 successful redirection if more than one address is generated. However,
2377 with -v the child addresses are also verified. Exim was outputting "A
2378 failed to verify" and then showing the successful verification for C,
2379 with its parentage. It now outputs "B failed to verify", showing B's
2380 parentage before showing the successful verification of C.
2382 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2383 look up a TXT record in a specific list after matching in a combined
2386 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2387 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2388 they consult the DNS. I had assumed they would set it the way they
2389 wanted; and indeed my experiments on Linux seem to show that in some
2390 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2391 To be on the safe side, however, I have now made the interface to
2392 host_find_byname() similar to host_find_bydns(), with an argument
2393 containing the DNS resolver options. The host_find_byname() function now
2394 sets these options at its start, just as host_find_bydns() does. The smtp
2395 transport options dns_qualify_single and dns_search_parents are passed to
2396 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2397 of host_find_byname() use the default settings of RES_DEFNAMES
2398 (qualify_single) but not RES_DNSRCH (search_parents).
2400 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2401 spool_read_header() do less string testing, by means of a preliminary
2402 switch on the second character of optional "-foo" lines. (This is
2403 overdue, caused by the large number of possibilities that now exist.
2404 Originally there were few.) While I was there, I also converted the
2405 str(n)cmp tests so they don't re-test the leading "-" and the first
2406 character, in the hope this might squeeze out yet more improvement.
2408 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2409 flag allowing group syntax was set by the header_syntax check but not
2410 turned off, possible causing trouble later; (2) The flag was not being
2411 set at all for the header_verify test, causing "group"-style headers to
2412 be rejected. I have now set it in this case, and also caused header_
2413 verify to ignore an empty address taken from a group. While doing this, I
2414 came across some other cases where the code for allowing group syntax
2415 while scanning a header line wasn't quite right (mostly, not resetting
2416 the flag correctly in the right place). These bugs could have caused
2417 trouble for malformed header lines. I hope it is now all correct.
2419 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2420 with the "reply" argument non-NULL. The code, however (which originally
2421 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2422 but it didn't always do it. This confused somebody who was copying the
2423 code for some other use. I have removed all the tests.
2425 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2426 feature that was used to support insecure browsers during the U.S. crypto
2427 embargo. It requires special client support, and Exim is probably the
2428 only MTA that supported it -- and would never use it because real RSA is
2429 always available. This code has been removed, because it had the bad
2430 effect of slowing Exim down by computing (never used) parameters for the
2431 RSA_EXPORT functionality.
2433 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2434 authenticator to fail if there's a tab character in the incoming data
2435 (there should never be unless someone is messing about, as it's supposed
2436 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2437 if the connection is using TLS or if the remote IP is the same as the
2438 local IP, and the "valid-client-cert option" if a client certificate has
2441 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2442 authenticators. This can be used for authorization after authentication
2443 succeeds. (In the case of plaintext, it servers for both authentication
2446 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2447 if any retry times were supplied.
2449 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2450 connection, where there is no client IP address to check. In this
2451 situation, the verify now always succeeds.
2453 PH/25 Applied John Jetmore's -Mset patch.
2455 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2457 PH/27 In a string expansion for a processed (not raw) header when multiple
2458 headers of the same name were present, leading whitespace was being
2459 removed from all of them, but trailing whitespace was being removed only
2460 from the last one. Now trailing whitespace is removed from each header
2461 before concatenation. Completely empty headers in a concatenation (as
2462 before) are ignored.
2464 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2465 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2467 PH/29 [Removed. This was a change that I later backed out, and forgot to
2468 correct the ChangeLog entry (that I had efficiently created) before
2469 committing the later change.]
2471 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2472 address errors (4xx response to RCPT) over the same connection as other
2473 messages routed to the same hosts. Such deliveries are always "forced",
2474 so retry times are not inspected. This resulted in far too many retries
2475 for the affected addresses. The effect occurred only when there were more
2476 hosts than the hosts_max_try setting in the smtp transport when it had
2477 the 4xx errors. Those hosts that it had tried were not added to the list
2478 of hosts for which the message was waiting, so if all were tried, there
2479 was no problem. Two fixes have been applied:
2481 (i) If there are any address or message errors in an SMTP delivery, none
2482 of the hosts (tried or untried) are now added to the list of hosts
2483 for which the message is waiting, so the message should not be a
2484 candidate for sending over the same connection that was used for a
2485 successful delivery of some other message. This seems entirely
2486 reasonable: after all the message is NOT "waiting for some host".
2487 This is so "obvious" that I'm not sure why it wasn't done
2488 previously. Hope I haven't missed anything, but it can't do any
2489 harm, as the worst effect is to miss an optimization.
2491 (ii) If, despite (i), such a delivery is accidentally attempted, the
2492 routing retry time is respected, so at least it doesn't keep
2493 hammering the server.
2495 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2496 in ${readsocket because some servers need this prod.
2498 PH/32 Added some extra debug output when updating a wait-xxx database.
2500 PH/33 The hint "could be header name not terminated by colon", which has been
2501 given for certain expansion errors for a long time, was not being given
2502 for the ${if def:h_colon_omitted{... case.
2504 PH/34 The spec says: "With one important exception, whenever a domain list is
2505 being scanned, $domain contains the subject domain." There was at least
2506 one case where this was not true.
2508 PH/35 The error "getsockname() failed: connection reset by peer" was being
2509 written to the panic log as well as the main log, but it isn't really
2510 panic-worthy as it just means the connection died rather early on. I have
2511 removed the panic log writing for the ECONNRESET error when getsockname()
2514 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2515 runs only) independently of the message's sender address. This meant
2516 that, if the 4xx error was in fact related to the sender, a different
2517 message to the same recipient with a different sender could confuse
2518 things. In particular, this can happen when sending to a greylisting
2519 server, but other circumstances could also provoke similar problems.
2520 I have changed the default so that the retry time for these errors is now
2521 based a combination of the sender and recipient addresses. This change
2522 can be overridden by setting address_retry_include_sender=false in the
2525 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2526 remote server are returned as part of bounce messages. This was not
2527 happening for LMTP over a pipe (the lmtp transport), but now it is the
2528 same for both kinds of LMTP.
2530 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2531 in header lines that were in fact CNAMEs. This is no longer the case.
2533 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2534 and queue runs started by the daemon processed all messages. This has
2535 been fixed so that -R and -S can now usefully be given with -q<time>.
2537 PH/40 Import PCRE release 6.7 (fixes some bugs).
2539 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2541 PH/42 Give an error if -q is specified more than once.
2543 PH/43 Renamed the variables $interface_address and $interface_port as
2544 $received_ip_address and $received_port, to make it clear that these
2545 values apply to message reception, and not to the outgoing interface when
2546 a message is delivered. (The old names remain recognized, of course.)
2548 PH/44 There was no timeout on the connect() call when using a Unix domain
2549 socket in the ${readsocket expansion. There now is.
2551 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2552 be meaningful with "accept".
2554 SC/01 Eximstats V1.43
2555 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2557 SC/02 Eximstats V1.44
2558 Use a glob alias rather than an array ref in the generated
2559 parser. This improves both readability and performance.
2561 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2562 Collect SpamAssassin and rejection statistics.
2563 Don't display local sender or destination tables unless
2564 there is data to show.
2565 Added average volumes into the top table text output.
2567 SC/04 Eximstats V1.46
2568 Collect data on the number of addresses (recipients)
2569 as well as the number of messages.
2571 SC/05 Eximstats V1.47
2572 Added 'Message too big' to the list of mail rejection
2573 reasons (thanks to Marco Gaiarin).
2575 SC/06 Eximstats V1.48
2576 Mainlog lines which have GMT offsets and are too short to
2577 have a flag are now skipped.
2579 SC/07 Eximstats V1.49 (Alain Williams)
2580 Added the -emptyok flag.
2582 SC/08 Eximstats V1.50
2583 Fixes for obtaining the IP address from reject messages.
2585 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2586 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2587 whitespace changes from 4.64-PH/27
2589 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2592 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2593 are found, allow negative numbers in numeric criteria)
2595 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2597 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2598 to match changes made in 4.64-PH/43
2600 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2602 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2603 "domain" is actually an IP address, and force a failure. This locks out
2604 those revolvers/nameservers that support "A-for-A" lookups, in
2605 contravention of the specifications.
2607 PH/48 When a host name was looked up from an IP address, and the subsequent
2608 forward lookup of the name timed out, the host name was left in
2609 $sender_host_name, contrary to the specification.
2611 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2612 restricted to single-key lookups, Exim was not diagnosing an error if
2613 * or *@ was used with a query-style lookup.
2615 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2617 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2618 long ago, but noone interested enough thought of it. Let's just say that
2619 the "1.1" means that there are some new functions that weren't there at
2620 some point in the past.
2622 PH/51 Error processing for expansion failure of helo_data from an smtp
2623 transport during callout processing was broken.
2625 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2626 tested/used via the -bh/-bhc/-bs options.
2628 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2629 bug, fixed in subsequent PCRE releases).
2631 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2632 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2634 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2640 SC/01 Use a glob alias rather than an array ref in eximstats generated
2641 parser. This improves both readability and performance.
2643 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2644 Don't display local sender or destination tables in eximstats unless
2645 there is data to show.
2646 Added average volumes into the eximstats top table text output.
2648 SC/03 Collect data on the number of addresses (recipients) as well
2649 as the number of messages in eximstats.
2651 TF/01 Correct an error in the documentation for the redirect router. Exim
2652 does (usually) call initgroups() when daemonizing.
2654 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2655 with consistent privilege compared to when running as a daemon.
2657 TF/03 Note in the spec that $authenticated_id is not set for local
2658 submissions from trusted users.
2660 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2661 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2663 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2664 by adding some example configuration directives to the default
2665 configuration file. A little bit of work is required to uncomment the
2666 directives and define how usernames and passwords are checked, but
2667 there is now a framework to start from.
2669 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2670 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2671 without this. I don't know how relevant this is to other LDAP libraries.
2673 PH/02 Add the verb name to the "unknown ACL verb" error.
2675 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2677 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2679 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2680 directory not expanded when it should be if an expanded home directory
2681 was set for the address (which is overridden by the transport).
2683 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2686 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2687 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2688 because it is too late at that time, and has no effect.
2690 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2691 security issue with \' (bugzilla #107). I could not use the
2692 PQescapeStringConn() function, because it needs a PGconn value as one of
2695 PH/08 When testing addresses using -bt, indicate those final addresses that
2696 are duplicates that would not cause an additional delivery. At least one
2697 person was confused, thinking that -bt output corresponded to deliveries.
2698 (Suppressing duplicates isn't a good idea as you lose the information
2699 about possibly different redirections that led to the duplicates.)
2701 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2702 systems where poll() doesn't work, in particular OS X.
2704 PH/10 Added more information to debugging output for retry time not reached.
2706 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2707 operations in malware.c.
2709 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2712 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2713 syslog_duplication set false, log lines that would normally be written
2714 both the the main log and to the reject log were not written to syslog at
2717 PH/14 In the default configuration, change the use of "message" in ACL warn
2718 statements to "add_header".
2720 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2721 not followed by a command (e.g. "seen endif").
2723 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2724 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2727 PH/17 Added extra conditions to the default value of delay_warning_condition
2731 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2732 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2733 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2736 The Auto-Submitted: and various List- headers are standardised, whereas I
2737 don't think Precedence: ever was.
2739 PH/18 Refactored debugging code in route_finduser() to show more information,
2740 in particular, the error code if getpwnam() issues one.
2742 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2743 This is apparently needed in addition to the PH/07 change above to avoid
2744 any possible encoding problems.
2746 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2747 but not after initializing Perl.
2749 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2750 output them only if debugging. By default they are written stderr,
2751 apparently, which is not desirable.
2753 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2756 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2759 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2761 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2762 authenticated or an ident call has been made. Suppress the default
2763 values for $authenticated_id and $authenticated_sender (but permit -oMai
2764 and -oMas) when testing with -bh.
2766 PH/24 Re-jigged the order of the tests in the default configuration so that the
2767 tests for valid domains and recipients precede the DNS black list and CSA
2768 tests, on the grounds that those ones are more expensive.
2770 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2771 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2772 This bug exists in every version of Exim that I still have, right back to
2775 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2776 However, an attempt to turn on case-sensitivity in a regex key by
2777 including (?-i) didn't work because the subject string was already
2778 lowercased, and the effects were non-intuitive. It turns out that a
2779 one-line patch can be used to allow (?-i) to work as expected.
2785 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2786 other effects) broke the use of negated acl sub-conditions.
2788 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2791 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2792 "Deny" causes Exim to reject the incoming connection with a 554 error.
2793 Unfortunately, if there is a major crisis, such as a disk failure,
2794 tcp-wrappers gives "deny", whereas what one would like would be some
2795 kind of temporary error. A kludge has been added to help with this.
2796 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2797 554 error is used if errno is still zero or contains ENOENT (which occurs
2798 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2801 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2803 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2804 errors. Otherwise a message that provokes a temporary error (when other
2805 messages do not) can cause a whole host to time out.
2807 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2808 addresses were routed directly to files or pipes from a redirect router.
2809 File deliveries just didn't batch; pipe deliveries might have suffered
2812 PH/06 A failure to get a lock for a hints database would erroneously always say
2813 "Failed to get write lock", even when it was really a read lock.
2815 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2816 of 0600. This has been changed to use the value of the lockfile_mode
2817 option (which defaults to 0600).
2819 PH/08 Applied small patch from the Sieve maintainer.
2821 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2822 folder from quota calculations, a direct delivery into this folder messed
2823 up the contents of the maildirsize file. This was because the regex was
2824 used only to exclude .Trash (or whatever) when the size of the mailbox
2825 was calculated. There was no check that a delivery was happening into an
2826 excluded directory. This bug has been fixed by ignoring all quota
2827 processing for deliveries into excluded directories.
2829 PH/10 Added the maildirfolder_create_regex option to appendfile.
2835 PH/01 The code for finding all the local interface addresses on a FreeBSD
2836 system running IPv6 was broken. This may well have applied to all BSD
2837 systems, as well as to others that have similar system calls. The broken
2838 code found IPv4 interfaces correctly, but gave incorrect values for the
2839 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2840 that it would not match correctly against @[] and not recognize the IPv6
2843 PH/02 The ipliteral router was not recognizing addresses of the form user@
2844 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2846 PH/03 Added disable_ipv6.
2848 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2849 decoded form, because it is most often used to construct To: headers
2850 lines in autoreplies, and the decoded form may well be syntactically
2851 invalid. However, $reply_address has leading white space removed, and all
2852 newlines turned into spaces so that the autoreply transport does not
2855 PH/05 If group was specified without a user on a router, and no group or user
2856 was specified on a transport, the group from the router was ignored.
2858 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2859 for visible compile-time settings that can be used to change these
2860 numbers, for those that want even more. Backwards compatibility with old
2861 spool files has been maintained. However, going back to a previous Exim
2862 release will lost any variables that are in spool files.
2864 PH/07 Two small changes when running in the test harness: increase delay when
2865 passing a TCP/IP connection to a new process, in case the original
2866 process has to generate a bounce, and remove special handling of
2867 127.0.0.2 (sic), which is no longer necessary.
2869 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2870 be the same on different OS.
2872 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2875 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2876 whether --show-vars was specified or not
2878 JJ/02 exipick: Added support for new ACL variable spool format introduced
2881 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2882 syntactically invalid From: or Reply-to: line, and a filter used this to
2883 generate an autoreply, and therefore failed to obtain an address for the
2884 autoreply, Exim could try to deliver to a non-existent relative file
2885 name, causing unrelated and misleading errors. What now happens is that
2886 it logs this as a hard delivery error, but does not attempt to create a
2889 PH/11 The exinext utility has a -C option for testing purposes, but although
2890 the given file was scanned by exinext itself; it wasn't being passed on
2891 when Exim was called.
2893 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2894 an end-of-file indication when reading a command response.
2896 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2897 compiled. In many other places in Exim, IPv6 addresses are always
2898 recognized, so I have changed this. It also means that IPv4 domain
2899 literals of the form [IPV4:n.n.n.n] are now always recognized.
2901 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2902 used if the router is not running as root, for example, when verifying at
2903 ACL time, or when using -bh. The debugging output from this situation was
2904 non-existent - all you got was a failure to exec. I have made two
2907 (a) Failures to set uid/gid, the current directory, or a process leader
2908 in a subprocess such as that created by queryprogram now generate
2909 suitable debugging output when -d is set.
2911 (b) The queryprogram router detects when it is not running as root,
2912 outputs suitable debugging information if -d is set, and then runs
2913 the subprocess without attempting to change uid/gid.
2915 PH/15 Minor change to Makefile for building test_host (undocumented testing
2918 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2919 additional section of a DNS packet that returns MX or SRV records.
2920 Instead, it always explicitly searches for A/AAAA records. This avoids
2921 major problems that occur when a DNS server includes only records of one
2922 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2923 fixed another bug: if SRV records were looked up and the corresponding
2924 address records were *not* found in the additional section, the port
2925 values from the SRV records were lost.
2927 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2928 using the correct key (the original address) when searching the retry
2929 rules in order to find which one to use for generating the retry hint.
2931 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2932 adding the default one. Similarly, if it contains a Reply-To: header, the
2933 errors_reply_to option, if set, is not used.
2935 PH/19 When calculating a retry time, Exim used to measure the "time since
2936 failure" by looking at the "first failed" field in the retry record. Now
2937 it does not use this if it is later than than the arrival time of the
2938 message. Instead it uses the arrival time. This makes for better
2939 behaviour in cases where some deliveries succeed, thus re-setting the
2940 "first failed" field. An example is a quota failure for a huge message
2941 when small messages continue to be delivered. Without this change, the
2942 "time since failure" will always be short, possible causing more frequent
2943 delivery attempts for the huge message than are intended.
2944 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2946 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2947 $1, $2, $3) because the numerical variables can be reset during some
2948 expansion items (e.g. "match"), thereby losing the authentication data.
2950 PH/21 Make -bV show the size of off_t variables so that the test suite can
2951 decide whether to run tests for quotas > 2G.
2953 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2954 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2955 filecount value is greater than 2G or if a quota value is greater than 2G
2956 on a system where the size of off_t is not greater than 4, a panic error
2959 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2960 never match. The debug and -bh output now contains an explicit error
2961 message indicating a malformed IPv4 address or mask.
2963 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2964 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2965 PH/23 above applies.
2967 PH/25 Do not write to syslog when running in the test harness. The only
2968 occasion when this arises is a failure to open the main or panic logs
2969 (for which there is an explicit test).
2971 PH/26 Added the /no_tell option to "control=freeze".
2973 PH/27 If a host name lookup failed very early in a connection, for example, if
2974 the IP address matched host_lookup and the reverse lookup yielded a name
2975 that did not have a forward lookup, an error message of the form "no IP
2976 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2977 could be logged. Now it outputs the IP address instead of "NULL".
2979 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2980 allows the sender and the authenticated sender to be set when
2981 submitting a message from within Exim. Since child_open_exim() is
2982 documented for local_scan(), the new function should be too.
2984 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2985 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2986 results in an empty string is now treated as unset.
2988 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2990 PH/31 Added sender_verify_fail logging option.
2992 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2993 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2994 tidied the source and removed it altogether.
2996 PH/33 When a queue run was abandoned because the load average was too high, a
2997 log line was always written; now it is written only if the queue_run log
2998 selector is set. In addition, the log line for abandonment now contains
2999 information about the queue run such as the pid. This is always present
3000 in "start" and "stop" lines but was omitted from the "abandon" line.
3002 PH/34 Omit spaces between a header name and the colon in the error message that
3003 is given when verify = headers_syntax fails (if there are lots of them,
3004 the message gets confusing).
3006 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3007 names, as there are now some PTR records that contain slashes. This check
3008 is only to protect against broken name servers that fall over on strange
3009 characters, so the fact that it applies to all lookups doesn't matter.
3011 PH/36 Now that the new test suite is complete, we can remove some of the
3012 special code in Exim that was needed for the old test suite. For example,
3013 sorting DNS records because real resolvers return them in an arbitrary
3014 order. The new test suite's fake resolver always returns records in the
3017 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3018 bounces) except when queue_only is set, to avoid logging races between
3019 the different processes.
3021 PH/38 Panic-die if .include specifies a non-absolute path.
3023 PH/39 A tweak to the "H" retry rule from its user.
3025 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3026 a label. They prevented compilation on older perls.
3028 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3029 a warning to be raised on newish perls.
3031 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3032 on queue. Changes to match documented behaviour of showing count of
3033 messages matching specified criteria.
3035 PH/40 Changed the default ident timeout from 30s to 5s.
3037 PH/41 Added support for the use of login_cap features, on those BSD systems
3038 that have them, for controlling the resources used by pipe deliveries.
3040 PH/42 The content-scanning code uses fopen() to create files in which to put
3041 message data. Previously it was not paying any attention to the mode of
3042 the files. Exim runs with umask(0) because the rest of the code creates
3043 files with open(), and sets the required mode explicitly. Thus, these
3044 files were ending up world-writeable. This was not a big issue, because,
3045 being within the spool directory, they were not world-accessible. I have
3046 created a function called modefopen, which takes an additional mode
3047 argument. It sets umask(777), creates the file, chmods it to the required
3048 mode, then resets the umask. All the relevant calls to fopen() in the
3049 content scanning code have been changed to use this function.
3051 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3052 to 24 hours. This avoids potential overflow problems when processing G
3053 and H retry rules. I suspect nobody ever tinkers with this value.
3055 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3057 PH/45 When the plaintext authenticator is running as a client, the server's
3058 challenges are checked to ensure they are valid base64 strings. By
3059 default, the authentication attempt is cancelled if an invalid string is
3060 received. Setting client_ignore_invalid_base64 true ignores these errors.
3061 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3062 they are received. Thus, the responses can be made to depend on the
3063 challenges. If an invalid string is ignored, an empty string is placed in
3066 PH/46 Messages that are created by the autoreply transport now contains a
3067 References: header, in accordance with RFCs 2822 and 3834.
3069 PH/47 Added authenticated_sender_force to the smtp transport.
3071 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3073 PH/49 Installed latest patch from the Sieve maintainer.
3075 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3076 was also set, the appendfile transport was unnecessarily scanning a
3077 directory of message files (e.g. for maildir delivery) to find the count
3078 of files (along with the size), even though it did not need this
3079 information. It now does the scan only if it needs to find either the
3080 size of the count of files.
3082 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3084 PH/52 Two bugs concerned with error handling when the smtp transport is
3087 (i) Exim was not creating retry information for temporary errors given
3088 for individual recipients after the DATA command when the smtp transport
3089 was used in LMTP mode. This meant that they could be retried too
3090 frequently, and not timed out correctly.
3092 (ii) Exim was setting the flag that allows error details to be returned
3093 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3094 recipients that were returned after the DATA command.
3096 PH/53 This is related to PH/52, but is more general: for any failing address,
3097 when detailed error information was permitted to be returned to the
3098 sender, but the error was temporary, then after the final timeout, only
3099 "retry timeout exceeded" was returned. Now it returns the full error as
3100 well as "retry timeout exceeded".
3102 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3103 do this, and (what is worse) MTAs that accept it.
3105 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3106 will now be deprecated.
3108 PH/56 New os.c-cygwin from the Cygwin maintainer.
3110 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3111 formats (previously only available in exim formats via -bpr, -bpru,
3112 and -bpra. Now also available in native and exiqgrep formats)
3114 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3115 with very large, slow to parse queues
3117 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3119 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3121 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3122 responses to authentication challenges, though it was showing the
3123 challenges; (ii) I've removed the CR characters from the debug output for
3126 PH/58 Allow for the insertion of a newline as well as a space when a string
3127 is turned into more than one encoded-word during RFC 2047 encoding. The
3128 Sieve code now uses this.
3130 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3131 data_4xx, lost_connection, tls_required.
3133 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3134 message was being sent as an SMTP response.
3136 PH/61 Add -l and -k options to exicyclog.
3138 PH/62 When verifying, if an address was redirected to one new address, so that
3139 verification continued, and the new address failed or deferred after
3140 having set something in $address_data, the value of $address_data was not
3141 passed back to the ACL. This was different to the case when no
3142 redirection occurred. The value is now passed back in both cases.
3144 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3145 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3146 use login_cap.h, so on its own it isn't the distinguishing feature. The
3147 new name refers directly to the setclassresources() function.
3149 PH/65 Added configuration files for NetBSD3.
3151 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3153 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3154 is preferred over IPv4.
3156 PH/68 The bounce_return_message and bounce_return_body options were not being
3157 honoured for bounces generated during the reception of non-SMTP messages.
3158 In particular, this applied to messages rejected by the ACL. This bug has
3159 been fixed. However, if bounce_return_message is true and bounce_return_
3160 body is false, the headers that are returned for a non-SMTP message
3161 include only those that have been read before the error was detected.
3162 (In the case of an ACL rejection, they have all been read.)
3164 PH/69 The HTML version of the specification is now built in a directory called
3165 spec_html instead of spec.html, because the latter looks like a path with
3166 a MIME-type, and this confuses some software.
3168 PH/70 Catch two compiler warnings in sieve.c.
3170 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3171 function verify_get_ident() calls ip_connect() to connect a socket, but
3172 if the "connect()" function timed out, ip_connect() used to close the
3173 socket. However, verify_get_ident() also closes the socket later, and in
3174 between Exim writes to the log, which may get opened at this point. When
3175 the socket was closed in ip_connect(), the log could get the same file
3176 descriptor number as the socket. This naturally causes chaos. The fix is
3177 not to close the socket in ip_connect(); the socket should be closed by
3178 the function that creates it. There was only one place in the code where
3179 this was missing, in the iplookup router, which I don't think anybody now
3180 uses, but I've fixed it anyway.
3182 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3183 well as to direct DNS lookups. Otherwise the handling of names in host
3184 lists is inconsistent and therefore confusing.
3190 PH/01 Two changes to the default runtime configuration:
3192 (1) Move the checks for relay_from_hosts and authenticated clients from
3193 after to before the (commented out) DNS black list checks.
3195 (2) Add control=submission to the relay_from_hosts and authenticated
3196 clients checks, on the grounds that messages accepted by these
3197 statements are most likely to be submissions.
3199 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3201 (1) Generate an error if the third argument for the ${prvs expansion is
3204 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3207 (3) Reset the variables that are obtained from the first argument of
3208 ${prvscheck and used in the second argument before leaving the code,
3209 because their memory is reclaimed, so using them afterwards may do
3212 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3213 one (it's much easier than Tom thought :-).
3215 (5) Because of (4), we can now allow for the use of $prvscheck_result
3216 inside the third argument.
3218 PH/03 For some reason, the default setting of PATH when running a command from
3219 a pipe transport was just "/usr/bin". I have changed it to
3222 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3223 anything to be listed in the output from -bV.
3225 PH/05 When a filter generated an autoreply, the entire To: header line was
3226 quoted in the delivery log line, like this:
3228 => >A.N.Other <ano@some.domain> <original@ddress> ...
3230 This has been changed so that it extracts the operative address. There
3231 may be more than one such address. If so, they are comma-separated, like
3234 => >ano@some.domain,ona@other.domain <original@ddress> ...
3236 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3237 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3238 not being looked up in the rDNS to get a host name, Exim was showing the
3239 IP address twice in Received: lines, even though the IP addresses were
3240 identical. For example:
3242 Received: from [1.2.3.4] (helo=[1.2.3.4])
3244 However, if the real host name was known, it was omitting the HELO data
3245 if it matched the actual IP address. This has been tidied up so that it
3246 doesn't show the same IP address twice.
3248 PH/07 When both +timestamp and +memory debugging was on, the value given by
3249 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3250 called by the debug printing, thereby overwriting the timestamp buffer.
3251 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3253 PH/08 When the original message was included in an autoreply transport, it
3254 always said "this is a copy of the message, including all the headers",
3255 even if body_only or headers_only was set. It now gives an appropriate
3258 PH/09 Applied a patch from the Sieve maintainer which:
3260 o fixes some comments
3261 o adds the (disabled) notify extension core
3262 o adds some debug output for the result of if/elsif tests
3263 o points to the current vacation draft in the documentation
3264 and documents the missing references header update
3268 o fixes a bug in processing the envelope test (when testing
3269 multiple envelope elements, the last element determined the
3272 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3273 Electronic Mail") by including:
3275 Auto-submitted: auto-generated
3277 in the messages that it generates (bounce messages and others, such as
3278 warnings). In the case of bounce messages for non-SMTP messages, there was
3279 also a typo: it was using "Auto_submitted" (underscore instead of
3280 hyphen). Since every message generated by Exim is necessarily in response
3281 to another message, thes have all been changed to:
3283 Auto-Submitted: auto-replied
3285 in accordance with these statements in the RFC:
3287 The auto-replied keyword:
3289 - SHOULD be used on messages sent in direct response to another
3290 message by an automatic process,
3292 - MUST NOT be used on manually-generated messages,
3294 - MAY be used on Delivery Status Notifications (DSNs) and Message
3295 Disposition Notifications (MDNs),
3297 - MUST NOT be used on messages generated by automatic or periodic
3298 processes, except for messages which are automatic responses to
3301 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3302 to the default Received: header definition.
3304 PH/12 Added log selector acl_warn_skipped (default on).
3306 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3307 variables because (a) they are in the wrong storage pool and (b) even if
3308 they were copied, it wouldn't work properly because of the caching.
3310 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3311 checking when decoding. Apparently there are clients that generate
3312 overlong encoded strings. Why am I not surprised?
3314 PH/15 If the first argument of "${if match_address" was not empty, but did not
3315 contain an "@" character, Exim crashed. Now it writes a panic log message
3316 and treats the condition as false.
3318 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3320 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3321 extension "envelope-auth". The code is finished and in agreement with
3322 other implementations, but there is no documentation so far and in fact,
3323 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3324 not changing the active code.
3326 Print executed "if" and "elsif" statements when debugging is used. This
3327 helps a great deal to understand what a filter does.
3329 Document more things not specified clearly in RFC3028. I had all this
3330 sorted out, when out of a sudden new issues came to my mind. Oops."
3332 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3335 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3336 canonical form (as documented). However, after a host name lookup from
3337 the IP address, check_host() was doing a simple string comparison with
3338 addresses acquired from the DNS when checking that the found name did
3339 have the original IP as one of its addresses. Since any found IPv6
3340 addresses are likely to be in abbreviated form, the comparison could
3341 fail. Luckily, there already exists a function for doing the comparison
3342 by converting both addresses to binary, so now that is used instead of
3343 the text comparison.
3345 PH/20 There was another similar case to PH/19, when a complete host name was
3346 given in a host list; looking up its IP address could give an abbreviated
3347 form, whereas the current host's name might or might not be abbreviated.
3348 The same fix has been applied.
3354 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3355 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3358 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3359 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3361 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3363 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3364 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3365 Also, for "Auto-Submitted", treat anything other than "no" as
3366 non-personal, in accordance with RFC 3834. (Previously it treated
3367 anything starting "auto-" as non-personal.)
3369 TF/01 The control=submission/name=... option had a problem with syntax
3370 errors if the name included a slash character. The /name= option
3371 now slurps the rest of the string, so it can include any characters
3372 but it must come last in the list of options (after /sender_retain
3375 PH/05 Some modifications to the interface to the fake nameserver for the new
3383 TK/01 Added the "success_on_redirect" address verification option. See
3384 NewStuff for rationale and an example.
3386 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3388 PH/02 Patch to exigrep to allow it to work on syslog lines.
3390 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3391 fread() to skip over the body file's header line, because in Cygwin the
3392 header line is locked and is inaccessible.
3394 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3395 co-exist for some time) to make it clear that it is the Exim ID that is
3396 referenced, not the Message-ID: header line.
3398 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3399 string_format() function, because snprintf() does not exist on all
3402 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3403 ${stat: expansion item.
3405 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3406 protocol synchronization error", to keep the pedants happy.
3408 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3409 well as for IRIX systems, when gcc is being used. See the host.c source
3412 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3414 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3417 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3418 transport and to the smtp transport in LMTP mode.
3420 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3422 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3423 run a filter in a subprocess. This could lead to confusion in subsequent
3424 lookups in the parent process. There should also be a search_tidyup() at
3425 the end of the subprocess.
3427 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3428 only if the host matched helo_try_verify_hosts, which caused the
3429 verification to occur when the EHLO/HELO command was issued. The ACL just
3430 tested the remembered result. Now, if a previous verification attempt has
3431 not happened, "verify = helo" does it there and then.
3433 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3435 TK/03 Fix log output including CR from clamd.
3437 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3438 exist provoked a memory error which could cause a segfault.
3440 PH/15 Installed PCRE 6.2
3442 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3444 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3445 of the problem. Specifically, suggested +O2 rather than +O1 for the
3448 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3450 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3451 record, the port was not correctly logged when the outgoing_port log
3452 selector was set (it logged the transort's default port).
3454 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3455 fallback_hosts, and "hosts" in the smtp transport.
3457 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3458 host errors such as "Connection refused".
3460 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3461 authentication with radiusclient 0.4.9:
3463 - Error returned from rc_read_config was caught wrongly
3464 - Username/password not passed on to radius server due to wrong length.
3466 The presumption is that some radiusclient API changes for 4.51/PH/17
3467 were not taken care of correctly. The code is still untested by me (my
3468 Linux distribution still has 0.3.2 of radiusclient), but it was
3469 contributed by a Radius user.
3471 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3472 expanding the "port" option of the smtp transport.
3474 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3475 while reading a MIME header. Thanks to Tom Hughes for a patch.
3477 PH/24 Include config.h inside local_scan.h so that configuration settings are
3480 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3481 that in an ACL for RCPT (for example), you can examine exactly what was
3484 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3485 commands, but it was not correctly comparing the address with the actual
3486 client host address. Thus, it would show the EHLO address in Received:
3487 header lines when this was not necessary.
3489 PH/27 Added the % operator to ${eval:}.
3491 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3492 it should be ignoring failures (because with -C, for example, it has lost
3493 privilege). It wasn't ignoring creation failures other than "already
3496 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3499 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3500 because an input line was too long, either on its own, or by virtue of
3501 too many continuations, the temporary file was not being removed, and the
3502 return code was incorrect.
3504 PH/31 Missing "BOOL" in function definition in filtertest.c.
3506 PH/32 Applied Sieve patches from the maintainer.
3508 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3510 PH/33 Added "verify = not_blind".
3512 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3513 Local/Makefile (with some defaults set). These are used in built scripts
3514 such as exicyclog, but they have never been used in the exim_install
3515 script (though there are many overriding facilities there). I have
3516 arranged that the exim_install script now takes note of these two
3519 PH/35 Installed configuration files for Dragonfly.
3521 PH/36 When a locally submitted message by a trusted user did not contain a
3522 From: header, and the sender address was obtained from -f or from an SMTP
3523 MAIL command, and the trusted user did not use -F to supply a sender
3524 name, $originator_name was incorrectly used when constructing a From:
3525 header. Furthermore, $originator_name was used for submission mode
3526 messages from external hosts without From: headers in a similar way,
3527 which is clearly wrong.
3529 PH/37 Added control=suppress_local_fixups.
3531 PH/38 When log_selector = +received_sender was set, and the addition of the
3532 sender made the log line's construction buffer exactly full, or one byte
3533 less than full, an overflow happened when the terminating "\n" was
3536 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3537 when the result of a list match is failure because a DNS lookup failed.
3539 PH/40 RM_COMMAND is now used in the building process.
3541 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3542 the "build-* directories that it finds.
3544 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3545 address in a domain literal was a prefix of an interface address.
3547 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3548 when verifying a sender address, unless rewrite_headers is false.
3550 PH/44 Wrote a long comment about why errors_to addresses are verified as
3551 recipients, not senders.
3553 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3554 the ratelimit ACL was added.
3556 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3558 PH/46 Added extra information about PostgreSQL errors to the error string.
3560 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3561 avoiding the need to install special zones in a real server. This is
3562 backwards compatible; if it can't find the fake resolver, it drops back.
3563 Thus, both old and new test suites can be run.
3565 TF/02 Added util/ratelimit.pl
3567 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3568 clock is set back in time.
3570 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3571 Candler <B.Candler@pobox.com>.
3573 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3574 OK for addresses that are the result of redirections.
3576 PH/48 A number of further additions for the benefit of the new test suite,
3577 including a fake gethostbyname() that interfaces to the fake DNS resolver
3580 TF/06 The fix for widen_domains has also been applied to qualify_single and
3581 search_parents which are the other dnslookup options that can cause
3584 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3587 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3589 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3590 service" if it encountered an MX record with an empty target hostname.
3591 The message is now "an MX or SRV record indicated no SMTP service".
3593 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3594 if the DNS of the sending site is misconfigured. This is quite a
3595 common situation. This change restores the behaviour of treating a
3596 helo verification defer as a failure.
3598 PH/51 If self=fail was set on a router, the bounce message did not include the
3599 actual error message.
3605 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3607 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3608 command itself ended in error, the underlying message about the transport
3609 filter timeout was being overwritten with the pipe command error. Now the
3610 underlying error message should be appended to the second error message.
3612 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3614 PH/02 Reduce the amount of output that "make" produces by default. Full output
3615 can still be requested.
3617 PH/03 The warning log line about a condition test deferring for a "warn" verb
3618 was being output only once per connection, rather than after each
3619 occurrence (because it was using the same function as for successful
3620 "warn" verbs). This seems wrong, so I have changed it.
3622 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3623 it should not have, which might have caused a crash in the right
3624 circumstances, but probably never did.
3626 PH/04 Installed a modified version of Tony Finch's patch to make submission
3627 mode fix the return path as well as the Sender: header line, and to
3628 add a /name= option so that you can make the user's friendly name appear
3631 TF/03 Added the control = fakedefer ACL modifier.
3633 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3634 Mark Lowes for thorough testing.
3636 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3638 TK/03 Merged latest SRS patch from Miles Wilton.
3640 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3641 with the definition in sysexits.h (which is #included earlier).
3642 Fortunately, Exim does not actually use EX_OK. The code used to try to
3643 preserve the sysexits.h value, by assuming that macro definitions were
3644 scanned for macro replacements. I have been disabused of this notion,
3645 so now the code just undefines EX_OK before #including unistd.h.
3647 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3648 in the smtp transport. When a block could not be written in a single
3649 write() function, the timeout was being re-applied to each part-write.
3650 This seems wrong - if the receiver was accepting one byte at a time it
3651 would take for ever. The timeout is now adjusted when this happens. It
3652 doesn't have to be particularly precise.
3654 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3655 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3657 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3658 without a domain if the check to <postmaster@domain> fails.
3660 SC/01 Eximstats: added -xls and the ability to specify output files
3661 (patch written by Frank Heydlauf).
3663 SC/02 Eximstats: use FileHandles for outputting results.
3665 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3667 SC/04 Eximstats: fixed display of large numbers with -nvr option
3669 SC/05 Eximstats: fixed merging of reports with empty tables.
3671 SC/06 Eximstats: added the -include_original_destination flag
3673 SC/07 Eximstats: removed tabs and trailing whitespace.
3675 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3677 TK/06 MBOX spool code: Add real "From " MBOX separator line
3678 so the .eml file is really in mbox format (even though
3679 most programs do not really care). Patch from Alex Miller.
3681 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3682 The latter is generated from $received_to and is only set if the
3683 message has one envelope recipient. SA can use these headers,
3684 obviously out-of-the-box. Patch from Alex Miller.
3686 PH/08 The ${def test on a variable was returning false if the variable's
3687 value was "0", contrary to what the specification has always said!
3688 The result should be true unless the variable is empty.
3690 PH/09 The syntax error of a character other than { following "${if
3691 def:variable_name" (after optional whitespace) was not being diagnosed.
3692 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3693 accidental colon was present, for example, could give incorrect results.
3695 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3696 result is used (not including appendfile, where other changes are about
3699 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3700 This involved changing a lot of size variables from int to off_t. It
3701 should work with maildirs and everything.
3703 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3704 spamd dying while we are connected to it.
3706 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3709 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3710 function for BDB 4.3.
3712 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3714 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3715 the sources for PCRE have been re-arranged and more files are now
3718 PH/15 The code I had for printing potentially long long variables in PH/11
3719 above was not the best (it lost precision). The length of off_t variables
3720 is now inspected at build time, and an appropriate printing format (%ld
3721 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3722 to be "long long int" or "long int". This is needed for the internal
3723 formatting function string_vformat().
3725 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3726 the configuration file to be ":syslog", then the script "guesses" where
3727 the logs files are, rather than using the compiled in default. In our
3728 case the guess is not the same as the compiled default, so the script
3729 suddenly stopped working when I started to use syslog. The patch checks
3730 to see if log_file_path is "". If so, it attempts to read it from exim
3731 with no configuration file to get the compiled in version, before it
3732 falls back to the previous guessing code."
3734 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3735 implementing BATV in an Exim configuration. See NewStuff for the gory
3738 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3739 Makefile that are specific to HP-UX.
3741 PH/18 If the "use_postmaster" option was set for a recipient callout together
3742 with the "random" option, the postmaster address was used as the MAIL
3743 FROM address for the random test, but not for the subsequent recipient
3744 test. It is now used for both.
3746 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3747 patch removes a few documentation additions to RFC 3028, because the
3748 latest draft now contains them. It adds the new en;ascii-case comparator
3749 and a new error check for 8bit text in MIME parts. Comparator and
3750 require names are now matched exactly. I enabled the subaddress
3751 extension, but it is not well tested yet (read: it works for me)."
3753 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3754 rework some of the code of TK/09 above to avoid the hardwired use of
3755 "%lld" and "long long". Replaced the call to snprintf() with a call to
3758 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3759 records point to non-existent hosts", "retry timeout exceeded", and
3760 "retry time not reached for any host after a long failure period".
3762 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3763 experimental DomainKeys support:
3765 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3766 (2) On an error such as an illegally used "control", the wrong name for
3767 the control was given.
3769 These problems did NOT occur unless DomainKeys support was compiled.
3771 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3773 PH/24 Added ${if match_ip condition.
3775 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3776 databases so that it will be absolutely obvious if a crash occurs in the
3777 DB library. This is a regular occurrence (often caused by mis-matched
3780 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3781 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3782 picked up on a user's system that detects such things. There doesn't seem
3783 to be a gcc warning option for this - only an attribute that has to be
3784 put on the function's prototype. It seems that in Fedora Core 4 they have
3785 set this on a number of new functions. No doubt there will be more in due
3788 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3789 specify a transport. However, if an address that was verified by such a
3790 router was the subject of a callout, Exim crashed because it tried to
3791 read the rcpt_include_affixes from the non-existent transport. Now it
3792 just assumes that the setting of that option is false. This bug was
3793 introduced by 4.51/PH/31.
3795 PH/28 Changed -d+all to exclude +memory, because that information is very
3796 rarely of interest, but it makes the output a lot bigger. People tend to
3797 do -d+all out of habit.
3799 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3800 code in os-type was giving problems when libc.so lives in lib64, like on
3803 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3804 aren't the modern standard, and it seems that some systems' include files
3805 don't always have them. Exim was already checking for some of the newer
3806 ones like T_AAAA, and defining it itself. I've added checks for all the
3807 record types that Exim uses.
3809 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3810 not automatically generating a new one, as it is supposed to. This
3811 prevented TLS from working. If the file did exist, but contained invalid
3812 data, a new version was generated, as expected. It was only the case of a
3813 non-existent file that was broken.
3815 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3816 with a change in libdomainkeys > 0.64.
3818 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3819 from DNS. If the selector record carries the flag, it now has
3820 precedence over the domain-wide flag.
3822 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3824 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3825 the use of a port name that isn't defined in /etc/services) occurred, the
3826 message was deferred as in a normal delivery, and thus remained on the
3827 spool, instead of being failed because of the mua_wrapper setting. This
3828 is now fixed, and I tidied up some of the mua_wrapper messages at the
3831 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3832 the messages in a hash of arrays rather than using individual hashes.
3833 This is a bit cleaner and results in dramatic memory savings, albeit
3834 at a slight CPU cost.
3836 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3837 as requested by Marc Sherman.
3839 SC/10 Eximstats: added histograms for user specified patterns as requested
3842 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3844 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3845 fopen() in the content-scanning modules that did not already have it.
3851 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3852 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3854 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3856 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3858 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3859 to test Sieve filters that use "vacation".
3861 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3862 that changes the way the GnuTLS parameters are stored in the cache file.
3863 The new format can be generated externally. For backward compatibility,
3864 if the data in the cache doesn't make sense, Exim assumes it has read an
3865 old-format file, and it generates new data and writes a new file. This
3866 means that you can't go back to an older release without removing the
3869 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3870 work if there are any delivery delays because "one_time" forces the
3871 parent to be marked "delivered", so its unseen clone is never tried
3872 again. For this reason, Exim now forbids the simultaneous setting of
3875 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3876 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3877 entry if you want to know the details. The fix, however, neglected to
3878 consider the case when local delivery batching is involved. The test for
3879 "previously delivered" was not happening when checking to see if an
3880 address could be batched with a previous (undelivered) one; under
3881 certain circumstances this could lead to multiple deliveries to the same
3884 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3885 in its include files, and this causes problems building Exim.
3887 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3888 header_syntax) but Exim was just ignoring anything given after a slash.
3889 In particular, this caused confusion with an attempt to use "verify =
3890 reverse_host_lookup/defer_ok". An error is now given when options are
3891 supplied for verify items that do not have them. (Maybe reverse_host_
3892 lookup should have a defer_ok option, but that's a different point.)
3894 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3895 defined by RFC 821) to 2048, because there were problems with some AUTH
3896 commands, and RFC 1869 says the size should be increased for extended
3897 SMTP commands that take arguments.
3899 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3902 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3903 "unknown" error; now it says that the functionality isn't in the binary.
3905 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3906 an address' error message when a string expansion fails (syntax or
3907 whatever). Otherwise the password may appear in the log. Following change
3908 PH/42 below, there is no longer a chance of it appearing in a bounce
3911 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3913 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3914 of its addresses was ever tried. (Bugzilla bug #2.)
3916 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3917 the result incorrectly in the debug output. (It correctly added a newline
3918 to what was transported.)
3920 TF/01 Added $received_time.
3922 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3923 commented out examples of how to interface to a virus scanner and to
3924 SpamAssassin. Also added commented examples of av_scanner and
3925 spamd_address settings.
3927 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3928 and controls are allowed in which ACLs. There were a couple of minor
3929 errors. Some of the entries in the conditions table (which is a table of
3930 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3931 negation of where the condition IS allowed.
3933 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3935 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3936 header file does not have a version number, so I've had to invent a new
3937 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3938 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3939 radiusclient), but it was contributed by a Radius user.
3941 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3942 files or directories, for OpenSSL.
3944 PH/19 When an Exim process that is running as root has to create an Exim log
3945 file, it does so in a subprocess that runs as exim:exim so as to get the
3946 ownership right at creation (otherwise, other Exim processes might see
3947 the file with the wrong ownership). There was no test for failure of this
3948 fork() call, which would lead to the process getting stuck as it waited
3949 for a non-existent subprocess. Forks do occasionally fail when resources
3950 run out. I reviewed all the other calls to fork(); they all seem to check
3953 PH/20 When checking for unexpected SMTP input at connect time (before writing
3954 the banner), Exim was not dealing correctly with a non-positive return
3955 from the read() function. If the client had disconnected by this time,
3956 the result was a log entry for a synchronization error with an empty
3957 string after "input=" when read() returned zero. If read() returned -1
3958 (an event I could not check), uninitialized data bytes were printed.
3959 There were reports of junk text (parts of files, etc) appearing after
3962 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3964 PH/22 Added support for macro redefinition, and (re)definition in between
3965 driver and ACL definitions.
3967 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3968 forgetting to use the resulting value; it was using the unexpanded value.
3970 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3971 hadn't been configured. The fix is from Juergen Kreileder, who
3972 understands it better than I do:
3974 "Here's what I see happening with three configured cyrus_sasl
3975 authenticators configured (plain, login, cram-md5):
3977 On startup auth_cyrus_sasl_init() gets called for each of these.
3978 This means three calls to sasl_listmech() without a specified mech_list.
3979 => SASL tests which mechs of all available mechs actually work
3980 => three warnings about OTP not working
3981 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3983 With the patch, sasl_listmech() also gets called three times. But now
3984 SASL's mech_list option is set to the server_mech specified in the the
3985 authenticator. Or in other words, the answer from sasl_listmech()
3986 gets limited to just the mech you're testing for (which is different
3988 => the return list contains just 'plain' or 'login', 'cram-md5' or
3989 nothing depending on the value of ob->server_mech.
3991 I've just tested the patch: Authentication still works fine,
3992 unavailable mechs specified in the exim configuration are still
3993 caught, and the auth.log warnings about OTP are gone."
3995 PH/25 When debugging is enabled, the contents of the command line are added
3996 to the debugging output, even when log_selector=+arguments is not
3999 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4000 answer is "GNU", and only if the return is "GNU/something" is the answer
4003 PH/27 $acl_verify_message is now set immediately after the failure of a
4004 verification in an ACL, and so is available in subsequent modifiers. In
4005 particular, the message can be preserved by coding like this:
4007 warn !verify = sender
4008 set acl_m0 = $acl_verify_message
4010 Previously, $acl_verify_message was set only while expanding "message"
4011 and "log_message" when a very denied access.
4013 PH/28 Modified OS/os.c-Linux with
4015 -#ifndef OS_LOAD_AVERAGE
4016 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4018 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4019 nomenclature these days.)
4021 PH/29 Installed patch from the Sieve maintainer that adds the options
4022 sieve_useraddress and sieve_subaddress to the redirect router.
4024 PH/30 In these circumstances:
4025 . Two addresses routed to the same list of hosts;
4026 . First host does not offer TLS;
4027 . First host accepts first address;
4028 . First host gives temporary error to second address;
4029 . Second host offers TLS and a TLS session is established;
4030 . Second host accepts second address.
4031 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4032 and peerdn, if requested) that were in fact used only for the second
4035 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4036 attention to any local part prefix or suffix that was matched by the
4037 router that accepted the address. It now behaves in the same way as it
4038 does for delivery: the affixes are removed from the local part unless
4039 rcpt_include_affixes is set on the transport.
4041 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4042 timeout during the DATA phase of an incoming message.
4044 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4045 applied a patch sanctioned by the Sieve maintainer.
4047 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4048 the uid or gid is negative. A case of a negative gid caused this to be
4049 noticed. The fix allows for either to be negative.
4051 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4052 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4055 PH/36 The change PH/12 above was broken. Fixed it.
4057 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4058 the grounds that routing the same address twice would always produce the
4059 same answer. This might have been true once, but it is certainly no
4060 longer true now. Routing a child address may depend on the previous
4061 routing that produced that child. Some complicated redirection strategies
4062 went wrong when messages had multiple recipients, and made Exim's
4063 behaviour dependent on the order in which the addresses were given.
4065 I have moved the duplicate checking until after the routing is complete.
4066 Exim scans the addresses that are assigned to local and remote
4067 transports, and removes any duplicates. This means that more work will be
4068 done, as duplicates will always all be routed, but duplicates are
4069 presumably rare, so I don't expect this is of any significance.
4071 For deliveries to pipes, files, and autoreplies, the duplicate checking
4072 still happens during the routing process, since they are not going to be
4075 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4076 It corrects a timeout issue with spamd. This is Ian's comment: "The
4077 background is that sometimes spamd either never reads data from a
4078 connection it has accepted, or it never writes response data. The exiscan
4079 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4080 blindly assumes that writes won't block so it may never time out."
4082 PH/39 Allow G after quota size as well as K and M.
4084 PH/40 The value set for $authenticated_id in an authenticator may not contain
4085 binary zeroes or newlines because the value is written to log lines and
4086 to spool files. There was no check on this. Now the value is run through
4087 the string_printing() function so that such characters are converted to
4088 printable escape sequences.
4090 PH/41 $message_linecount is a new variable that contains the total number of
4091 lines in the message. Compare $body_linecount, which is the count for the
4094 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4095 bounce and delay warning messages, except in certain special cases, which
4098 (a) An SMTP error message from a remote host;
4099 (b) A message specified in a :fail: redirection;
4100 (c) A message specified in a "fail" command in a system filter;
4101 (d) A message specified in a FAIL return from the queryprogram router;
4102 (e) A message specified by the cannot_route_message router option.
4104 In these cases only, Exim does include the error details in bounce and
4105 warning messages. There are also a few cases where bland messages such
4106 as "unrouteable address" or "local delivery error" are given.
4108 PH/43 $value is now also set for the "else" part of a ${run expansion.
4110 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4111 being worked on, but at least Exim now implements the latest version to
4114 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4115 process to complete was treated as a delivery failure, a timeout while
4116 writing the message to the pipe was logged, but erroneously treated as a
4117 successful delivery. Such timeouts include transport filter timeouts. For
4118 consistency with the overall process timeout, these timeouts are now
4119 treated as errors, giving rise to delivery failures by default. However,
4120 there is now a new Boolean option for the pipe transport called
4121 timeout_defer, which, if set TRUE, converts the failures into defers for
4122 both kinds of timeout. A transport filter timeout is now identified in
4125 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4126 systems where "make" and "gmake" are different, calling "gmake" at top
4127 level broke things. I've arranged for the value of $(MAKE) to be passed
4128 from the Makefile to this script so that it can call the same version of
4132 A note about Exim versions 4.44 and 4.50
4133 ----------------------------------------
4135 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4136 changes of various kinds. As a consequence, a big documentation update was
4137 needed. This delayed the release for rather longer than seemed good, especially
4138 in the light of a couple of (minor) security issues. Therefore, the changes
4139 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4140 release. So 4.44 and 4.50 are in effect two different branches that both start
4143 I have left the 4.50 change log unchanged; it contains all the changes since
4144 4.43. The change log for 4.44 is below; many of its items are identical to
4145 those for 4.50. This seems to be the most sensible way to preserve the
4146 historical information.
4152 1. Minor wording change to the doc/README.SIEVE file.
4154 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4155 computation of the current number of files was incorrect.
4157 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4158 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4161 4. Give more explanation in the error message when the command for a transport
4162 filter fails to execute.
4164 5. There are several places where Exim runs a non-Exim command in a
4165 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4166 was being done only for the command run by the queryprogram router. It is
4167 now done for all such subprocesses. The other cases are: ${run, transport
4168 filters, and the commands run by the lmtp and pipe transports.
4170 6. Added CONFIGURE_GROUP build-time option.
4172 7. Some older OS have a limit of 256 on the maximum number of file
4173 descriptors. Exim was using setrlimit() to set 1000 as a large value
4174 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4175 systems. I've change it so that if it can't get 1000, it tries for 256.
4177 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4178 was an oversight, and furthermore, ever since the addition of extra
4179 controls (e.g. 4.43/32), the checks on when to allow different forms of
4180 "control" were broken. There should now be diagnostics for all cases when a
4181 control that does not make sense is encountered.
4183 9. Added the /retain_sender option to "control=submission".
4185 10. $recipients is now available in the predata ACL (oversight).
4187 11. Tidy the search cache before the fork to do a delivery from a message
4188 received from the command line. Otherwise the child will trigger a lookup
4189 failure and thereby defer the delivery if it tries to use (for example) a
4190 cached ldap connection that the parent has called unbind on.
4192 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4193 of $address_data from the recipient verification was clobbered by the
4194 sender verification.
4196 13. The value of address_data from a sender verification is now available in
4197 $sender_address_data in subsequent conditions in the ACL statement.
4199 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4201 15. Added a new option "connect=<time>" to callout options, to set a different
4204 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4205 was its contents. (It was OK if the option was not defined at all.)
4207 17. A "Completed" log line is now written for messages that are removed from
4208 the spool by the -Mrm option.
4210 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4211 information about exactly what failed.
4213 19. Added -dd to debug only the daemon process.
4215 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4216 handles timeouts, both on the server side and network timeouts. Renamed the
4217 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4219 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4220 It is now set to "smtps".
4222 22. $host_address is now set to the target address during the checking of
4223 ignore_target_hosts.
4225 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4226 being passed; this would have caused $sender_host_name to have been used if
4227 matching the list had actually called for a host name (not very likely,
4228 since this list is usually IP addresses). A host name is now passed as
4231 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4232 code that specifies a non-restarting handler (typically sigaction() in
4233 modern systems) in an attempt to fix a rare and obscure crash bug.
4235 25. Narrowed the window for a race in the daemon that could cause it to ignore
4236 SIGCHLD signals. This is not a major problem, because they are used only to
4237 wake it up if nothing else does.
4239 26. A malformed maildirsize file could cause Exim to calculate negative values
4240 for the mailbox size or file count. Odd effects could occur as a result.
4241 The maildirsize information is now recalculated if the size or filecount
4244 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4245 support for a long time. Removed HAVE_SYS_VFS_H.
4247 28. Installed the latest version of exipick from John Jetmore.
4249 29. In an address list, if the pattern was not a regular expression, an empty
4250 subject address (from a bounce message) matched only if the pattern was an
4251 empty string. Non-empty patterns were not even tested. This was the wrong
4252 because it is perfectly reasonable to use an empty address as part of a
4253 database query. An empty address is now tested by patterns that are
4254 lookups. However, all the other forms of pattern expect the subject to
4255 contain a local part and a domain, and therefore, for them, an empty
4256 address still always fails if the pattern is not itself empty.
4258 30. Exim went into a mad DNS loop when attempting to do a callout where the
4259 host was specified on an smtp transport, and looking it up yielded more
4260 than one IP address.
4262 31. Re-factored the code for checking spool and log partition space into a
4263 function that finds that data and another that does the check. The former
4264 is then used to implement four new variables: $spool_space, $log_space,
4265 $spool_inodes, and $log_inodes.
4267 32. The RFC2047 encoding function was originally intended for short strings
4268 such as real names; it was not keeping to the 75-character limit for
4269 encoded words that the RFC imposes. It now respects the limit, and
4270 generates multiple encoded words if necessary. To be on the safe side, I
4271 have increased the buffer size for the ${rfc2047: expansion operator from
4274 33. It is now permitted to omit both strings after an "if" condition; if the
4275 condition is true, the result is "true". As before, when the second string
4276 is omitted, a false condition yields an empty string. This makes it less
4277 cumbersome to write custom ACL and router conditions.
4279 34. Failure to deliver a bounce message always caused it to be frozen, even if
4280 there was an errors_to setting on the router. The errors_to setting is now
4283 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4284 canonical form (fully expanded) before being placed in
4285 $sender_host_address.
4287 36. The table in the code that translates DNS record types into text (T_A to
4288 "A" for instance) was missing entries for NS and CNAME. It is just possible
4289 that this could have caused confusion if both these types were looked up
4290 for the same domain, because the text type is used as part of Exim's
4291 per-process caching. But the chance of anyone hitting this buglet seems
4294 37. The dnsdb lookup has been extended in a number of ways.
4296 (1) There is a new type, "zns", which walks up the domain tree until it
4297 finds some nameserver records. It should be used with care.
4299 (2) There is a new type, "mxh", which is like "mx" except that it returns
4300 just the host names, not the priorities.
4302 (3) It is now possible to give a list of domains (or IP addresses) to be
4303 looked up. The behaviour when one of the lookups defers can be
4304 controlled by a keyword.
4306 (4) It is now possible to specify the separator character for use when
4307 multiple records are returned.
4309 38. The dnslists ACL condition has been extended: it is now possible to supply
4310 a list of IP addresses and/or domains to be looked up in a particular DNS
4313 39. Added log_selector=+queue_time_overall.
4315 40. When running the queue in the test harness, wait just a tad after forking a
4316 delivery process, to get repeatability of debugging output.
4318 41. Include certificate and key file names in error message when GnuTLS fails
4319 to set them up, because the GnuTLS error message doesn't include the name
4320 of the failing file when there is a problem reading it.
4322 42. Allow both -bf and -bF in the same test run.
4324 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4326 44. The "Exiscan patch" is now merged into the mainline Exim source.
4328 45. Sometimes the final signoff response after QUIT could fail to get
4329 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4330 < 0 before doing a fflush(). This bug looks as though it goes back to the
4331 introduction of TLS in release 3.20, but "sometimes" must have been rare
4332 because the tests only now provoked it.
4334 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4335 (this can affect the format of dates).
4337 47. exim_tidydb, when checking for the continued existence of a message for
4338 which it has found a message-specific retry record, was not finding
4339 messages that were in split spool directories. Consequently, it was
4340 deleting retry records that should have stayed in existence.
4342 48. Steve fixed some bugs in eximstats.
4344 49. The SPA authentication driver was not abandoning authentication and moving
4345 on to the next authenticator when an expansion was forced to fail,
4346 contradicting the general specification for all authenticators. Instead it
4347 was generating a temporary error. It now behaves as specified.
4349 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4350 (the order specifies the preference for clients). The order is now AES256,
4351 AES128, 3DES, ARCFOUR128.
4353 51. Small patch to Sieve code - explicitly set From: when generating an
4356 52. Exim crashed if a remote delivery caused a very long error message to be
4357 recorded - for instance if somebody sent an entire SpamAssassin report back
4358 as a large number of 550 error lines. This bug was coincidentally fixed by
4359 increasing the size of one of Exim's internal buffers (big_buffer) that
4360 happened as part of the Exiscan merge. However, to be on the safe side, I
4361 have made the code more robust (and fixed the comments that describe what
4364 53. Now that there can be additional text after "Completed" in log lines (if
4365 the queue_time_overall log selector is set), a one-byte patch to exigrep
4366 was needed to allow it to recognize "Completed" as not the last thing in
4369 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4370 patch that reportedly fixes this has been added. I am not expert enough to
4371 create a test for it. This is what the patch creator wrote:
4373 "I found a little strange behaviour of ldap code when working with
4374 Windows 2003 AD Domain, where users was placed in more than one
4375 Organization Units. When I tried to give exim partial DN, the exit code
4376 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4377 But simultaneously result of request was absolutely normal ldap result,
4378 so I produce this patch..."
4380 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4381 REFERENCE, so I have modified the code to exclude the patch when that macro
4384 55. Some experimental protocols are using DNS PTR records for new purposes. The
4385 keys for these records are domain names, not reversed IP addresses. The
4386 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4387 leaves it alone. Component reversal etc. now happens only for IP addresses.
4390 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4392 57. Double the size of the debug message buffer (to 2048) so that more of very
4393 long debug lines gets shown.
4395 58. The exicyclog utility now does better if the number of log files to keep
4396 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4398 59. Two changes related to the smtp_active_hostname option:
4400 (1) $smtp_active_hostname is now available as a variable.
4401 (2) The default for smtp_banner uses $smtp_active_hostname instead
4402 of $primary_hostname.
4404 60. The host_aton() function is supposed to be passed a string that is known
4405 to be a valid IP address. However, in the case of IPv6 addresses, it was
4406 not checking this. This is a hostage to fortune. Exim now panics and dies
4407 if the condition is not met. A case was found where this could be provoked
4408 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4409 components; fortuitously, this particular loophole had already been fixed
4410 by change 4.50/55 above.
4412 If there are any other similar loopholes, the new check in host_aton()
4413 itself should stop them being exploited. The report I received stated that
4414 data on the command line could provoke the exploit when Exim was running as
4415 exim, but did not say which command line option was involved. All I could
4416 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4417 running as the user.
4420 61. There was a buffer overflow vulnerability in the SPA authentication code
4421 (which came originally from the Samba project). I have added a test to the
4422 spa_base64_to_bits() function which I hope fixes it.
4425 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4426 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4428 63. The daemon start-up calls getloadavg() while still root for those OS that
4429 need the first call to be done as root, but it missed one case: when
4430 deliver_queue_load_max is set with deliver_drop_privilege. This is
4431 necessary for the benefit of the queue runner, because there is no re-exec
4432 when deliver_drop_privilege is set.
4434 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4435 This has been fixed.
4437 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4438 was in use, was not putting the data itself into the right store pool;
4439 consequently, it could be overwritten for a subsequent message in the same
4440 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4443 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4445 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4446 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4447 the function were treating the return as a boolean value, which happened to
4448 work because 0=false and not-0=true, but is not correct code.
4450 68. The host_aton() function was not handling scoped IPv6 addresses (those
4451 with, for example, "%eth0" on the end) correctly.
4453 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4454 negated items (that is, ~something) in unsigned ints. Some compilers
4455 apparently mutter when there is no cast.
4457 70. If an address verification called from an ACL failed, and did not produce a
4458 user-specific message (i.e. there was only a "system" message), nothing was
4459 put in $acl_verify_message. In this situation, it now puts the system
4462 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4463 session; change 4.31/43 added the unwanted input to the log line - except
4464 that it did not do this in the start of session case. It now does.
4466 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4467 This is wrong and can cause the other end to generate a synchronization
4468 error if it is another Exim or anything else that does the synchronization
4469 check. A QUIT command is no longer sent after a timeout.
4471 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4472 during host lookups.
4474 74. The defer_ok option of callout verification was not working if it was used
4475 when verifying addresses in header lines, that is, for this case:
4477 verify = header_sender/callout=defer_ok
4479 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4480 those file descriptors could be used for SMTP connections. If anything
4481 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4482 could be sent to the SMTP client, causing chaos. The daemon now opens
4483 stdin, stdout, and stderr to /dev/null when it puts itself into the
4486 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4487 log by default. The user can override this with suitable Perl magic.
4489 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4490 the log message when discard triggers, was not working for the DATA ACL or
4491 for the non-SMTP ACL.
4493 78. Error message wording change in sieve.c.
4495 79. If smtp_accept_max_per_host was set, the number of connections could be
4496 restricted to fewer than expected, because the daemon was trying to set up
4497 a new connection before checking whether the processes handling previous
4498 connections had finished. The check for completed processes is now done
4499 earlier. On busy systems, this bug wouldn't be noticed because something
4500 else would have woken the daemon, and it would have reaped the completed
4503 80. If a message was submitted locally by a user whose login name contained one
4504 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4505 It caused a spool format error. I have fixed the spool reading code. A
4506 related problem was that the "from" clause in the Received: line became
4507 illegal because of the space(s). It is now covered by ${quote_local_part.
4509 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4512 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4514 83. Patch from Sieve maintainer to fix unterminated string problem in
4515 "vacation" handling.
4517 84. Some minor changes to the Linux configuration files to help with other
4518 OS variants using glibc.
4520 85. One more patch for Sieve to update vacation handling to latest spec.
4523 ----------------------------------------------------
4524 See the note above about the 4.44 and 4.50 releases.
4525 ----------------------------------------------------
4531 1. Change 4.43/35 introduced a bug that caused file counts to be
4532 incorrectly computed when quota_filecount was set in an appendfile
4535 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4536 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4539 3. Give more explanation in the error message when the command for a transport
4540 filter fails to execute.
4542 4. There are several places where Exim runs a non-Exim command in a
4543 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4544 was being done only for the command run by the queryprogram router. It is
4545 now done for all such subprocesses. The other cases are: ${run, transport
4546 filters, and the commands run by the lmtp and pipe transports.
4548 5. Some older OS have a limit of 256 on the maximum number of file
4549 descriptors. Exim was using setrlimit() to set 1000 as a large value
4550 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4551 systems. I've change it so that if it can't get 1000, it tries for 256.
4553 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4554 was an oversight, and furthermore, ever since the addition of extra
4555 controls (e.g. 4.43/32), the checks on when to allow different forms of
4556 "control" were broken. There should now be diagnostics for all cases when a
4557 control that does not make sense is encountered.
4559 7. $recipients is now available in the predata ACL (oversight).
4561 8. Tidy the search cache before the fork to do a delivery from a message
4562 received from the command line. Otherwise the child will trigger a lookup
4563 failure and thereby defer the delivery if it tries to use (for example) a
4564 cached ldap connection that the parent has called unbind on.
4566 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4567 of $address_data from the recipient verification was clobbered by the
4568 sender verification.
4570 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4571 was its contents. (It was OK if the option was not defined at all.)
4573 11. A "Completed" log line is now written for messages that are removed from
4574 the spool by the -Mrm option.
4576 12. $host_address is now set to the target address during the checking of
4577 ignore_target_hosts.
4579 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4580 being passed; this would have caused $sender_host_name to have been used if
4581 matching the list had actually called for a host name (not very likely,
4582 since this list is usually IP addresses). A host name is now passed as
4585 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4586 code that specifies a non-restarting handler (typically sigaction() in
4587 modern systems) in an attempt to fix a rare and obscure crash bug.
4589 15. Narrowed the window for a race in the daemon that could cause it to ignore
4590 SIGCHLD signals. This is not a major problem, because they are used only to
4591 wake it up if nothing else does.
4593 16. A malformed maildirsize file could cause Exim to calculate negative values
4594 for the mailbox size or file count. Odd effects could occur as a result.
4595 The maildirsize information is now recalculated if the size or filecount
4598 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4599 support for a long time. Removed HAVE_SYS_VFS_H.
4601 18. Updated exipick to current release from John Jetmore.
4603 19. Allow an empty sender to be matched against a lookup in an address list.
4604 Previously the only cases considered were a regular expression, or an
4607 20. Exim went into a mad DNS lookup loop when doing a callout where the
4608 host was specified on the transport, if the DNS lookup yielded more than
4611 21. The RFC2047 encoding function was originally intended for short strings
4612 such as real names; it was not keeping to the 75-character limit for
4613 encoded words that the RFC imposes. It now respects the limit, and
4614 generates multiple encoded words if necessary. To be on the safe side, I
4615 have increased the buffer size for the ${rfc2047: expansion operator from
4618 22. Failure to deliver a bounce message always caused it to be frozen, even if
4619 there was an errors_to setting on the router. The errors_to setting is now
4622 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4623 canonical form (fully expanded) before being placed in
4624 $sender_host_address.
4626 24. Updated eximstats to version 1.33
4628 25. Include certificate and key file names in error message when GnuTLS fails
4629 to set them up, because the GnuTLS error message doesn't include the name
4630 of the failing file when there is a problem reading it.
4632 26. Expand error message when OpenSSL has problems setting up cert/key files.
4635 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4636 (this can affect the format of dates).
4638 28. exim_tidydb, when checking for the continued existence of a message for
4639 which it has found a message-specific retry record, was not finding
4640 messages that were in split spool directories. Consequently, it was
4641 deleting retry records that should have stayed in existence.
4643 29. eximstats updated to version 1.35
4644 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4645 1.35 - bugfix such that pie charts by volume are generated correctly
4647 30. The SPA authentication driver was not abandoning authentication and moving
4648 on to the next authenticator when an expansion was forced to fail,
4649 contradicting the general specification for all authenticators. Instead it
4650 was generating a temporary error. It now behaves as specified.
4652 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4653 (the order specifies the preference for clients). The order is now AES256,
4654 AES128, 3DES, ARCFOUR128.
4656 31. Small patch to Sieve code - explicitly set From: when generating an
4659 32. Exim crashed if a remote delivery caused a very long error message to be
4660 recorded - for instance if somebody sent an entire SpamAssassin report back
4661 as a large number of 550 error lines. This bug was coincidentally fixed by
4662 increasing the size of one of Exim's internal buffers (big_buffer) that
4663 happened as part of the Exiscan merge. However, to be on the safe side, I
4664 have made the code more robust (and fixed the comments that describe what
4667 33. Some experimental protocols are using DNS PTR records for new purposes. The
4668 keys for these records are domain names, not reversed IP addresses. The
4669 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4670 leaves it alone. Component reversal etc. now happens only for IP addresses.
4673 34. The host_aton() function is supposed to be passed a string that is known
4674 to be a valid IP address. However, in the case of IPv6 addresses, it was
4675 not checking this. This is a hostage to fortune. Exim now panics and dies
4676 if the condition is not met. A case was found where this could be provoked
4677 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4678 components; fortuitously, this particular loophole had already been fixed
4679 by change 4.50/55 or 4.44/33 above.
4681 If there are any other similar loopholes, the new check in host_aton()
4682 itself should stop them being exploited. The report I received stated that
4683 data on the command line could provoke the exploit when Exim was running as
4684 exim, but did not say which command line option was involved. All I could
4685 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4686 running as the user.
4689 35. There was a buffer overflow vulnerability in the SPA authentication code
4690 (which came originally from the Samba project). I have added a test to the
4691 spa_base64_to_bits() function which I hope fixes it.
4694 36. The daemon start-up calls getloadavg() while still root for those OS that
4695 need the first call to be done as root, but it missed one case: when
4696 deliver_queue_load_max is set with deliver_drop_privilege. This is
4697 necessary for the benefit of the queue runner, because there is no re-exec
4698 when deliver_drop_privilege is set.
4700 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4701 was in use, was not putting the data itself into the right store pool;
4702 consequently, it could be overwritten for a subsequent message in the same
4703 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4706 38. Sometimes the final signoff response after QUIT could fail to get
4707 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4708 < 0 before doing a fflush(). This bug looks as though it goes back to the
4709 introduction of TLS in release 3.20, but "sometimes" must have been rare
4710 because the tests only now provoked it.
4716 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4717 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4718 Now it returns an integer. A number of calls were still expecting a T/F
4719 return. Fortuitously, in all cases, the tests worked in OK situations,
4720 which is the norm. However, things would have gone wrong on any write
4721 failures on the smtp file descriptor. This function is used when sending
4722 messages over SMTP and also when doing verify callouts.
4724 2. When Exim is called to do synchronous delivery of a locally submitted
4725 message (the -odf or -odi options), it no longer closes stderr before doing
4728 3. Implemented the mua_wrapper option.
4730 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4732 5. Implemented the functions header_remove(), header_testname(),
4733 header_add_at_position(), and receive_remove_recipient(), and exported them
4736 6. If an ACL "warn" statement specified the addition of headers, Exim already
4737 inserted X-ACL-Warn: at the start if there was no header name. However, it
4738 was not making this test for the second and subsequent header lines if
4739 there were newlines in the string. This meant that an invalid header could
4740 be inserted if Exim was badly configured.
4742 7. Allow an ACL "warn" statement to add header lines at the start or after all
4743 the Received: headers, as well as at the end.
4745 8. Added the rcpt_4xx retry error code.
4747 9. Added postmaster_mailfrom=xxx to callout verification option.
4749 10. Added mailfrom=xxxx to the callout verification option, for verify=
4752 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4753 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4755 12. Inserted some casts to stop certain compilers warning when using pointer
4756 differences as field lengths or precisions in printf-type calls (mostly
4757 affecting debugging statements).
4759 13. Added optional readline() support for -be (dynamically loaded).
4761 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4762 same clock tick as a message's arrival, so that its received time was the
4763 same as the "first fail" time on the retry record, and that message
4764 remained on the queue past the ultimate address timeout, every queue runner
4765 would try a delivery (because it was past the ultimate address timeout) but
4766 after another failure, the ultimate address timeout, which should have then
4767 bounced the address, did not kick in. This was a "< instead of <=" error;
4768 in most cases the first failure would have been in the next clock tick
4769 after the received time, and all would be well.
4771 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4772 being recognized when the domain list was tested by the match_domain
4773 condition in an expansion string.
4775 16. Added the ${str2b64: operator.
4777 17. Exim was always calling setrlimit() to set a large limit for the number of
4778 processes, without checking whether the existing limit was already
4779 adequate. (It did check for the limit on file descriptors.) Furthermore,
4780 errors from getrlimit() and setrlimit() were being ignored. Now they are
4781 logged to the main and panic logs, but Exim does carry on, to try to do its
4782 job under whatever limits there are.
4784 18. Imported PCRE 5.0.
4786 19. Trivial typo in log message " temporarily refused connection" (the leading
4789 20. If the log selector return_path_on_delivery was set and an address was
4790 redirected to /dev/null, the delivery process crashed because it assumed
4791 that a return path would always be set for a "successful" delivery. In this
4792 case, the whole delivery is bypassed as an optimization, and therefore no
4795 21. Internal re-arrangement: the function for sending a challenge and reading
4796 a response while authentication was assuming a zero-terminated challenge
4797 string. It's now changed to take a pointer and a length, to allow for
4798 binary data in such strings.
4800 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4802 23. Exim was not respecting finduser_retries when seeking the login of the
4803 uid under which it was called; it was always trying 10 times. (The default
4804 setting of finduser_retries is zero.) Also, it was sleeping after the final
4805 failure, which is pointless.
4807 24. Implemented tls_on_connect_ports.
4809 25. Implemented acl_smtp_predata.
4811 26. If the domain in control=submission is set empty, Exim assumes that the
4812 authenticated id is a complete email address when it generates From: or
4813 Sender: header lines.
4815 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4816 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4817 chown and chgrp in /bin and hostname in /usr/bin.
4819 28. Exim was keeping the "process log" file open after each use, just as it
4820 does for the main log. This opens the possibility of it remaining open for
4821 long periods when the USR1 signal hits a daemon. Occasional processlog
4822 errors were reported, that could have been caused by this. Anyway, it seems
4823 much more sensible not to leave this file open at all, so that is what now
4826 29. The long-running daemon process does not normally write to the log once it
4827 has entered its main loop, and it closes the log before doing so. This is
4828 so that log files can straightforwardly be renamed and moved. However,
4829 there are a couple of unusual error situations where the daemon does write
4830 log entries, and I had neglected to close the log afterwards.
4832 30. The text of an SMTP error response that was received during a remote
4833 delivery was being truncated at 512 bytes. This is too short for some of
4834 the long messages that one sometimes sees. I've increased the limit to
4837 31. It is now possible to make retry rules that apply only when a message has a
4838 specific sender, in particular, an empty sender.
4840 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4841 it possible to be selective about when SMTP synchronization is enforced.
4843 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4845 32. Implemented hosts_connection_nolog.
4847 33. Added an ACL for QUIT.
4849 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4852 35. Added mailbox_size and mailbox_filecount to appendfile.
4854 36. Added control = no_multiline_responses to ACLs.
4856 37. There was a bug in the logic of the code that waits for the clock to tick
4857 in the case where the clock went backwards by a substantial amount such
4858 that the microsecond fraction of "now" was more than the microsecond
4859 fraction of "then" (but the whole seconds number was less).
4861 38. Added support for the libradius Radius client library this is found on
4862 FreeBSD (previously only the radiusclient library was supported).
4868 1. When certain lookups returned multiple values in the form name=value, the
4869 quoting of the values was not always being done properly. Specifically:
4870 (a) If the value started with a double quote, but contained no whitespace,
4872 (b) If the value contained whitespace other than a space character (i.e.
4873 tabs or newlines or carriage returns) it was not quoted.
4874 This fix has been applied to the mysql and pgsql lookups by writing a
4875 separate quoting function and calling it from the lookup code. The fix
4876 should probably also be applied to nisplus, ibase and oracle lookups, but
4877 since I cannot test any of those, I have not disturbed their existing code.
4879 2. A hit in the callout cache for a specific address caused a log line with no
4880 reason for rejecting RCPT. Now it says "Previous (cached) callout
4881 verification failure".
4883 3. There was an off-by-one bug in the queryprogram router. An over-long
4884 return line was truncated at 256 instead of 255 characters, thereby
4885 overflowing its buffer with the terminating zero. As well as fixing this, I
4886 have increased the buffer size to 1024 (and made a note to document this).
4888 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4889 when Exim is waiting for an SMTP response from a remote server, Exim
4890 restarts its select() call on the socket, thereby resetting its timeout.
4891 This is not a problem when such interrupts are rare. Somebody set up a cron
4892 job to run exiwhat every 2 minutes, which is less than the normal select()
4893 timeout (5 or 10 minutes). This meant that the select() timeout never
4894 kicked in because it was always reset. I have fixed this by comparing the
4895 time when an interrupt arrives with the time at the start of the first call
4896 to select(). If more time than the timeout has elapsed, the interrupt is
4897 treated as a timeout.
4899 5. Some internal re-factoring in preparation for the addition of Sieve
4900 extensions (by MH). In particular, the "personal" test is moved to a
4901 separate function, and given an option for scanning Cc: and Bcc: (which is
4902 not set for Exim filters).
4904 6. When Exim created an email address using the login of the caller as the
4905 local part (e.g. when creating a From: or Sender: header line), it was not
4906 quoting the local part when it contained special characters such as @.
4908 7. Installed new OpenBSD configuration files.
4910 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4911 try to make them clearer.
4913 9. Callout options, other than the timeout value, were being ignored when
4914 verifying sender addresses in header lines. For example, when using
4916 verify = header_sender/callout=no_cache
4918 the cache was (incorrectly) being used.
4920 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4921 only the Cygwin environment.
4923 11. When return_path_on_delivery was set as a log selector, if different remote
4924 addresses in the same message used different return paths and parallel
4925 remote delivery occurred, the wrong values would sometimes be logged.
4926 (Whenever a remote delivery process finished, the return path value from
4927 the most recently started remote delivery process was logged.)
4929 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4930 lines when AUTH and/or TLS are in use. This is the "received protocol"
4931 field. Exim used to use "asmtp" for authenticated SMTP, without any
4932 indication (in the protocol name) for TLS use. Now it follows the RFC and
4933 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4934 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4935 names appear in log lines as well as in Received: header lines.
4937 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4938 extensions, and comparison tests, and to fix some bugs.
4940 14. Changes to the "personal" filter test:
4942 (1) The test was buggy in that it was just doing the equivalent of
4943 "contains" tests on header lines. For example, if a user's address was
4944 anne@some.where, the "personal" test would incorrectly be true for
4946 To: susanne@some.where
4948 This test is now done by extracting each address from the header in turn,
4949 and checking the entire address. Other tests that are part of "personal"
4950 are now done using regular expressions (for example, to check local parts
4951 of addresses in From: header lines).
4953 (2) The list of non-personal local parts in From: addresses has been
4954 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4955 taken from the Sieve specification recommendations.
4957 (3) If the message contains any header line starting with "List-" it is
4958 treated as non-personal.
4960 (4) The test for "circular" in the Subject: header line has been removed
4961 because it now seems ill-conceived.
4963 15. Minor typos in src/EDITME comments corrected.
4965 16. Installed latest exipick from John Jetmore.
4967 17. If headers_add on a router specified a text string that was too long for
4968 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4969 of string_sprintf() is now avoided.
4971 18. $message_body_size was not set (it was always zero) when running the DATA
4972 ACL and the local_scan() function.
4974 19. For the "mail" command in an Exim filter, no default was being set for
4975 the once_repeat time, causing a random time value to be used if "once" was
4976 specified. (If the value happened to be <= 0, no repeat happened.) The
4977 default is now 0s, meaning "never repeat". The "vacation" command was OK
4978 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4979 (I found it when inspecting the code).
4981 20. There is now an overall timeout for performing a callout verification. It
4982 defaults to 4 times the callout timeout, which applies to individual SMTP
4983 commands during the callout. The overall timeout applies when there is more
4984 than one host that can be tried. The timeout is checked before trying the
4985 next host. This prevents very long delays if there are a large number of
4986 hosts and all are timing out (e.g. when the network connections are timing
4987 out). The value of the overall timeout can be changed by specifying an
4988 additional sub-option for "callout", called "maxwait". For example:
4990 verify = sender/callout=5s,maxwait=20s
4992 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4993 to the end before writing, but this should make it even safer).
4995 22. Exim was forgetting that it had advertised PIPELINING for the second and
4996 subsequent messages on an SMTP connection. It was also not resetting its
4997 memory on STARTTLS and an internal HELO.
4999 23. When Exim logs an SMTP synchronization error within a session, it now
5000 records whether PIPELINING has been advertised or not.
5002 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5003 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5004 rather than long int.
5006 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5008 26. Added the never_mail option to autoreply.
5014 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5015 crash if the getsockname() call failed; this can happen if a connection is
5016 closed very soon after it is established. The problem was simply in the
5017 order in which certain operations were done, causing Exim to try to write
5018 to the SMTP stream before it had set up the file descriptor. The bug has
5019 been fixed by making things happen in the correct order.
5025 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5026 before the connection was closed, thus losing the rejection response.
5028 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5029 some early Solaris releases, but causes trouble in current releases where
5030 socklen_t is defined.
5032 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5035 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5038 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5039 that releasing the top of it at the end releases what was used for sub-
5040 expansions (unless the block got too big). However, discard this block if
5041 the first thing is a variable or header, so that we can use its block when
5042 it is dynamic (useful for very large $message_headers, for example).
5044 6. Lookups now cache *every* query, not just the most recent. A new, separate
5045 store pool is used for this. It can be recovered when all lookup caches are
5046 flushed. Lookups now release memory at the end of their result strings.
5047 This has involved some general refactoring of the lookup sources.
5049 7. Some code has been added to the store_xxx() functions to reduce the amount
5050 of flapping under certain conditions.
5052 8. log_incoming_interface used to affect only the <= reception log lines. Now
5053 it causes the local interface and port to be added to several more SMTP log
5054 lines, for example "SMTP connection from", and rejection lines.
5056 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5058 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5060 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5061 could be overwritten at the end of the current message (or the start of a
5062 new message if it was set in a HELO ACL). The value is now preserved for
5063 the duration of the SMTP connection.
5065 12. If a transport had a headers_rewrite setting, and a matching header line
5066 contained an unqualified address, that address was qualified, even if it
5067 did not match any rewriting rules. The underlying bug was that the values
5068 of the flags that permit the existence of unqualified sender and recipient
5069 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5070 non-local messages, and by -bnq for local messages) were not being
5071 preserved with the message after it was received.
5073 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5074 "next input=" as part of the text comprising the host identity instead of
5075 the correct text. The code was using the same buffer for two different
5076 strings. However, depending on which order the printing function evaluated
5077 its arguments, the bug did not always show up. Under Linux, for example, my
5078 test suite worked just fine.
5080 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5081 doesn't work with some older versions of Perl. It has been changed to "my",
5082 which in any case is probably the better facility to use.
5084 15. A really picky compiler found some instances of statements for creating
5085 error messages that either had too many or two few arguments for the format
5088 16. The size of the buffer for calls to the DNS resolver has been increased
5089 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5090 for addresses that have a lot of PTR records. This alleviates a problem; it
5091 does not fully solve it.
5093 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5094 buffer now truncates the list and logs the incident, which is the same
5095 action as happens when Exim is looking up a host name and its aliases.
5096 Previously in this situation something unpredictable would happen;
5097 sometimes it was "internal error: store_reset failed".
5099 18. If a server dropped the connection unexpectedly when an Exim client was
5100 using GnuTLS and trying to read a response, the client delivery process
5101 crashed while trying to generate an error log message.
5103 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5104 string, for example:
5106 warn message = H1: something\nH2: something
5108 the text was added as a single header line from Exim's point of view
5109 though it ended up OK in the delivered message. However, searching for the
5110 second and subsequent header lines using $h_h2: did not work. This has been
5111 fixed. Similarly, if a system filter added multiple headers in this way,
5112 the routers could not see them.
5114 20. Expanded the error message when iplsearch is called with an invalid key to
5115 suggest using net-iplsearch in a host list.
5117 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5118 ACLs are no longer actually imposed (and a message to that effect is
5121 22. If a "gecos" field in a passwd entry contained escaped characters, in
5122 particular, if it contained a \" sequence, Exim got it wrong when building
5123 a From: or a Sender: header from that name. A second bug also caused
5124 incorrect handling when an unquoted " was present following a character
5125 that needed quoting.
5127 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5128 was not being matched caselessly.
5130 24. Arranged for all hyphens in the exim.8 source to be escaped with
5133 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5134 properly. Recipient callout cache records were still being keyed to include
5135 the sender, even when use_sender was set false. This led to far more
5136 callouts that were necessary. The sender is no longer included in the key
5137 when use_sender is false.
5139 26. Added "control = submission" modifier to ACLs.
5141 27. Added the ${base62d: operator to decode base 62 numbers.
5143 28. dnsdb lookups can now access SRV records.
5145 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5146 the configuration file.
5148 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5149 (-v) mode. This makes the output for a verbose queue run more intelligible.
5151 31. Added a use_postmaster feature to recipient callouts.
5153 32. Added the $body_zerocount variable, containing the number of binary zero
5154 bytes in the message body.
5156 33. The time of last modification of the "new" subdirectory is now used as the
5157 "mailbox time last read" when there is a quota error for a maildir
5160 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5162 35. Added +ignore_unknown as a special item in host lists.
5164 36. Code for decoding IPv6 addresses in host lists is now included, even if
5165 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5166 address was recognized as an IP address, but was then not correctly decoded
5167 into binary, causing unexpected and incorrect effects when compared with
5174 1. Very minor rewording of debugging text in manualroute to say "list of
5175 hosts" instead of "hostlist".
5177 2. If verify=header_syntax was set, and a header line with an unqualified
5178 address (no domain) and a large number of spaces between the end of the
5179 name and the colon was received, the reception process suffered a buffer
5180 overflow, and (when I tested it) crashed. This was caused by some obsolete
5181 code that should have been removed. The fix is to remove it!
5183 3. When running in the test harness, delay a bit after writing a bounce
5184 message to get a bit more predictability in the log output.
5186 4. Added a call to search_tidyup() just before forking a reception process. In
5187 theory, someone could use a lookup in the expansion of smtp_accept_max_
5188 per_host which, without the tidyup, could leave open a database connection.
5190 5. Added the variables $recipient_data and $sender_data which get set from a
5191 lookup success in an ACL "recipients" or "senders" condition, or a router
5192 "senders" option, similar to $domain_data and $local_part_data.
5194 6. Moved the writing of debug_print from before to after the "senders" test
5197 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5198 problems for message scanning, either using a data ACL, or using
5199 local_scan() because the Received: header was not generated till after they
5200 were called (in order to set the time as the time of reception completion).
5201 I have revised the way this works. The header is now generated after the
5202 body is received, but before the ACL or local_scan() are called. After they
5203 are run, the timestamp in the header is updated.
5209 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5210 before starting a queue runner without re-exec. This happened only when
5211 deliver_drop_privilege was set or when the Exim user was set to root. The
5212 effect of the bug was that timeouts during subsequent deliveries caused
5213 crashes instead of being properly handled. The handler is now left at its
5214 default (and expected) setting.
5216 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5217 message, again when deliver_drop_privilege is set or Exim is run as root.
5218 The bug described in (1) was not present in this case, but the tidying up
5219 of the other signals was missing. I have made the two cases consistent.
5221 3. The ignore_target_hosts setting on a manualroute router was being ignored
5222 for hosts that were looked up using the /MX notation.
5224 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5227 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5228 operated on the sender address. After changing the $sender_address to <>
5229 for the sender address verify, Exim was re-instated it as the original
5230 (before rewriting) address, but remembering that it had rewritten it, so it
5231 wasn't rewriting it again. This bug also had the effect of breaking the
5232 sender address verification caching when the sender address was rewritten.
5234 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5235 This has been changed so that if the ip literal address matches
5236 ignore_target_hosts, the router declines.
5238 7. Added expansion conditions match_domain, match_address, and match_local_
5239 part (NOT match_host).
5241 8. The placeholder for the Received: header didn't have a length field set.
5243 9. Added code to Exim itself and to exim_lock to test for a specific race
5244 condition that could lead to file corruption when using MBX delivery. The
5245 issue is with the lockfile that is created in /tmp. If this file is removed
5246 after a process has opened it but before that process has acquired a lock,
5247 there is the potential for a second process to recreate the file and also
5248 acquire a lock. This could lead to two Exim processes writing to the file
5249 at the same time. The added code performs the same test as UW imapd; it
5250 checks after acquiring the lock that its file descriptor still refers to
5251 the same named file.
5253 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5254 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5257 11. Added the smtp_active_hostname option. If used, this will typically be made
5258 to depend on the incoming interface address. Because $interface_address is
5259 not set up until the daemon has forked a reception process, error responses
5260 that can happen earlier (such as "too many connections") no longer contain
5263 12. If an expansion in a condition on a "warn" statement fails because a lookup
5264 defers, the "warn" statement is abandoned, and the next ACL statement is
5265 processed. Previously this caused the whole ACL to be aborted.
5267 13. Added the iplsearch lookup type.
5269 14. Added ident_timeout as a log selector.
5271 15. Added tls_certificate_verified as a log selector.
5273 16. Added a global option tls_require_ciphers (compare the smtp transport
5274 option of the same name). This controls incoming TLS connections.
5276 17. I finally figured out how to make tls_require_ciphers do a similar thing
5277 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5278 before starting the TLS session.
5280 18. Tabs are now shown as \t in -bP output.
5282 19. If the log selector return_path_on_delivery was set, Exim crashed when
5283 bouncing a message because it had too many Received: header lines.
5285 20. If two routers both had headers_remove settings, and the first one included
5286 a superfluous trailing colon, the final name in the first list and the
5287 first name in the second list were incorrectly joined into one item (with a
5288 colon in the middle).
5294 1. Added -C and -D options to the exinext utility, mainly to make it easier
5295 to include in the automated testing, but these could be helpful when
5296 multiple configurations are in use.
5298 2. The exinext utility was not formatting the output nicely when there was
5299 an alternate port involved in the retry record key, nor when there was a
5300 message id as well (for retries that were specific to a specific message
5301 and a specific host). It was also confused by IPv6 addresses, because of
5302 the additional colons they contain. I have fixed the IPv4 problem, and
5303 patched it up to do a reasonable job for IPv6.
5305 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5306 delivery, the log line now contains "pipelined" if PIPELINING was used.
5308 4. An SMTP transport process used to panic and die if the bind() call to set
5309 an explicit outgoing interface failed. This has been changed; it is now
5310 treated in the same way as a connect() failure.
5312 5. A reference to $sender_host_name in the part of a conditional expansion
5313 that was being skipped was still causing a DNS lookup. This no longer
5316 6. The def: expansion condition was not recognizing references to header lines
5317 that used bh_ and bheader_.
5319 7. Added the _cache feature to named lists.
5321 8. The code for checking quota_filecount in the appendfile transport was
5322 allowing one more file than it should have been.
5324 9. For compatibility with Sendmail, the command line option
5332 and sets the incoming protocol and host name (for trusted callers). The
5333 host name and its colon can be omitted when only the protocol is to be set.
5334 Note the Exim already has two private options, -pd and -ps, that refer to
5335 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5336 "s", but I don't think that's a major issue.
5338 10. A number of refactoring changes to the code, none of which should affect
5341 (a) The number of logging options was getting close to filling up the
5342 32-bit word that was used as a bit map. I have split them into two classes:
5343 those that are passed in the argument to log_write(), and those that are
5344 only ever tested independently outside of that function. These are now in
5345 separate 32-bit words, so there is plenty of room for expansion again.
5346 There is no change in the user interface or the logging behaviour.
5348 (b) When building, for example, log lines, the code previously used a
5349 macro that called string_cat() twice, in order to add two strings. This is
5350 not really sufficiently general. Furthermore, there was one instance where
5351 it was actually wrong because one of the argument was used twice, and in
5352 one call a function was used. (As it happened, calling the function twice
5353 did not affect the overall behaviour.) The macro has been replaced by a
5354 function that can join an arbitrary number of extra strings onto a growing
5357 (c) The code for expansion conditions now uses a table and a binary chop
5358 instead of a serial search (which was left over from when there were very
5359 few conditions). Also, it now recognizes conditions like "pam" even when
5360 the relevant support is not compiled in: a suitably worded error message is
5361 given if an attempt is made to use such a condition.
5363 11. Added ${time_interval:xxxxx}.
5365 12. A bug was causing one of the ddress fields not to be passed back correctly
5366 from remote delivery subprocesses. The field in question was not being
5367 subsequently used, so this caused to problems in practice.
5369 13. Added new log selectors queue_time and deliver_time.
5371 14. Might have fixed a bug in maildirsizefile handling that threw up
5372 "unexpected character" debug warnings, and recalculated the data
5373 unnecessarily. In any case, I expanded the warning message to give more
5376 15. Added the message "Restricted characters in address" to the statements in
5377 the default ACL that block characters like @ and % in local parts.
5379 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5380 Three changes have been made:
5382 (a) There was a serious bug; a negative response to MAIL caused the whole
5383 recipient domain to be cached as invalid, thereby blocking all messages
5384 to all local parts at the same domain, from all senders. This bug has
5385 been fixed. The domain is no longer cached after a negative response to
5386 MAIL if the sender used is not empty.
5388 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5391 (c) A new callout option, "use_sender" has been added for people who want
5392 the modified behaviour.
5398 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5401 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5402 indeed breaks things for older releases.
5404 3. Added additional logging to the case where there is a problem reading data
5405 from a filter that is running in a subprocess using a pipe, in order to
5406 try to track down a specific problem.
5408 4. Testing facility fudge: when running in the test harness and attempting
5409 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5410 getting "No route to host". Convert this to a timeout.
5412 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5415 6. Some OS don't have socklen_t but use size_t instead. This affects the
5416 fifth argument of getsockopt() amongst other things. This is now
5417 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5418 can be set for individual OS. I have set it for SunOS5, OSF1, and
5419 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5420 some earlier ones do not.
5422 7. Change 4.30/15 was not doing the test caselessly.
5424 8. The standard form for an IPv6 address literal was being rejected by address
5425 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5426 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5427 this, as well as the form without the "IPv6" on the front (but only when
5428 address literals are enabled, of course).
5430 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5432 10. Exim crashed if a message with an empty sender address specified by -f
5433 encountered a router with an errors_to setting. This could be provoked only
5434 by a command such as
5438 where an empty string was supplied; "<>" did not hit this bug.
5440 11. Installed PCRE release 4.5.
5442 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5443 remained set. It is now erased.
5445 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5446 times from message ids (which are base 36 rather than the normal 62).
5448 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5449 were being counted as actual protocol errors, and logged if the log
5450 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5451 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5452 and DATA following a set of rejected RCPTs do not count as protocol errors.
5453 In other words, Exim assumes they were pipelined, though this may not
5454 actually be the case. Of course, in all cases the client gets an
5455 appropriate error code.
5457 15. If a lookup fails in an ACL condition, a message about the failure may
5458 be available; it is used if testing the ACL cannot continue, because most
5459 such messages specify what the cause of the deferral is. However, some
5460 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5461 that caused an old message to be retained and used if a later statement
5462 caused a defer, replacing the real cause of the deferral.
5464 16. If an IP address had so many PTR records that the DNS lookup buffer
5465 was not large enough to hold them, Exim could crash while trying to process
5466 the truncated data. It now detects and logs this case.
5468 17. Further to 4.21/58, another change has been made: if (and only if) the
5469 first line of a message (the first header line) ends with CRLF, a bare LF
5470 in a subsequent header line has a space inserted after it, so as not to
5471 terminate the header.
5473 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5474 unnecessarily, used atoi() instead of strtol(), and didn't check the
5475 termination when getting file sizes from file names by regex.
5477 19. Completely re-implemented the support for maildirsize files, in the light
5478 of a number of problems with the previous contributed implementation
5479 (4.30/29). In particular:
5481 . If the quota is zero, the maildirsize file is maintained, but no quota is
5484 . If the maildir directory does not exist, it is created before any attempt
5485 to write a maildirsize file.
5487 . The quota value in the file is just a cache; if the quota is changed in
5488 the transport, the new value overrides.
5490 . A regular expression is available for excluding directories from the
5493 20. The autoreply transport checks the characters in options that define the
5494 message's headers; it allows continued headers, but it was checking with
5495 isspace() after an embedded newline instead of explicitly looking for a
5498 21. If all the "regular" hosts to which an address was routed had passed their
5499 expiry times, and had not reached their retry times, the address was
5500 bounced, even if fallback hosts were defined. Now Exim should go on to try
5503 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5504 equivalent code in the SMTP transport. Some hosts send humungous responses
5505 to HELO/EHLO, more than 1024 it seems.
5507 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5508 gives compiler warnings in some environments. I've now done it "properly",
5511 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5512 (because of problems with the built-in one) was declared to return uschar *
5513 instead of char *, causing compiler failure.
5515 25. Fixed a file descriptor leak when processing alias/forward files.
5517 26. Fixed a minor format string issue in dbfn.c.
5519 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5521 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5522 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5525 29. When a sender address is verified, it is cached, to save repeating the test
5526 when there is more than one recipient in a message. However, when the
5527 verification involves a callout, it is possible for different callout
5528 options to be set for different recipients. It is too complicated to keep
5529 track of this in the cache, so now Exim always runs a verification when a
5530 callout is required, relying on the callout cache for the optimization.
5531 The overhead is duplication of the address routing, but this should not be
5534 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5535 to be verified with callout=postmaster, and the main callout worked but the
5536 postmaster check failed, the verification correctly failed. However, if a
5537 subsequent RCPT command asked for sender verification *without* the
5538 postmaster check, incorrect caching caused this verification also to fail,
5541 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5542 it was not caching the DNS options (qualify_single, search_parents) that
5543 were used when the lookup failed. A subsequent lookup with different
5544 options therefore always gave the same answer, though there were cases
5545 where it should not have. (Example: a "domains = !$mx_any" option on a
5546 dnslookup router: the "domains" option is always processed without any
5547 widening, but the router might have qualify_single set.) Now Exim uses the
5548 cached value only when the same options are set.
5550 32. Added John Jetmore's "exipick" utility to the distribution.
5552 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5553 than a timeout (e.g. a certificate is required, and is not provided), an
5554 Exim server now closes the connection immediately. Previously it waited for
5555 the client to close - but if the client is SSL, it seems that they each
5556 wait for each other, leading to a delay before one of them times out.
5558 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5559 maintained 0.8.x compatibility because I don't think many are using it, and
5560 it is clearly obsolete.
5562 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5565 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5566 client certificate was expired. A simple patch fixes this, though I don't
5567 understand the full logic of why the verify callback is called multiple
5570 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5571 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5572 which causes problems with some clients (such as the Certicom SSL Plus
5573 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5574 disables the coutermeasure allowing Eudora to connect."
5576 38. Exim was not checking that a write() to a log file succeeded. This could
5577 lead to Bad Things if a log got too big, in particular if it hit a file
5578 size limit. Exim now panics and dies if it cannot write to a log file, just
5579 as it does if it cannot open a log file.
5581 39. Modified OS/Makefile-Linux so that it now contains
5583 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5585 The two -D definitions ensure that Exim is compiled with large file
5586 support, which makes it possible to handle log files that are bigger than
5589 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5590 instance) a domain was checked against a named list that involved a lookup,
5591 causing $domain_data to be set, then another domain was checked against the
5592 same list, then the first domain was re-checked, the value of $domain_data
5593 after the final check could be wrong. In particular, if the second check
5594 failed, it could be set empty. This bug probably also applied to
5597 41. The strip_trailing_dot option was not being applied to the address given
5598 with the -f command-line option.
5600 42. The code for reading a message's header from the spool was incrementing
5601 $received_count, but never initializing it. This meant that the value was
5602 incorrect (doubled) while delivering a message in the same process in which
5603 it was received. In the most common configuration of Exim, this never
5604 happens - a fresh exec is done - but it can happen when
5605 deliver_drop_privilege is set.
5607 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5608 it now includes up to 150 characters of the unexpected data in the log
5611 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5612 and building data strings. The size of both of these buffers was 10 000
5613 bytes - far larger than anybody would *ever* want, thought I. Needless to
5614 say, somebody hit the limit. I have increased the maximum line length to
5615 20 000 and the maximum data length of concatenated lines to 100 000. I have
5616 also fixed two bugs, because there was no checking on these buffers. Tsk,
5617 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5618 buffer is too small.
5620 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5621 lsearch lookups. Now it does.
5623 46. When parsing a route_list item in a manualroute router, a fixed-length
5624 buffer was used for the list of hosts. I made this 1024 bytes long,
5625 thinking that nobody would ever have a list of hosts that long. Wrong.
5626 Somebody had a whole pile of complicated expansion conditions, and the
5627 string was silently truncated, leading to an expansion error. It turns out
5628 that it is easier to change to an unlimited length (owing to other changes
5629 that have happened since this code was originally written) than to build
5630 structure for giving a limitation error. The length of the item that
5631 expands into the list of hosts is now unlimited.
5633 47. The lsearch lookup could not handle data where the length of text line was
5634 more than 4095 characters. Such lines were truncated, leading to shortened
5635 data being returned. It should now handle lines of any length.
5637 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5638 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5641 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5642 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5643 no longer happens in comment lines. A list of replacements is now placed
5644 at the head of all of the source files, except those whose only change is
5645 to replace PERL_COMMAND in the very first #! line.
5647 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5648 messages on the queue, with a bottom-up merge sort, using code contributed
5649 by Michael Haardt. This should make operations like -bp somewhat faster on
5650 large queues. It won't affect queue runners, except when queue_run_in_order
5653 51. Installed eximstats 1.31 in the distribution.
5655 52. Added support for SRV lookups to the dnslookup router.
5657 53. If an ACL referred to $message_body or $message_body_end, the value was not
5658 reset for any messages that followed in the same SMTP session.
5660 54. The store-handling optimization for building very long strings was not
5661 differentiating between the different store pools. I don't think this
5662 actually made any difference in practice, but I've tidied it.
5664 55. While running the routers to verify a sender address, $sender_address
5665 was still set to the sender address. This is wrong, because when routing to
5666 send a bounce to the sender, it would be empty. Therefore, I have changed
5667 it so that, while verifying a sender address, $sender_address is set to <>.
5668 (There is no change to what happens when verifying a recipient address.)
5670 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5671 target A or AAAA records (if not already returned) without resetting the
5672 qualify_single or search_parents options of the DNS resolver. These are
5673 inappropriate in this case because the targets of MX and SRV records must
5674 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5675 target that, when qualified, matched something in the local domain. These
5676 two options are now turned off when doing these lookups.
5678 57. It seems that at least some releases of Reiserfs (which does not have the
5679 concept of a fixed number of inodes) returns zero and not -1 for the
5680 number of available inodes. This interacted badly with check_spool_inodes,
5681 which assumed that -1 was the "no such thing" setting. What I have done is
5682 to check that the total number of inodes is greater than zero before doing
5683 the test of how many are available.
5685 58. When a "warn" ACL statement has a log_message modifier, the message is
5686 remembered, and not repeated. This is to avoid a lot of repetition when a
5687 message has many recipients that cause the same warning to be written.
5688 However, Exim was preserving the list of already written lines for an
5689 entire SMTP session, which doesn't seem right. The memory is now reset if a
5690 new message is started.
5692 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5693 debugging output unless log_rewrite was also set.
5695 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5696 of (int)(handle) when we know that handle contains (void *)(-1).
5698 61. The Exim daemon panic-logs an error return when it closes the incoming
5699 connection. However "connection reset by peer" seems to be common, and
5700 isn't really an error worthy of noting specially, so that particular error
5703 62. When Exim is trying to find all the local interfaces, it used to panic and
5704 die if the ioctl to get the interface flags failed. However, it seems that
5705 on at least one OS (Solaris 9) it is possible to have an interface that is
5706 included in the list of interfaces, but for which you get a failure error
5707 for this call. This happens when the interface is not "plumbed" into a
5708 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5709 failure of the "get flags" call assumes that the interface is down.
5711 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5712 makes life easier for people who are doing arithmetic on fields extracted
5713 from dates, where you often get leading zeros that should not be
5714 interpreted as octal.
5716 64. Added qualify_domain to the redirect router, to override the global
5719 65. If a pathologically long header line contained very many addresses (the
5720 report of this problem mentioned 10 000) and each of them was rewritten,
5721 Exim could use up a very large amount of memory. (It kept on making new
5722 copies of the header line as it rewrote, and never released the old ones.)
5723 At the expense of a bit more processing, the header rewriting function has
5724 been changed so that it no longer eats memory in this way.
5726 66. The generation of the Received: header has been moved from the time that a
5727 message starts to be received, to the time that it finishes. The timestamp
5728 in the Received: header should now be very close to that of the <= log
5729 line. There are two side-effects of this change:
5731 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5732 logged header lines no longer include the local Received: line, because
5733 it has not yet been created. The same applies to a copy of the message
5734 that is returned to a non-SMTP sender when a message is rejected.
5736 (b) When a filter file is tested using -bf, no additional Received: header
5737 is added to the test message. After some thought, I decided that this
5740 This change does not affect the value of $received_for. It is still set
5741 after address rewriting, but before local_scan() is called.
5743 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5745 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5746 gave an unhelpful panic error message, and a defer error. I have managed to
5747 change this behaviour so that it now rejects any supplied certificate,
5748 which seems right, as the list of acceptable certificates is empty.
5750 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5751 gave an unhelpful defer error. I have not managed to make this reject any
5752 supplied certificates, but the error message it gives is "no certificate
5753 supplied", which is not helpful.
5755 70. exigrep's output now also includes lines that are not associated with any
5756 message, but which match the given pattern. Implemented by a patch from
5757 Martin Sluka, which also tidied up the Perl a bit.
5759 71. Recipient callout verification, like sender verification, was using <> in
5760 the MAIL FROM command. This isn't really the right thing, since the actual
5761 sender may affect whether the remote host accepts the recipient or not. I
5762 have changed it to use the actual sender in the callout; this means that
5763 the cache record is now keyed on a recipient/sender pair, not just the
5764 recipient address. There doesn't seem to be a real danger of callout loops,
5765 since a callout by the remote host to check the sender would use <>.
5766 [SEE ABOVE: changed after hitting problems.]
5768 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5769 temporary errors. However, in the case of such a code being given after
5770 the end of a data transmission (i.e. after ".") Exim was failing to write
5771 a retry record for the message. (Yes, there was some broken host that was
5772 actually sending 8xx at this point.)
5774 73. An unknown lookup type in a host list could cause Exim to panic-die when
5775 the list was checked. (An example that provoked this was putting <; in the
5776 middle of a list instead of at the start.) If this happened during a DATA
5777 ACL check, a -D file could be left lying around. This kind of configuration
5778 error no longer causes Exim to die; instead it causes a defer error. The
5779 incident is still logged to the main and panic logs.
5781 74. Buglet left over from Exim 3 conversion. The message "too many messages
5782 in one connection" was written to the rejectlog but not the mainlog, except
5783 when address rewriting (yes!) was being logged.
5785 75. Added write_rejectlog option.
5787 76. When a system filter was run not as root (that is, when system_filter_user
5788 was set), the values of the $n variables were not being returned to the
5789 main process; thus, they were not subsequently available in the $sn
5792 77. Added +return_path_on_delivery log selector.
5794 78. A connection timeout was being treated differently from recipients deferred
5795 when testing hosts_max_try with a message that was older than the host's
5796 retry timeout. (The host should not be counted, thus allowing all hosts to
5797 be tried at least once before bouncing.) This may have been the cause of an
5798 occasionally reported bug whereby a message would remain on the queue
5799 longer than the retry timeout, but would be bounced if a delivery was
5800 forced. I say "may" because I never totally pinned down the problem;
5801 setting up timeout/retry tests is difficult. See also the next item.
5803 79. The ultimate address timeout was not being applied to errors that involved
5804 a combination of host plus message (for example, a timeout on a MAIL
5805 command). When an address resolved to a number of possible hosts, and they
5806 were not all tried for each delivery (e.g. because of hosts_max_try), a
5807 message could remain on the queue longer than the retry timeout.
5809 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5812 81. Fixed an obscure SMTP outgoing bug which required at least the following
5813 conditions: (a) there was another message waiting for the same server;
5814 (b) the server returned 5xx to all RCPT commands in the first message so
5815 that the message was not completed; (c) the server dropped the connection
5816 or gave a negative response to the RSET that Exim sends to abort the
5817 transaction. The observed case was a dropped connection after DATA that had
5818 been sent in pipelining mode. That is, the server had advertised PIPELINING
5819 but was not implementing it correctly. The effect of the bug was incorrect
5820 behaviour, such as trying another host, and this could lead to a crash.
5826 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5827 and daemon.c were passed as pointers to ints; they should have been
5828 pointers to socklen_t variables (which are typically unsigned ints).
5830 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5833 3. Fixed a really odd bug that affected only the testing scheme; patching a
5834 certain fixed string in the binary changed the value of another string that
5835 happened to be identical to the end of the original first string.
5837 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5838 name", it returns that address as the IP address. On some operating
5839 systems (e.g. Solaris), it also passes back the IP address string as the
5840 "host name". However, on others (e.g. Linux), it passes back an empty
5841 string. Exim wasn't checking for this, and was changing the host name to an
5842 empty string, assuming it had been canonicalized.
5844 5. Although rare, it is permitted to have more than one PTR record for a given
5845 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5846 all the names associated with an address, because they do in Solaris.
5847 However, it seems that they do not in Linux for data that comes from the
5848 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5849 I found this out when I moved to a new Linux workstation and tried to run
5850 the Exim test suite.
5852 To get round this problem I have changed the code so that it now does its
5853 own call to the DNS to look up PTR records when searching for a host name.
5854 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5855 addresses that are only in /etc/hosts are still found.
5857 This behaviour is, however, controlled by an option called host_lookup_
5858 order, which defaults to "bydns:byaddr". If people want to use the other
5859 order, or indeed, just use one or the other means of lookup, they can
5860 specify it in this variable.
5862 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5863 some operating systems, this comes back from gethostbyaddr() as an empty
5864 string, and this is what Exim used to test for. However, it seems that in
5865 other systems, "." is yielded. Exim now tests for this case too.
5867 7. The values of check_spool_space and check_log_space are now held internally
5868 as a number of kilobytes instead of an absolute number of bytes. If a
5869 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5870 kilobyte. This means that much larger values can be stored.
5872 8. Exim monitor: an attempt to get the action menu when not actually pointing
5873 at a message produces an empty menu entitled "No message selected". This
5874 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5875 no entries in it ("Shell widget menu has zero width and/or height"). So I
5876 have added a single, blank menu entry in this case.
5878 9. Added ${quote_local_part.
5880 10. MIME decoding is now applied to the contents of Subject: header lines when
5883 11. Now that a reference to $sender_host_address automatically causes a reverse
5884 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5885 host lookup before query-style lookups in lists that might use this
5886 variable. This has therefore been abolished, and the "net-" prefix is no
5887 longer necessary for query-style lookups.
5889 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5890 have been a typo for LFLAGS, so it has been changed.
5892 13. The install script calls Exim with "-C /dev/null" in order to find the
5893 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5894 to be output. However, since Exim outputs its version number before the
5895 error, it didn't break the script. It just looked ugly. I fixed this by
5896 always allowing "-C /dev/null" if the caller is root.
5898 14. Ignore overlarge ACL variable number when reading spool file - insurance
5899 against a later release with more variables having written the file.
5901 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5902 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5903 this, as well as the form without the "IPv6" on the front.
5905 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5906 OS/Makefile-Darwin file.
5908 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5909 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5911 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5912 variable. However, some OS use ints for inodes. I've added cast to long int
5913 to get rid of the compiler warning.
5915 19. I had forgotten to lock out "/../" in configuration file names when
5916 ALT_CONFIG_PREFIX was set.
5918 20. Routers used for verification do not need to specify transports. However,
5919 if such a router generated a host list, and callout was configured, Exim
5920 crashed, because it could not find a port number from the (non-existent)
5921 transport. It now assumes port 25 in this circumstance.
5923 21. Added the -t option to exigrep.
5925 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5926 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5927 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5928 them is set, LOOKUP_LSEARCH is forced.
5930 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5932 24. Added sender and host information to the "rejected by local_scan()" log
5933 line; previously there was no indication of these.
5935 25. Added .include_if_exists.
5937 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5938 Linux. It turns out that not all file systems support this. Apparently some
5939 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5940 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5941 not supported on the file descriptor, is now ignored when Exim is trying to
5942 sync a directory. This applies only to Linux.
5944 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5946 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5947 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5950 29. Added support for maildirsize files from supplied patch (modified a bit).
5952 30. The use of :fail: followed by an empty string could lead Exim to respond to
5953 sender verification failures with (e.g.):
5955 550 Verification failed for <xxx>
5956 550 Sender verify failed
5958 where the first response line was missing the '-' that indicates it is not
5959 the final line of the response.
5961 31. The loop for finding the name of the user that called Exim had a hardwired
5962 limit of 10; it now uses the value of finduser_retries, which is used for
5963 all other user lookups.
5965 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5968 33. Exim was neglecting to zero errno before one call of strtol() when
5969 expanding a string and expecting an integer value. On some systems this
5970 resulted in spurious "integer overflow" errors. Also, it was casting the
5971 result into an int without checking.
5973 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5974 did not work. The code looks as if it has *never* worked, though it appears
5975 to have been documented since at least release 1.62. I have made it work.
5977 35. The "timeout_DNS" error in retry rules, also documented since at least
5978 1.62, also never worked. As it isn't clear exactly what this means, and
5979 clearly it isn't a major issue, I have abolished the feature by treating it
5980 as "timeout", and writing a warning to the main and panic logs.
5982 36. The display of retry rules for -brt wasn't always showing the error code
5985 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5986 timeout_connect_A, timeout_connect_MX.
5988 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5989 to the empty sender.
5991 39. The daemon was not analysing the content of -oX till after it had closed
5992 stderr and disconnected from the controlling terminal. This meant that any
5993 syntax errors were only noted on the panic log, and the return code from
5994 the command was 0. By re-arranging the code a little, I've made the
5995 decoding happen first, so such errors now appear on stderr, and the return
5996 code is 1. However, the actual setting up of the sockets still happens in
5997 the disconnected process, so errors there are still only recorded on the
6000 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6001 connections (as happens on some IP stacks) was logged at start up time as
6002 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6003 it from "IPv6 and IPv4", which means that two separate sockets are being
6006 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6007 says whether AF_INET or AF_INET6 was passed as an argument.
6009 42. Exiwhat output was messed up when time zones were included in log
6012 43. Exiwhat now gives more information about the daemon's listening ports,
6013 and whether -tls-on-connect was used.
6015 44. The "port" option of the smtp transport is now expanded.
6017 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6018 silently ignored. Now an error message is written to the main and panic
6021 46. There's a new ACL modifier called "logwrite" which writes to a log file
6022 as soon as it is encountered.
6024 47. Added $local_user_uid and $local_user_gid at routing time.
6026 48. Exim crashed when trying to verify a sender address that was being
6029 49. Exim was recognizing only a space character after ".include". It now also
6030 recognizes a tab character.
6032 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6033 extracting the relevant information from the specification. The man page no
6034 longer contains scrambled data for the -d option, and I've added a section
6035 at the front about calling Exim under different names.
6037 51. Added "extra_headers" argument to the "mail" command in filter files.
6039 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6042 53. Installed eximstats 1.29.
6044 54. Added transport_filter_timeout as a generic transport option.
6046 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6047 Cc: header lines. This was required by RFC 822, but it not required by RFC
6050 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6051 incoming messages that did not have them. Now it does so only if the
6052 message originates locally, that is, if there is no associated remote host
6053 address. When Resent- header lines are present, this applies to the Resent-
6054 lines rather than the non-Resent- lines.
6056 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6057 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6059 58. Messages for configuration errors now include the name of the main
6060 configuration file - useful now that there may be more than one file in a
6061 list (.included file names were always shown).
6063 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6064 for those rare installations that do not start the daemon as root or run it
6065 setuid root. I've cut out the call to initgroups() if the daemon is not
6068 60. The Exim user and group can now be bound into the binary as text strings
6069 that are looked up at the start of Exim's processing.
6071 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6073 62. Added $mailstore_basename variable.
6075 63. Installed patch to sieve.c from Michael Haardt.
6077 64. When Exim failed to open the panic log after failing to open the main log,
6078 the original message it was trying to log was written to stderr and debug
6079 output, but if they were not available (the usual case in production), it
6080 was lost. Now it is written to syslog before the two lines that record the
6081 failures to open the logs.
6083 65. Users' Exim filters run in subprocesses under the user's uid. It is
6084 possible for a "deliver" command or an alias in a "personal" command to
6085 provoke an address rewrite. If logging of address rewriting is configured,
6086 this fails because the process is not running as root or exim. There may be
6087 a better way of dealing with this, but for the moment (because 4.30 needs
6088 to be released), I have disabled address rewrite logging when running a
6089 filter in a non-root, non-exim process.
6095 1. The buildconfig auxiliary program wasn't quoting the value set for
6096 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6097 not defined. This bug was masked in 4.22 by the effect that was fixed in
6100 2. Some messages that were rejected after a message id was allocated were
6101 shown as "incomplete" by exigrep. It no longer does this for messages that
6102 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6104 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6105 have allow_domain_literals set, the ID did not get logged in the <= line.
6106 Domain literals are now always recognized in Message-ID: header lines.
6108 4. The first argument for a ${extract expansion item is the key name or field
6109 number. Leading and trailing spaces in this item were not being ignored,
6110 causing some misleading effects.
6112 5. When deliver_drop_privilege was set, single queue runner processes started
6113 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6114 same command in the process it spins off) were not dropping privilege.
6116 6. When the daemon running as "exim" started a queue runner, it always
6117 re-executed Exim in the spun-off process. This is a waste of effort when
6118 deliver_drop_privilege is set. The new process now just calls the
6119 queue-runner function directly.
6125 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6128 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6129 ignored. Though the use of -oP was forcing the writing of a pid file, it
6130 was always written to the default place.
6132 3. If the message "no IP address found for host xxxx" is generated during
6133 incoming verification, it is now followed by identification of the incoming
6134 connection (so you can more easily find what provoked it).
6136 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6138 5. Added some features to "harden" Exim a bit more against certain attacks:
6140 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6141 be put in Local/Makefile. This is like the never_users runtime option,
6142 but it cannot be overridden. The default setting is "root".
6144 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6145 prefix string with which any file named in a -C command line option
6148 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6149 is retained for -C and -D only if the caller of Exim is root. Without
6150 it, the exim user may also use -C and -D and retain privilege.
6152 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6153 command line option is disabled.
6155 6. Macro names set by the -D option must start with an upper case letter, just
6156 like macro names defined in the configuration file.
6158 7. Added "dereference=" facility to LDAP.
6160 8. Two instances of the typo "uknown" in the source files are fixed.
6162 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6163 the Configure-Makefile script screwed up while processing it.
6165 10. Incorporated PCRE 4.4.
6167 11. The SMTP synchronization check was not operating right at the start of an
6168 SMTP session. For example, it could not catch a HELO sent before the client
6169 waited for the greeting. There is now a check for outstanding input at the
6170 point when the greeting is written. Because of the duplex, asynchronous
6171 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6172 way, but not yet received, when the check is performed.
6174 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6175 on TCP/IP sockets, because this apparently causes some broken clients to
6178 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6179 unchanged) from the Cygwin maintainer.
6181 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6182 was supported instead of testing for mbx. Effectively a typo.
6184 15. The spa authenticator server code was not checking that the input it
6185 received was valid base64.
6187 16. The debug output line for the "set" modifier in ACLs was not showing the
6188 name of the variable that was being set.
6190 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6192 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6193 lookup on its own. Something else was needed to trigger the lookup. For
6194 example, a match in host_lookup or the need for a host name in a host list.
6195 Now, if $sender_host_name is referenced and the host name has not yet been
6196 looked up, a lookup is performed. If the lookup fails, the variable remains
6197 empty, and $host_lookup_failed is set to "1".
6199 19. Added "eqi" as a case-independent comparison operator.
6201 20. The saslauthd authentication condition could segfault if neither service
6202 nor realm was specified.
6204 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6205 error message that was logged was misleading, and incoming SMTP
6206 connections were dropped. The message is now more accurate, and temporary
6207 errors are given to SMTP connections.
6209 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6210 (except RSET) with a 421 error, until QUIT is received. However, it was
6211 failing to send a response to QUIT.
6213 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6214 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6215 ACL is now run after the helo_try_verify_hosts code.
6217 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6218 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6219 case-independent, so other case variants are also recognized). Apparently
6220 some systems use these upper case variants.
6222 25. If more than two messages were waiting for the same host, and a transport
6223 filter was specified for the transport, Exim sent two messages over the
6224 same TCP/IP connection, and then failed with "socket operation on non-
6225 socket" when it tried to send the third.
6227 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6229 27. The extern definition of crypt16() in expand.c was not being excluded when
6230 the OS had its own crypt16() function.
6232 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6233 as a preferred synonym for return_size_limit, both as an option and as an
6236 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6238 30. Changed the default configuration ACL to relax the local part checking rule
6239 for addresses that are not in any local domains. For these addresses,
6240 slashes and pipe symbols are allowed within local parts, but the sequence
6241 /../ is explicitly forbidden.
6243 31. SPA server authentication was not clearing the challenge buffer before
6246 32. log_message in a "warn" ACL statement was writing to the reject log as
6247 well as to the main log, which contradicts the documentation and doesn't
6248 seem right (because no rejection is happening). So I have stopped it.
6250 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6251 However, I am unable to do any testing of this.
6253 34. Fixed an infelicity in the appendfile transport. When checking directories
6254 for a mailbox, to see if any needed to be created, it was accidentally
6255 using path names with one or more superfluous leading slashes; tracing
6256 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6258 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6259 contents are added to the log line that is written for every discarded
6260 recipient. (Previously a log_message setting was ignored.)
6262 36. The ${quote: operator now quotes the string if it is empty.
6264 37. The install script runs exim in order to find its version number. If for
6265 some reason other than non-existence or emptiness, which it checks, it
6266 could not run './exim', it was installing it with an empty version number,
6267 i.e. as "exim-". This error state is now caught, and the installation is
6270 38. An argument was missing from the function that creates an error message
6271 when Exim fails to connect to the socket for saslauthd authentication.
6272 This could cause Exim to crash, or give a corrupted message.
6274 39. Added isip, isip4, and isip6 to ${if conditions.
6276 40. The ACL variables $acl_xx are now saved with the message, and can be
6277 accessed later in routers, transports, and filters.
6279 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6280 strings in the file are not string-expanded.
6282 42. If a MAIL command specified a SIZE value that was too large to fit into an
6283 int variable, the check against message_size_limit failed. Such values are
6284 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6285 day this will have to be increased, but I don't think I want to be around
6286 when emails are that large.
6293 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6294 iconv() is not standard in FreeBSD.
6296 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6297 IPv6 enabled. The observed symptom was a segmentation fault on return from
6298 the function os_common_find_running_interfaces() in src/os.c.
6300 3. In the check_special_case() function in daemon.c I had used "errno" as an
6301 argument name, which causes warnings on some systems. This was basically a
6302 typo, since it was named "eno" in the comments!
6304 4. The code that waits for the clock to tick (at a resolution of some fraction
6305 of a second) so as to ensure message-id uniqueness was always waiting for
6306 at least one whole tick, when it could have waited for less. [This is
6307 almost certainly not relevant at current processor speeds, where it is
6308 unlikely to ever wait at all. But we try to future-proof.]
6310 5. The function that sleeps for a time interval that includes fractions of a
6311 second contained a race. It did not block SIGALRM between setting the
6312 timer, and suspending (a couple of lines later). If the interval was short
6313 and the sigsuspend() was delayed until after it had expired, the suspension
6314 never ended. On busy systems this could lead to processes getting stuck for
6317 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6318 process, before it forks any delivery processes. The open lookup caching
6319 mechanism meant that the open file or database connection was passed into
6320 the delivery process. The problem was that delivery processes always tidy
6321 up cached lookup data. This could cause a problem for the next delivery
6322 process started by the queue runner, because the external queue runner
6323 process does not know about the closure. So the next delivery process
6324 still has data in the lookup cache. In the case of a file lookup, there was
6325 no problem because closing a file descriptor in a subprocess doesn't affect
6326 the parent. However, if the lookup was caching a connection to a database,
6327 the connection was closed, and the second delivery process was likely to
6328 see errors such as "PGSQL: query failed: server closed the connection
6329 unexpectedly". The problem has been fixed by closing all cached lookups
6330 in a queue runner before running a delivery process.
6332 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6333 seem to have the "const" qualifier which it has on other OS. I've
6336 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6337 *of the same type* (client or server) with the same public name that an
6338 error should be diagnosed.
6340 9. When Exim looked up a host name for an IP address, but failed to find the
6341 original IP address when looking up the host name (a safety check), it
6342 output the message "<ip address> does not match any IP for NULL", which was
6343 confusing, to say the least. The bug was that the host name should have
6344 appeared instead of "NULL".
6346 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6347 user that is built into the binary, and the -C or -D options is used, root
6348 privilege is dropped before the configuration file is read. In addition,
6349 logging is switched to stderr instead of the normal log files. If the
6350 configuration then re-defines the Exim user, the unprivileged environment
6351 is probably not what is expected, so Exim logs a panic warning message (but
6354 However, if deliver_drop_privilege is set, the unprivileged state may well
6355 be exactly what is intended, so the warning has been cut out in that case,
6356 and Exim is allowed to try to write to its normal log files.
6362 1. smtp_return_error_details was not giving details for temporary sender
6363 or receiver verification errors.
6365 2. Diagnose a configuration error if two authenticators have the same public
6368 3. Exim used not to create the message log file for a message until the first
6369 delivery attempt. This could be confusing when incoming messages were held
6370 for policy or load reasons. The message log file is now created at the time
6371 the message is received, and an initial "Received" line is written to it.
6373 4. The automatically generated man page for command line options had a minor
6374 bug that caused no ill effects; however, a more serious problem was that
6375 the procedure for building the man page automatically didn't always
6376 operate. Consequently, release 4.20 contains an out-of-date version. This
6377 shouldn't happen again.
6379 5. When building Exim with embedded Perl support, the script that builds the
6380 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6381 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6383 6. The freeze_tell option was not being used for messages that were frozen on
6384 arrival, either by an ACL or by local_scan().
6386 7. Added the smtp_incomplete_transaction log selector.
6388 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6389 was accepting AUTH without a new EHLO.
6391 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6392 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6395 10. Logging of TCP/IP connections (when configured) now happens in the main
6396 daemon process instead of the child process, so that the TCP/IP connection
6397 count is more accurate (but it can never be perfect).
6399 11. The use of "drop" in a nested ACL was not being handled correctly in the
6400 outer ACL. Now, if condition failure induced by the nested "drop" causes
6401 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6402 or "require"), the connection is dropped.
6404 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6405 that yield "discard" can now be used with an "accept" or a "discard" verb,
6406 but an error is generated for any others (because I can't see a useful way
6407 to define what should happen).
6409 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6410 are now processed in the same way as lines in the Exim configuration file.
6411 In particular, continuation lines are supported.
6413 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6415 15. Added -ti meaning -t -i.
6417 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6418 domains, and warn by logging if others are found.
6420 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6421 returned from GIFCONF when Exim is trying to find the list of interfaces on
6422 a host. The code in os.c has been modified to copy each ifreq to an aligned
6423 structure in all cases.
6425 Also, in some cases, the returned ifreq's were being copied to a 'struct
6426 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6427 means the last couple of bytes of an IPv6 address could be chopped if the
6428 ifreq contained only a normal sockaddr (14 bytes storage).
6430 18. Named domain lists were not supported in the hosts_treat_as_local option.
6431 An entry such as +xxxx was not recognized, and was treated as a literal
6434 19. Ensure that header lines added by a DATA ACL are included in the reject log
6435 if the ACL subsequently rejects the message.
6437 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6438 MD5 (which is deprecated).
6440 21. When testing a filter file using -bf, Exim was writing a message when it
6441 took the sender from a "From " line in the message, but it was not doing so
6442 when it took $return_path from a Return-Path: header line. It now does.
6444 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6445 with a valid header line field name (a series of printing characters
6446 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6448 23. Changed "disc" in the source to "disk" to conform to the documentation and
6449 the book and for uniformity.
6451 24. Ignore Sendmail's -Ooption=value command line item.
6453 25. When execve() failed while trying to run a command in a pipe transport,
6454 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6455 could be confused with a return value of 69 from the command itself. This
6456 has been changed to 127, the value the shell returns if it is asked to run
6457 a non-existent command. The wording for the related log line suggests a
6458 non-existent command as the problem.
6460 26. If received_header_text expands to an empty string, do not add a Received:
6461 header line to the message. (Well, it adds a token one on the spool, but
6462 marks it "old" so that it doesn't get used or transmitted.)
6464 27. Installed eximstats 1.28 (addition of -nt option).
6466 28. There was no check for failure on the call to getsockname() in the daemon
6467 code. This can fail if there is a shortage of resources on the system, with
6468 ENOMEM, for example. A temporary error is now given on failure.
6470 29. Contrary to the C standard, it seems that in some environments, the
6471 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6472 program. Exim now does this explicitly; it affects the formatting of
6473 timestamps using strftime().
6475 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6476 complaints. I've now initialized all the variables, to avoid this.
6478 32. Header lines added by a system filter were not being "seen" during
6479 transport-time rewrites.
6481 33. The info_callback() function passed to OpenSSL is set up with type void
6482 (*)(SSL *, int, int), as described somewhere. However, when calling the
6483 function (actually a macro) that sets it up, the type void(*)() is
6484 expected. I've put in a cast to prevent warnings from picky compilers.
6486 34. If a DNS black list lookup found a CNAME record, but there were no A
6487 records associated with the domain it pointed at, Exim crashed.
6489 35. If a DNS black list lookup returned more than one A record, Exim ignored
6490 all but the first. It now scans all returned addresses if a particular IP
6491 value is being sought. In this situation, the contents of the
6492 $dnslist_value variable are a list of all the addresses, separated by a
6495 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6496 to accept a host name and all its aliases if the forward lookup for any of
6497 them yielded the IP address of the incoming connection. Now it accepts only
6498 those names whose forward lookup yields the correct IP address. Any other
6499 names are discarded. This closes a loophole whereby a rogue DNS
6500 administrator could create reverse DNS records to break through a
6501 wildcarded host restriction in an ACL.
6503 37. If a user filter or a system filter that ran in a subprocess used any of
6504 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6505 the wrong values were passed to the pipe command ($thisaddress had the
6506 value of $0, $0 had the value of $1, etc). This bug was introduced by
6507 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6509 38. Improved the line breaking for long SMTP error messages from ACLs.
6510 Previously, if there was no break point between 40 and 75 characters, Exim
6511 left the rest of the message alone. Two changes have been made: (a) I've
6512 reduced the minimum length to 35 characters; (b) if it can't find a break
6513 point between 35 and 75 characters, it looks ahead and uses the first one
6514 that it finds. This may give the occasional overlong line, but at least the
6515 remaining text gets split now.
6517 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6518 file descriptors might be low, and that setting 1000 would always raise it.
6519 It turns out that in some environments, the limit is already over 1000 and
6520 that lowering it causes trouble. So now Exim takes care not to decrease it.
6522 40. When delivering a message, the value of $return_path is set to $sender_
6523 address at the start of routing (routers may change the value). By an
6524 oversight, this default was not being set up when an address was tested by
6525 -bt or -bv, which affected the outcome if any router or filter referred to
6528 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6529 log only when "message" or "log_message" are set. However, if one of the
6530 conditions was an address verification, or a call to a nested ACL, the
6531 messages generated by the underlying test were being passed through. This
6532 no longer happens. The underlying message is available in $acl_verify_
6533 message for both "message" and "log_message" expansions, so it can be
6534 passed through if needed.
6536 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6537 new expansion $bh_ to give the encoded byte string without charset
6538 translation. Translation happens only if iconv() is available; HAVE_ICONV
6539 indicates this at build time. HEADERS_CHARSET gives the charset to
6540 translate to; headers_charset can change it in the configuration, and
6541 "headers charset" can change it in an individual filter file.
6543 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6544 that creates RFC 2047 encoded "words" labels them as that charset instead
6545 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6546 expansion operator; (ii) when Exim creates a From: line for a local
6547 message; (iii) when a header line is rewritten to include a "phrase" part.
6549 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6550 buggy, causing it to skip the first lines of messages whose message ID
6551 ended in 'D'. This would not have bitten before Exim release 4.14, because
6552 message IDs were unlikely to end in 'D' before then. The effect was to have
6553 incorrect size information for certain domains.
6555 45. #include "config.h" was missing at the start of the crypt16.c module. This
6556 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6559 46. If there was a timeout during a "random" callout check, Exim treated it as
6560 a failure of the random address, and carried on sending RSET and the real
6561 address. If the delay was just some slowness somewhere, the response to the
6562 original RCPT would be taken as a response to RSET and so on, causing
6563 mayhem of various kinds.
6565 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6566 when I implemented it. It didn't allow for the fact that some option values
6567 may legitimately be negative (e.g. size_addition), and it didn't even do
6568 the right test for positive values.
6570 48. Domain names in DNS records are case-independent. Exim always looks them up
6571 in lower case. Some resolvers return domain names in exactly the case they
6572 appear in the zone file, that is, they may contain uppercase letters. Not
6573 all resolvers do this - some return always lower case. Exim was treating a
6574 change of case by a resolver as a change of domain, similar to a widening
6575 of a domain abbreviation. This triggered its re-routing code and so it was
6576 trying to route what was effectively the same domain again. This normally
6577 caused routing to fail (because the router wouldn't handle the domain
6578 twice). Now Exim checks for this case specially, and just changes the
6579 casing of the domain that it ultimately uses when it transmits the message
6582 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6585 50. If a filter generated a file delivery with a non-absolute name (possible if
6586 no home directory exists for the router), the forbid_file option was not
6589 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6590 the existing equality matching.
6592 52. Exim was using ints instead of ino_t variables in some places where it was
6593 dealing with inode numbers.
6595 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6596 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6597 called TMPDIR, and if it finds it is different, it changes its value.
6599 54. The smtp_printf() function is now made available to local_scan() so
6600 additional output lines can be written before returning. There is also an
6601 smtp_fflush() function to enable the detection of a dropped connection.
6602 The variables smtp_input and smtp_batched_input are exported to
6605 55. Changed the default runtime configuration: the message "Unknown user"
6606 has been removed from the ACL, and instead placed on the localuser router,
6607 using the cannot_route_message feature. This means that any verification
6608 failures that generate their own messages won't get overridden. Similarly,
6609 the "Unrouteable address" message that was in the ACL for unverifiable
6610 relay addresses has also been removed.
6612 56. Added hosts_avoid_esmtp to the smtp transport.
6614 57. The exicyclog script was not checking for the esoteric option
6615 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6616 will work only if exicyclog is run under the appropriate euid.
6618 58. Following a discussion on the list, the rules by which Exim recognises line
6619 endings on incoming messages have been changed. The -dropcr and drop_cr
6620 options are now no-ops, retained only for backwards compatibility. The
6621 following line terminators are recognized: LF CRLF CR. However, special
6622 processing applies to CR:
6624 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6625 nor a local message in the state where . is a terminator.
6627 (ii) If a bare CR is encountered in a header line, an extra space is added
6628 after the line terminator so as not to end the header. The reasoning
6629 behind this is that bare CRs in header lines are most likely either
6630 to be mistakes, or people trying to play silly games.
6632 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6633 was being incorrectly given as 18 bytes larger than it should have been.
6634 This is a VOB (very old bug).
6636 60. This may never have affected anything current, but just in case it has:
6637 When the local host is found other than at the start of a list of hosts,
6638 the local host, those with the same MX, and any that follow, are discarded.
6639 When the list in question was part of a longer list of hosts, the following
6640 hosts (not currently being processed) were also being discarded. This no
6641 longer happens. I'm not sure if this situation could ever has previously
6644 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6647 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6648 header. This is something that is recommended in a new Internet Draft, and
6649 is something that is documented as being done by Sendmail. There are two
6650 possible values. For messages generated by the autoreply transport, Exim
6653 Auto-Submitted: auto-replied
6655 whereas for all other generated messages (e.g. bounces) it adds
6657 Auto-Submitted: auto-generated
6659 63. The "personal" condition in filters now includes a test for the
6660 Auto-Submitted: header. If it contains the string "auto-" the message it
6661 not considered personal.
6663 64. Added rcpt_include_affixes as a generic transport option.
6665 65. Added queue_only_override (default true).
6667 66. Added the syslog_duplication option.
6669 67. If what should have been the first header line of a message consisted of
6670 a space followed by a colon, Exim was mis-interpreting it as a header line.
6671 It isn't of course - it is syntactically invalid and should therefore be
6672 treated as the start of the message body. The misbehaviour could have
6673 caused a number of strange effects, including loss of data in subsequent
6674 header lines, and spool format errors.
6676 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6677 client host had authenticated. This control can now be exercised by an ACL
6678 for more flexibility.
6680 69. By default, callouts do not happen when testing with -bh. There is now a
6681 variant, -bhc, which does actually run the callout code, including
6682 consulting and updating the callout cache.
6684 70. Added support for saslauthd authentication, courtesy of Alexander
6687 71. If statvfs() failed on the spool or log directories while checking their
6688 size for availability, Exim confusingly gave the error "space shortage".
6689 Furthermore, in debugging mode it crashed with a floating point exception.
6690 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6691 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6692 serious problem, Exim now writes to the main and panic logs when this
6693 happens, with details of the failure. It then refuses to accept the
6694 incoming message, giving the message "spool directory problem" or "log
6695 directory problem" with a 421 code for SMTP messages.
6697 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6698 0, 1, and 2 exist, because some OS complain for execs without them (see
6699 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6700 descriptors. However, the code omitted to check that the open succeeded,
6701 causing mysterious errors if for some reason the permissions on /dev/null
6702 got screwed. Now Exim writes a message to the main and panic logs, and
6703 bombs out if it can't open /dev/null.
6705 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6706 interact so that it is all more flexible. It is supposed to remain
6707 backwards compatible. Also added extra_local_interfaces.
6709 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6710 to bomb out with an assertion failure - to the client this appears as a
6711 connection drop. This problem occurs in the part of the code that was taken
6712 from the Samba project. Fortunately, the assertion is in a very simple
6713 function, so I have fixed this by reproducing the function inline in the
6714 one place where it is called, and arranging for authentication to fail
6715 instead of killing the process with assert().
6717 75. The SPA client code was not working when the server requested OEM rather
6718 than Unicode encoding.
6720 76. Added code to make require_files with a specific uid setting more usable in
6721 the case where statting the file as root fails - usually a non-root-mounted
6722 NFS file system. When this happens and the failure is EACCES, Exim now
6723 forks a subprocess and does the per-uid checking as the relevant uid.
6725 77. Added process_log_path.
6727 78. If log_file_path was not explicitly set, a setting of check_log_space or
6728 check_log_inodes was ignored.
6730 79. If a space check for the spool or log partitions fails, the incident is now
6731 logged. Of course, in the latter case the data may get lost...
6733 80. Added the %p formatting code to string_format() so that it can be used to
6734 print addresses in debug_print(). Adjusted all the address printing in the
6735 debugging in store.c to use %p rather than %d.
6737 81. There was a concern that a line of code in smtp_in.c could overflow a
6738 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6739 initially expressed, the concern was not well-founded, because trailing
6740 spaces are removed early. However, if the trailing spaces were followed by
6741 a NULL, they did not get removed, so the overflow was possible. Two fixes
6744 (a) I re-wrote the offending code in a cleaner fashion.
6745 (b) If an incoming SMTP command contains a NULL character, it is rejected
6748 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6749 runs initgroups(), so that if the Exim user is in any additional groups,
6750 they will be used during message reception.
6756 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob