1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 cipher, Certificate or OCSP information.
33 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
34 than -2003 did; needs libidn2 in addition to libidn.
36 JH/02 The path option on a pipe transport is now expanded before use.
38 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
39 Patch provided by "Björn", documentation fix added too.
41 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
42 missing a wire-to-host endian conversion.
44 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
45 close after a BDAT command line could be taken as a following command,
46 giving a synch failure. Fix by only checking for synch immediately
47 before acknowledging the chunk.
49 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
50 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
51 macro. Patches provided by Josh Soref.
53 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
54 Previously we did not; the RFC seems ambiguous and VRFY is not listed
55 by IANA as a service extension. However, John Klensin suggests that we
58 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
59 the dkim code may be unix-mode line endings rather than smtp wire-format
60 CRLF, so prepend a CR to any bare LF.
62 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
63 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
65 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
66 the first were themselves being wrongly included in the feed into dkim
67 processing; with most chunk sizes in use this resulted in an incorrect
68 body hash calculated value.
70 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
71 DKIM signature block, for verification. Although advised against by
72 standards it is specifically not ruled illegal.
74 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
76 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
77 missing a body hash (the bh= tag).
79 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
80 It seems that HAProxy sends the Proxy Protocol information in clear and
81 only then does a TLS startup, so do the same.
83 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
84 TCP connections (such as for Spamd) unless the daemon successfully set
85 Fast Open mode on its listening sockets. This fixes breakage seen on
86 too-old kernels or those not configured for Fast Open, at the cost of
87 requiring both directions being enabled for TFO, and TFO never being used
88 by non-daemon-related Exim processes.
90 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
91 endings, at least on the first header line. Try to canonify any that get
92 past that check, despite the cost.
94 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
95 now limited to an arbitrary five deep, while parsing addresses with the
96 strip_excess_angle_brackets option enabled.
98 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
99 instead leave the unprompted TLS handshake in socket buffer for the
100 TLS library to consume.
102 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
104 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
106 JH/16 Drop variables when they go out of scope. Memory management drops a whole
107 region in one operation, for speed, and this leaves assigned pointers
108 dangling. Add checks run only under the testsuite which checks all
109 variables at a store-reset and panics on a dangling pointer; add code
110 explicitly nulling out all the variables discovered. Fixes one known
111 bug: a transport crash, where a dangling pointer for $sending_ip_address
112 originally assigned in a verify callout, is re-used.
114 PP/06 Drop '.' from @INC in various Perl scripts.
116 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
118 PP/08 Reduce a number of compilation warnings under clang; building with
119 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
120 should be warning-free.
122 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
124 HS/01 Fix portability problems introduced by PP/08 for platforms where
125 realloc(NULL) is not equivalent to malloc() [SunOS et al].
127 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
128 chunk. This allows us to accept broken chunked messages. We need a more
129 general solution here.
131 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
132 already-broken messages in the queue.
134 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
136 JH/19 Fix reference counting bug in routing-generated-address tracking.
142 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
143 supports it and a size is available (ie. the sending peer gave us one).
145 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
146 years of being deprecated). The replacements are the ACLs
147 acl_smtp_mime and acl_not_smtp_mime.
149 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
150 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
151 or even an in-clear connection were permitted. Now, if the host lookup
152 was dnssec and dane was requested then the host is only used if the
153 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
154 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
155 if one fails this test.
156 This means that a poorly-configured remote DNS will make it incommunicado;
157 but it protects against a DNS-interception attack on it.
159 JH/04 Bug 1810: make continued-use of an open smtp transport connection
160 non-noisy when a race steals the message being considered.
162 JH/05 If main configuration option tls_certificate is unset, generate a
163 self-signed certificate for inbound TLS connections.
165 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
166 in rewrites and routers.
168 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
169 and logged a warning sing 4.83; now they are a configuration file error.
171 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
172 (lacking @domain). Apply the same qualification processing as RCPT.
174 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
176 JH/10 Support ${sha256:} applied to a string (as well as the previous
179 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
180 a cutthrough deliver is pending, as we always want to make a connection.
181 This also avoids re-routing the message when later placing the cutthrough
182 connection after a verify cache hit.
183 Do not update it with the verify result either.
185 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
186 when routing results in more than one destination address.
188 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
189 signing (which inhibits the cutthrough capability). Previously only
190 the presence of an option was tested; now an expansion evaluating as
191 empty is permissible (obviously it should depend only on data available
192 when the cutthrough connection is made).
194 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
195 the relevant preceding SMTP command did not note the pipelining mode.
197 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
198 Previously they were not counted.
200 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
201 as one having no matching records. Previously we deferred the message
202 that needed the lookup.
204 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
205 distinguished as "(=".
207 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
208 for missing MX records. Previously it only worked for missing A records.
210 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
212 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
213 after the data-go-ahead and data-ack. Patch from Jason Betts.
215 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
216 even for a "none" policy. Patch from Tony Meyer.
218 JH/22 Fix continued use of a connection for further deliveries. If a port was
219 specified by a router, it must also match for the delivery to be
222 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
223 When one of the recipients of a message was unsuitable for the connection
224 (has no matching addresses), we lost track of needing to mark it
225 deferred. As a result mail would be lost.
227 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
229 JH/25 Decoding ACL controls is now done using a binary search; the source code
230 takes up less space and should be simpler to maintain. Merge the ACL
231 condition decode tables also, with similar effect.
233 JH/26 Fix problem with one_time used on a redirect router which returned the
234 parent address unchanged. A retry would see the parent address marked as
235 delivered, so not attempt the (identical) child. As a result mail would
238 JH/27 Fix a possible security hole, wherein a process operating with the Exim
239 UID can gain a root shell. Credit to http://www.halfdog.net/ for
240 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
243 JH/28 Enable {spool,log} filesystem space and inode checks as default.
244 Main config options check_{log,spool}_{inodes,space} are now
245 100 inodes, 10MB unless set otherwise in the configuration.
247 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
248 Previously it only applied to the main-section connection policy
251 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
253 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
254 by me. Added RFC7919 DH primes as an alternative.
256 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
257 are not in the system include path.
259 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
260 GnuTLS, when a session startup failed (eg because the client disconnected)
261 Exim did stdio operations after fclose. This was exposed by a recent
262 change which nulled out the file handle after the fclose.
264 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
265 signed directly by the cert-signing cert, rather than an intermediate
266 OCSP-signing cert. This is the model used by LetsEncrypt.
268 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
270 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
271 an incoming connection.
273 HS/02 Bug 1802: Do not half-close the connection after sending a request
276 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
277 fallback to "prime256v1".
279 JH/34 SECURITY: Use proper copy of DATA command in error message.
280 Could leak key material. Remotely exploitable. CVE-2016-9963.
286 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
287 and 3.4.4 - once the server is enabled to respond to an OCSP request
288 it does even when not requested, resulting in a stapling non-aware
289 client dropping the TLS connection.
291 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
292 support variable-length bit vectors. No functional change.
294 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
295 The I= interface field on outgoing lines is now after the H= remote
296 host field, same as incoming lines. There is a separate
297 outgoing_interface log selector which allows you to disable the
300 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
301 If not running log_selector +smtp_connection the mainlog would be held
302 open indefinitely after a "too many connections" event, including to a
303 deleted file after a log rotate. Leave the per net connection logging
304 leaving it open for efficiency as that will be quickly detected by the
305 check on the next write.
307 HS/01 Bug 1671: Fix post transport crash.
308 Processing the wait-<transport> messages could crash the delivery
309 process if the message IDs didn't exist for some reason. When
310 using 'split_spool_directory=yes' the construction of the spool
311 file name failed already, exposing the same netto behaviour.
313 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
314 mime_regex ACL conditions.
316 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
317 to DSN fail messages (bounces): remote IP, remote greeting, remote response
318 to HELO, local diagnostic string.
320 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
321 log line to debug. Even when configured with a tls authenticator many
322 client connections are expected to not authenticate in this way, so
323 an authenticate fail is not an error.
325 HS/02 Add the Exim version string to the process info. This way exiwhat
326 gives some more detail about the running daemon.
328 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
329 matter for fast-change records such as DNSBLs.
331 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
332 retry record, even if constant. There may be multiple transports with
333 different interface settings and the retry behaviour needs to be kept
336 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
338 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
340 JH/10 Bug 840: fix log_defer_output option of pipe transport
342 JH/11 Bug 830: use same host for all RCPTS of a message, even under
343 hosts_randomize. This matters a lot when combined with mua_wrapper.
345 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
346 ${quote_pgsql:<string>} operator.
348 JH/13 Bug 1708: avoid misaligned access in cached lookup.
350 JH/14 Change header file name for freeradius-client. Relevant if compiling
351 with Radius support; from the Gentoo tree and checked under Fedora.
353 JH/15 Bug 1712: Introduce $prdr_requested flag variable
355 JH/16 Bug 1714: Permit an empty string as expansion result for transport
356 option transport_filter, meaning no filtering.
358 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
360 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
361 defaults to "*" (all hosts). The variable is now available when not built
362 with TLS, default unset, mainly to enable keeping the testsuite sane.
363 If a server certificate is not supplied (via tls_certificate) an error is
364 logged, and clients will find TLS connections fail on startup. Presumably
365 they will retry in-clear.
366 Packagers of Exim are strongly encouraged to create a server certificate
367 at installation time.
369 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
370 with the $config_file variable.
372 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
373 in transport context, after the attempt, and per-recipient. The latter type
374 is per host attempted. The event data is the error message, and the errno
375 information encodes the lookup type (A vs. MX) used for the (first) host,
376 and the trailing two digits of the smtp 4xx response.
378 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
379 to write to mainlog (or rejectlog, paniclog) in the window between file
380 creation and permissions/ownership being changed. Particularly affects
381 installations where exicyclog is run as root, rather than exim user;
382 result is that the running daemon panics and dies.
384 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
386 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
387 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
388 "pri" and "weight". Note that the previous implicit priority given by the
389 list order is no longer honoured.
391 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
394 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
395 by defining SUPPORT_SOCKS.
397 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
398 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
399 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
400 variables are renamed to proxy_{local,external}_{address,port}.
402 JH/27 Move Internationalisation support from Experimental to mainline, enabled
403 for a build by defining SUPPORT_I18N
405 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
406 of the query string, and make ${quote_redis:} do that quoting.
408 JH/29 Move Events support from Experimental to mainline, enabled by default
409 and removable for a build by defining DISABLE_EVENT.
411 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
413 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
414 cached by the daemon.
416 JH/32 Move Redis support from Experimental to mainline, enabled for a build
417 by defining LOOKUP_REDIS. The libhiredis library is required.
419 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
420 keys are given for lookup.
422 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
423 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
424 only supported when built with TLS support. The PolarSSL SHA routines
425 are still used when the TLS library is too old for convenient support.
427 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
428 openssl_options), for security. OpenSSL forces this from version 1.1.0
429 server-side so match that on older versions.
431 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
432 allocation for $value could be released as the expansion processing
433 concluded, but leaving the global pointer active for it.
435 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
436 and to use the domains and local_parts ACL conditions.
438 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
439 incorrectly not doubled on cutthrough transmission, hence seen as a
440 body-termination at the receiving system - resulting in truncated mails.
441 Commonly the sender saw a TCP-level error, and retransmitted the message
442 via the normal store-and-forward channel. This could result in duplicates
443 received - but deduplicating mailstores were liable to retain only the
444 initial truncated version.
446 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
448 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
450 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
451 we're in there, support oversigning also; bug 1309.
453 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
455 HS/04 Add support for keep_environment and add_environment options.
457 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
458 either intentional arithmetic overflow during PRNG, or testing config-
461 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
462 delivery resulted in actual delivery. Cancel cutthrough before DATA
465 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
466 rejects a recipient: pass the reject to the originator.
468 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
469 Many were false-positives and ignorable, but it's worth fixing the
472 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
473 for the new environment-manipulation done at startup. Move the routines
474 from being local to tls.c to being global via the os.c file.
476 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
477 an extract embedded as result-arg for a map, the first arg for extract
478 is unavailable so we cannot tell if this is a numbered or keyed
479 extraction. Accept either.
485 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
488 JH/02 The smtp transport option "multi_domain" is now expanded.
490 JH/03 The smtp transport now requests PRDR by default, if the server offers
493 JH/04 Certificate name checking on server certificates, when exim is a client,
494 is now done by default. The transport option tls_verify_cert_hostnames
495 can be used to disable this per-host. The build option
496 EXPERIMENTAL_CERTNAMES is withdrawn.
498 JH/05 The value of the tls_verify_certificates smtp transport and main options
499 default to the word "system" to access the system default CA bundle.
500 For GnuTLS, only version 3.0.20 or later.
502 JH/06 Verification of the server certificate for a TLS connection is now tried
503 (but not required) by default. The verification status is now logged by
504 default, for both outbound TLS and client-certificate supplying inbound
507 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
510 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
511 Status Notification (bounce) messages are now MIME format per RFC 3464.
512 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
513 under the control of the dsn_advertise_hosts option, and routers may
514 have a dsn_lasthop option.
516 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
517 default, modifiable by a malware= option. The list separator for
518 the options can now be changed in the usual way. Bug 68.
520 JH/10 The smtp_receive_timeout main option is now expanded before use.
522 JH/11 The incoming_interface log option now also enables logging of the
523 local interface on delivery outgoing connections.
525 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
526 if the interface and destination host and port all match.
528 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
531 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
532 Patch from Andrew Lewis.
534 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
535 now supports optional time-restrictions, weighting, and priority
536 modifiers per server. Patch originally by <rommer@active.by>.
538 JH/16 The spamd_address main option now supports a mixed list of local
539 and remote servers. Remote servers can be IPv6 addresses, and
540 specify a port-range.
542 JH/17 Bug 68: The spamd_address main option now supports an optional
543 timeout value per server.
545 JH/18 Bug 1581: Router and transport options headers_add/remove can
546 now have the list separator specified.
548 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
551 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
554 JH/21 Support for the A6 type of dns record is withdrawn.
556 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
557 rather than the verbs used.
559 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
560 from 255 to 1024 chars.
562 JH/24 Verification callouts now attempt to use TLS by default.
564 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
565 are generic router options now. The defaults didn't change.
567 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
568 Original patch from Alexander Shikoff, worked over by JH.
570 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
571 files if ZCAT_COMMAND is not executable.
573 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
575 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
577 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
578 Normally benign, it bites when the pair was led to by a CNAME;
579 modern usage is to not canonicalize the domain to a CNAME target
580 (and we were inconsistent anyway for A-only vs AAAA+A).
582 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
584 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
585 when evaluating $sender_host_dnssec.
587 JH/31 Check the HELO verification lookup for DNSSEC, adding new
588 $sender_helo_dnssec variable.
590 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
592 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
594 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
596 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
597 documented as working, but never had. Support all but $spam_report.
599 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
600 added for tls authenticator.
602 HS/03 Add perl_taintmode main config option
608 TL/01 When running the test suite, the README says that variables such as
609 no_msglog_check are global and can be placed anywhere in a specific
610 test's script, however it was observed that placement needed to be near
611 the beginning for it to behave that way. Changed the runtest perl
612 script to read through the entire script once to detect and set these
613 variables, reset to the beginning of the script, and then run through
614 the script parsing/test process like normal.
616 TL/02 The BSD's have an arc4random API. One of the functions to induce
617 adding randomness was arc4random_stir(), but it has been removed in
618 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
619 function when detected.
621 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
622 cause callback expansion.
624 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
625 syntax errors in an expansion can be treated as a string instead of
626 logging or causing an error, due to the internal use of bool_lax
627 instead of bool when processing it.
629 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
630 server certificates when making smtp deliveries.
632 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
634 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
636 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
638 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
639 Merged patch from Sebastian Wiedenroth.
641 JH/05 Fix results-pipe from transport process. Several recipients, combined
642 with certificate use, exposed issues where response data items split
643 over buffer boundaries were not parsed properly. This eventually
644 resulted in duplicates being sent. This issue only became common enough
645 to notice due to the introduction of connection certificate information,
646 the item size being so much larger. Found and fixed by Wolfgang Breyha.
648 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
649 size buffer was used, resulting in syntax errors when an expansion
652 JH/07 Add support for directories of certificates when compiled with a GnuTLS
653 version 3.3.6 or later.
655 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
656 is EXPERIMENTAL_EVENT, the main-configuration and transport options
657 both become "event_action", the variables become $event_name, $event_data
658 and $event_defer_errno. There is a new variable $verify_mode, usable in
659 routers, transports and related events. The tls:cert event is now also
660 raised for inbound connections, if the main configuration event_action
663 TL/06 In test suite, disable OCSP for old versions of openssl which contained
664 early OCSP support, but no stapling (appears to be less than 1.0.0).
666 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
667 server certificate names available under the smtp transport option
668 "tls_verify_cert_hostname" now do not permit multi-component wildcard
671 JH/10 Time-related extraction expansions from certificates now use the main
672 option "timezone" setting for output formatting, and are consistent
673 between OpenSSL and GnuTLS compilations. Bug 1541.
675 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
676 encoded parameter in the incoming message. Bug 1558.
678 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
679 include certificate info, eximon was claiming there were spoolfile
682 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
684 JH/14 Log delivery-related information more consistently, using the sequence
685 "H=<name> [<ip>]" wherever possible.
687 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
688 are problematic for Debian distribution, omit them from the release
691 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
693 JH/16 Fix string representation of time values on 64bit time_t architectures.
696 JH/17 Fix a null-indirection in certextract expansions when a nondefault
697 output list separator was used.
702 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
703 checkers that were complaining about end of non-void function with no
706 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
707 This was a regression introduced in 4.83 by another bugfix.
709 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
711 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
712 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
718 TF/01 Correctly close the server side of TLS when forking for delivery.
720 When a message was received over SMTP with TLS, Exim failed to clear up
721 the incoming connection properly after forking off the child process to
722 deliver the message. In some situations the subsequent outgoing
723 delivery connection happened to have the same fd number as the incoming
724 connection previously had. Exim would try to use TLS and fail, logging
725 a "Bad file descriptor" error.
727 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
728 utilities have not been installed.
730 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
731 temporary space as the ACL may create new global variables.
733 TL/01 LDAP support uses per connection or global context settings, depending
734 upon the detected version of the libraries at build time.
736 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
737 to extract and use the src ip:port in logging and expansions as if it
738 were a direct connection from the outside internet. PPv2 support was
739 updated based on HAProxy spec change in May 2014.
741 JH/02 Add ${listextract {number}{list}{success}{fail}}.
743 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
744 Properly escape header and check for NULL return.
746 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
749 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
751 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
752 characters in header names, implemented as a verify condition.
753 Contributed by Michael Fischer v. Mollard.
755 TL/05 Rename SPF condition results err_perm and err_temp to standardized
756 results permerror and temperror. Previous values are deprecated but
757 still accepted. In a future release, err_perm and err_temp will be
758 completely removed, which will be a backward incompatibility if the
759 ACL tests for either of these two old results. Patch contributed by
760 user bes-internal on the mailing list.
762 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
764 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
765 selectors, in both main and reject logs.
767 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
770 JH/07 Add malware type "sock" for talking to simple daemon.
772 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
774 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
775 routers/transports under cutthrough routing.
777 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
778 numbers. Touch up "bool" conditional to keep the same definition.
780 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
782 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
784 JH/12 Expand items in router/transport headers_add or headers_remove lists
785 individually rather than the list as a whole. Bug 1452.
787 Required for reasonable handling of multiple headers_ options when
788 they may be empty; requires that headers_remove items with embedded
789 colons must have them doubled (or the list-separator changed).
791 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
792 view the policy declared in the DMARC record. Currently, $dmarc_status
793 is a combined value of both the record presence and the result of the
796 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
798 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
799 dnslookup router and the smtp transport (applying to the forward
802 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
803 of ldap servers used for a specific lookup. Patch provided by Heiko
806 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
807 New variable $lookup_dnssec_authenticated for observability.
809 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
810 Patch submitted by Lars Timman.
812 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
814 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
815 Requires trusted mode and valid format message id, aborts otherwise.
816 Patch contributed by Heiko Schlichting.
818 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
819 certextract with support for various fields. Bug 1358.
821 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
822 is requested by default, modifiable by smtp transport option
825 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
826 operate on certificate variables to give certificate fingerprints
827 Also new ${sha256:cert_variable}.
829 JH/23 The PRDR feature is moved from being Experimental into the mainline.
831 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
832 Christian Aistleitner.
834 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
836 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
837 file. Patch from Wolfgang Breyha.
839 JH/25 Expand the coverage of the delivery $host and $host_address to
840 client authenticators run in verify callout. Bug 1476.
842 JH/26 Port service names are now accepted for tls_on_connect_ports, to
843 align with daemon_smtp_ports. Bug 72.
845 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
846 support and error reporting did not work properly.
848 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
849 and is readable. Patch from Andrew Colin Kissa.
851 TL/14 Enhance documentation of ${run expansion and how it parses the
852 commandline after expansion, particularly in the case when an
853 unquoted variable expansion results in an empty value.
855 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
857 PP/02 Fix internal collision of T_APL on systems which support RFC3123
858 by renaming away from it. Addresses GH issue 15, reported by
861 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
863 TL/15 SECURITY: prevent double expansion in math comparison functions
864 (can expand unsanitized data). Not remotely exploitable.
871 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
873 PP/02 Make -n do something, by making it not do something.
874 When combined with -bP, the name of an option is not output.
876 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
879 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
880 $sender_host_name and config options to manage this, and basic check
883 PP/05 DSCP support for outbound connections and control modifier for inbound.
885 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
886 (Only plugin which currently uses this is kerberos4, which nobody should
887 be using, but we should make it available and other future plugins might
888 conceivably use it, even though it would break NAT; stuff *should* be
889 using channel bindings instead).
891 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
892 name; added for Sendmail compatibility; requires admin caller.
893 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
894 just ignore it); requires trusted caller.
895 Also parse but ignore: -Ac -Am -X<logfile>
898 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
900 TL/02 Add +smtp_confirmation as a default logging option.
902 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
903 Patch by Magnus Holmgren from 2007-02-20.
905 TL/04 Bugzilla 1281 - Spec typo.
906 Bugzilla 1283 - Spec typo.
907 Bugzilla 1290 - Spec grammar fixes.
909 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
911 TL/06 Add Experimental DMARC support using libopendmarc libraries.
913 TL/07 Fix an out of order global option causing a segfault. Reported to dev
914 mailing list by by Dmitry Isaikin.
916 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
918 JH/02 Support "G" suffix to numbers in ${if comparisons.
920 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
922 NM/01 Bugzilla 1197 - Spec typo
923 Bugzilla 1196 - Spec examples corrections
925 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
927 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
928 gnutls_enable_pkcs11, but renamed to more accurately indicate its
931 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
932 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
934 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
935 "acl {{name}{arg}...}", and optional args on acl condition
938 JH/05 Permit multiple router/transport headers_add/remove lines.
940 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
942 JH/07 Avoid using a waiting database for a single-message-only transport.
943 Performance patch from Paul Fisher. Bugzilla 1262.
945 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
948 JH/09 Add $headers_added variable, with content from use of ACL modifier
949 add_header (but not yet added to the message). Bugzilla 199.
951 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
952 Pulled from Bugzilla 817 by Wolfgang Breyha.
954 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
956 (nb: this is the same fix as in Exim 4.80.1)
958 JH/11 Add A= logging on delivery lines, and a client_set_id option on
961 JH/12 Add optional authenticated_sender logging to A= and a log_selector
964 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
966 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
967 advertise SMTP AUTH mechanism to us, instead of a generic
968 protocol violation error. Also, make Exim more robust to bad
969 data from the Dovecot auth socket.
971 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
973 When a queue runner is handling a message, Exim first routes the
974 recipient addresses, during which it prunes them based on the retry
975 hints database. After that it attempts to deliver the message to
976 any remaining recipients. It then updates the hints database using
979 So if a recipient address works intermittently, it can get repeatedly
980 deferred at routing time. The retry hints record remains fresh so the
981 address never reaches the final cutoff time.
983 This is a fairly common occurrence when a user is bumping up against
984 their storage quota. Exim had some logic in its local delivery code
985 to deal with this. However it did not apply to per-recipient defers
986 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
988 This change adds a proper retry rule check during routing so that the
989 final cutoff time is checked against the message's age. We only do
990 this check if there is an address retry record and there is not a
991 domain retry record; this implies that previous attempts to handle
992 the address had the retry_use_local_parts option turned on. We use
993 this as an approximation for the destination being like a local
994 delivery, as in LMTP.
996 I suspect this new check makes the old local delivery cutoff check
997 redundant, but I have not verified this so I left the code in place.
999 TF/02 Correct gecos expansion when From: is a prefix of the username.
1001 Test 0254 submits a message to Exim with the header
1005 When I ran the test suite under the user fanf2, Exim expanded
1006 the header to contain my full name, whereas it should have added
1007 a Resent-Sender: header. It erroneously treats any prefix of the
1008 username as equal to the username.
1010 This change corrects that bug.
1012 GF/01 DCC debug and logging tidyup
1013 Error conditions log to paniclog rather than rejectlog.
1014 Debug lines prefixed by "DCC: " to remove any ambiguity.
1016 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1018 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1019 Bug spotted by Jeremy Harris; was flawed since initial commit.
1020 Would have resulted in OCSP responses post-SNI triggering an Exim
1021 NULL dereference and crash.
1023 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1025 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1026 Bug detection, analysis and fix by Samuel Thibault.
1027 Bugzilla 1331, Debian bug #698092.
1029 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1031 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1032 Server implementation by Todd Lyons, client by JH.
1033 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1034 config variable "prdr_enable" controls whether the server
1035 advertises the facility. If the client requests PRDR a new
1036 acl_data_smtp_prdr ACL is called once for each recipient, after
1037 the body content is received and before the acl_smtp_data ACL.
1038 The client is controlled by both of: a hosts_try_prdr option
1039 on the smtp transport, and the server advertisement.
1040 Default client logging of deliveries and rejections involving
1041 PRDR are flagged with the string "PRDR".
1043 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1044 fclose(). Diagnosis by Todd Lyons.
1046 PP/17 Update configure.default to handle IPv6 localhost better.
1047 Patch by Alain Williams (plus minor tweaks).
1050 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1051 This is now consistent with GnuTLS, and is now documented: the
1052 previous undocumented portable approach to treating the option as
1053 unset was to force an expansion failure. That still works, and
1054 an empty string is now equivalent.
1056 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1057 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1058 not performing validation itself.
1060 PP/20 Added force_command boolean option to pipe transport.
1061 Patch from Nick Koston, of cPanel Inc.
1063 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1066 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1068 PP/21 Fix eximon continuous updating with timestamped log-files.
1069 Broken in a format-string cleanup in 4.80, missed when I repaired the
1070 other false fix of the same issue.
1071 Report and fix from Heiko Schlichting.
1074 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1075 Report from Prashanth Katuri.
1077 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1078 It's SecureTransport, so affects any MacOS clients which use the
1079 system-integrated TLS libraries, including email clients.
1081 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1082 using a MIME ACL for non-SMTP local injection.
1083 Report and assistance in diagnosis by Warren Baker.
1085 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1087 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1089 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1090 last id that failed so it may be referenced in subsequent ACL's.
1092 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1095 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1096 ldap library initialization, allowing self-signed CA's to be
1097 used. Also properly sets require_cert option later in code by
1098 using NULL (global ldap config) instead of ldap handle (per
1099 session). Bug diagnosis and testing by alxgomz.
1101 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1102 the src/util/ subdirectory.
1104 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1105 renamed to Transport Post Delivery Action by Jeremy Harris, as
1108 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1109 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1110 redis_servers = needs to be configured which will be used by the redis
1111 lookup. Patch from Warren Baker, of The Packet Hub.
1113 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1115 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1116 hostname or reverse DNS when processing a host list. Used suggestions
1117 from multiple comments on this bug.
1119 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1121 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1122 Missed a few lines, added it to make the runtest require no keyboard
1125 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1126 contains upper case chars. Make router use caseful_local_part.
1128 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1129 support when GnuTLS has been built with p11-kit.
1135 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1137 This, or similar/improved, will also be change PP/11 of 4.82.
1143 PP/01 Handle short writes when writing local log-files.
1144 In practice, only affects FreeBSD (8 onwards).
1145 Bugzilla 1053, with thanks to Dmitry Isaikin.
1147 NM/01 Bugzilla 949 - Documentation tweak
1149 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1152 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1154 PP/02 Implemented gsasl authenticator.
1156 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1158 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1159 `pkg-config foo` for cflags/libs.
1161 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1162 with rest of GSASL and with heimdal_gssapi.
1164 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1165 `pkg-config foo` for cflags/libs for the TLS implementation.
1167 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1168 properties get this fed in as external SSF. A number of robustness
1169 and debugging improvements to the cyrus_sasl authenticator.
1171 PP/08 cyrus_sasl server now expands the server_realm option.
1173 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1174 Patch by Jeremy Harris.
1176 PP/10 Added dbmjz lookup type.
1178 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1180 PP/12 MAIL args handles TAB as well as SP, for better interop with
1181 non-compliant senders.
1182 Analysis and variant patch by Todd Lyons.
1184 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1185 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1186 Patch from Dirk Mueller <dmueller@suse.com>
1188 PP/13 tls_peerdn now print-escaped for spool files.
1189 Observed some $tls_peerdn in wild which contained \n, which resulted
1190 in spool file corruption.
1192 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1193 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1194 or write after TLS renegotiation, which otherwise led to messages
1197 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1198 as a tracking header (ie: a signed header comes before the signature).
1199 Patch from Wolfgang Breyha.
1201 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1202 comma-sep list; embedded commas doubled.
1204 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1206 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1208 Report and patch from Dmitry Banschikov.
1210 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1211 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1212 failures. We appear to now support TLS1.1+ with Exim.
1214 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1215 lets Exim select keys and certificates based upon TLS SNI from client.
1216 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1217 before an outbound SMTP session. New log_selector, +tls_sni.
1219 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1220 NULL dereference. Report and patch from Alun Jones.
1222 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1223 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1224 Not seeing resolver debug output on NetBSD, but suspect this is a
1225 resolver implementation change.
1227 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1228 Left warnings. Added "eximon gdb" invocation mode.
1230 PP/21 Defaulting "accept_8bitmime" to true, not false.
1232 PP/22 Added -bw for inetd wait mode support.
1234 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1235 locate the relevant includes and libraries. Made this the default.
1237 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1238 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1240 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1241 This may cause build issues on older platforms.
1243 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1244 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1245 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1246 Added SNI support via GnuTLS too.
1247 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1249 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1251 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1252 Applied second patch from Janne, implementing suggestion to default
1253 multiple-strings-in-record handling to match SPF spec.
1255 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1257 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1258 read-only, out of scope).
1259 Patch from Wolfgang Breyha, report from Stuart Northfield.
1261 PP/29 Fix three issues highlighted by clang analyser static analysis.
1262 Only crash-plausible issue would require the Cambridge-specific
1263 iplookup router and a misconfiguration.
1264 Report from Marcin Mirosław.
1266 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1268 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1269 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1270 As part of this, removing so much warning spew let me fix some minor
1271 real issues in debug logging.
1273 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1274 assignment on my part. Fixed.
1276 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1277 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1278 Janne Snabb (who went above and beyond: thank you).
1280 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1281 string otherwise requires a connection and a bunch more work and it's
1282 relatively easy to get wrong. Should also expose TLS library linkage
1285 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1286 64-bit ${eval} (JH/03).
1288 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1289 GNU libc to support some of the 64-bit stuff, should not lead to
1290 conflicts. Defined before os.h is pulled in, so if a given platform
1291 needs to override this, it can.
1293 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1294 protection layer was required, which is not implemented.
1295 Bugzilla 1254, patch from Wolfgang Breyha.
1297 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1298 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1299 tls_dhparam take prime identifiers. Also unbreak combination of
1300 OpenSSL+DH_params+TLSSNI.
1302 PP/39 Disable SSLv2 by default in OpenSSL support.
1308 PP/01 Solaris build fix for Oracle's LDAP libraries.
1309 Bugzilla 1109, patch from Stephen Usher.
1311 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1313 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1316 TF/02 Fix a couple more cases where we did not log the error message
1317 when unlink() failed. See also change 4.74-TF/03.
1319 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1320 lock up or crash if it happened to be inside a call to libc when it
1321 got a SIGUSR1 from exiwhat.
1323 The SIGUSR1 handler appends the current process status to the process
1324 log which is later printed by exiwhat. It used to use the general
1325 purpose logging code to do this, but several functions it calls are
1326 not safe for signals.
1328 The new output code in the SIGUSR1 handler is specific to the process
1329 log, and simple enough that it's easy to inspect for signal safety.
1330 Removing some special cases also simplifies the general logging code.
1331 Removing the spurious timestamps from the process log simplifies
1334 TF/04 Improved ratelimit ACL condition.
1336 The /noupdate option has been deprecated in favour of /readonly which
1337 has clearer semantics. The /leaky, /strict, and /readonly update modes
1338 are mutually exclusive. The update mode is no longer included in the
1339 database key; it just determines when the database is updated. (This
1340 means that when you upgrade Exim will forget old rate measurements.)
1342 Exim now checks that the per_* options are used with an update mode that
1343 makes sense for the current ACL. For example, when Exim is processing a
1344 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1345 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1346 must specify per_mail/readonly. If you omit the update mode it defaults to
1347 /leaky where that makes sense (as before) or /readonly where required.
1349 The /noupdate option is now undocumented but still supported for
1350 backwards compatibility. It is equivalent to /readonly except that in
1351 ACLs where /readonly is required you may specify /leaky/noupdate or
1352 /strict/noupdate which are treated the same as /readonly.
1354 A useful new feature is the /count= option. This is a generalization
1355 of the per_byte option, so that you can measure the throughput of other
1356 aggregate values. For example, the per_byte option is now equivalent
1357 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1359 The per_rcpt option has been generalized using the /count= mechanism
1360 (though it's more complicated than the per_byte equivalence). When it is
1361 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1362 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1363 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1364 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1365 Note that using per_rcpt with a non-readonly update mode in more than
1366 one ACL will cause the recipients to be double-counted. (The per_mail
1367 and per_byte options don't have this problem.)
1369 The handling of very low rates has changed slightly. If the computed rate
1370 is less than the event's count (usually one) then this event is the first
1371 after a long gap. In this case the rate is set to the same as this event's
1372 count, so that the first message of a spam run is counted properly.
1374 The major new feature is a mechanism for counting the rate of unique
1375 events. The new per_addr option counts the number of different
1376 recipients that someone has sent messages to in the last time period. It
1377 behaves like per_rcpt if all the recipient addresses are different, but
1378 duplicate recipient addresses do not increase the measured rate. Like
1379 the /count= option this is a general mechanism, so the per_addr option
1380 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1381 example, measure the rate that a client uses different sender addresses
1382 with the options per_mail/unique=$sender_address. There are further
1383 details in the main documentation.
1385 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1387 TF/06 Removed a few PCRE remnants.
1389 TF/07 Automatically extract Exim's version number from tags in the git
1390 repository when doing development or release builds.
1392 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1393 Bugzilla 879. Patch from Paul Fisher.
1395 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1396 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1399 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1401 PP/05 Variable $av_failed, true if the AV scanner deferred.
1402 Bugzilla 1078. Patch from John Horne.
1404 PP/06 Stop make process more reliably on build failure.
1405 Bugzilla 1087. Patch from Heiko Schlittermann.
1407 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1408 Bugzilla 1089. Patch from Heiko Schlittermann.
1410 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1411 Bugzilla 1131. Patch from Holger Weiß.
1413 PP/09 Handle IPv6 addresses with SPF.
1414 Bugzilla 860. Patch from Wolfgang Breyha.
1416 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1418 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1421 PP/11 match_* no longer expand right-hand-side by default.
1422 New compile-time build option, EXPAND_LISTMATCH_RHS.
1423 New expansion conditions, "inlist", "inlisti".
1425 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1427 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1429 PP/14 fix log_write() format string regression from TF/03.
1430 Bugzilla 1152. Patch from Dmitry Isaikin.
1436 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1438 PP/02 Harmonised TLS library version reporting; only show if debugging.
1439 Layout now matches that introduced for other libraries in 4.74 PP/03.
1441 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1443 PP/04 New "dns_use_edns0" global option.
1445 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1448 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1449 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1451 TK/01 Updated PolarSSL code to 0.14.2.
1452 Bugzilla 1097. Patch from Andreas Metzler.
1454 PP/07 Catch divide-by-zero in ${eval:...}.
1455 Fixes bugzilla 1102.
1457 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1460 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1461 format-string attack -- SECURITY: remote arbitrary code execution.
1463 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1464 time unintentionally subject to list matching rules, letting the header
1465 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1466 arbitrary string expansion). This allowed for information disclosure.
1468 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1469 INT_MIN/-1 -- value coerced to INT_MAX.
1475 NM/01 Workaround for PCRE version dependency in version reporting
1478 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1479 This fixes portability to compilers other than gcc, notably
1480 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1482 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1483 makefiles for portability to HP-UX and POSIX correctness.
1485 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1486 Also via indented variable definition in the Makefile.
1487 (Debugging by Oliver Heesakkers).
1489 PP/02 Restore caching of spamd results with expanded spamd_address.
1490 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1492 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1493 Improves build reliability. Fix from: Frank Elsner
1495 NM/02 Fix wide character breakage in the rfc2047 coding
1496 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1498 NM/03 Allow underscore in dnslist lookups
1499 Fixes bug 1026. Patch from Graeme Fowler
1501 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1502 Code patches from Adam Ciarcinski of NetBSD.
1504 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1507 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1508 is logged, to avoid truncation. Patch from John Horne.
1510 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1511 Patch from Jakob Hirsch.
1513 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1514 SQL string expansion failure details.
1515 Patch from Andrey Oktyabrski.
1517 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1518 Patch from Simon Arlott.
1520 PP/09 New lookups functionality failed to compile on old gcc which rejects
1521 extern declarations in function scope.
1522 Patch from Oliver Fleischmann
1524 PP/10 Use sig_atomic_t for flags set from signal handlers.
1525 Check getgroups() return and improve debugging.
1526 Fixed developed for diagnosis in bug 927 (which turned out to be
1529 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1530 Patch from Mark Zealey.
1532 PP/12 Bugzilla 1056: Improved spamd server selection.
1533 Patch from Mark Zealey.
1535 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1536 Based on patch from Heiko Schlittermann.
1538 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1539 Patch from Uwe Doering, sign-off by Michael Haardt.
1541 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1542 variable declaration deep within a block. Bug and patch from
1545 PP/15 lookups-Makefile IRIX compatibility coercion.
1547 PP/16 Make DISABLE_DKIM build knob functional.
1549 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1550 Patch by Simon Arlott
1552 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1553 variable argument macros. Our copy now differs from upstream.
1559 TF/01 Failure to get a lock on a hints database can have serious
1560 consequences so log it to the panic log.
1562 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1563 controlled using the smtp_confirmation log selector.
1565 TF/03 Include the error message when we fail to unlink a spool file.
1567 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1568 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1569 for maintaining out-of-tree patches for some time.
1571 PP/01 Bugzilla 139: Documentation and portability issues.
1572 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1573 Handle per-OS dynamic-module compilation flags.
1575 PP/02 Let /dev/null have normal permissions.
1576 The 4.73 fixes were a little too stringent and complained about the
1577 permissions on /dev/null. Exempt it from some checks.
1578 Reported by Andreas M. Kirchwitz.
1580 PP/03 Report version information for many libraries, including
1581 Exim version information for dynamically loaded libraries. Created
1582 version.h, now support a version extension string for distributors
1583 who patch heavily. Dynamic module ABI change.
1585 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1586 privilege escalation vulnerability whereby the Exim run-time user
1587 can cause root to append content of the attacker's choosing to
1590 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1593 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1594 If dropping privileges for untrusted macros, we disabled normal logging
1595 on the basis that it would fail; for the Exim run-time user, this is not
1596 the case, and it resulted in successful deliveries going unlogged.
1597 Fixed. Reported by Andreas Metzler.
1603 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1604 only prepend for the Resent-* case. Fixes regression introduced in
1605 Exim 4.70 by NM/22 for Bugzilla 607.
1607 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1608 increasing numbers of administrators be bitten by this.
1610 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1612 PP/03 Bugzilla 994: added openssl_options main configuration option.
1614 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1616 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1618 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1620 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1621 without a peer certificate, leading to a segfault because of an
1622 assumption that peers always have certificates. Be a little more
1623 paranoid. Problem reported by Martin Tscholak.
1625 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1626 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1627 NB: ClamAV planning to remove STREAM in "middle of 2010".
1628 CL also introduces -bmalware, various -d+acl logging additions and
1629 more caution in buffer sizes.
1631 PP/09 Implemented reverse_ip expansion operator.
1633 PP/10 Bugzilla 937: provide a "debug" ACL control.
1635 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1637 PP/12 Bugzilla 973: Implement --version.
1639 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1641 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1643 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1645 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1646 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1647 ignore trailing whitespace.
1649 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1651 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1654 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1655 Notification from Dr Andrew Aitchison.
1657 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1658 ExtendedDetectionInfo response format.
1659 Notification from John Horne.
1661 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1664 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1665 XSL and documented dependency on system catalogs, with examples of how
1668 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1671 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1672 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1673 configuration file which is writeable by the Exim user or group.
1675 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1676 of configuration files to cover files specified with the -C option if
1677 they are going to be used with root privileges, not just the default
1680 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1681 option (effectively making it always true).
1683 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1684 files to be used while preserving root privileges.
1686 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1687 that rogue child processes cannot use them.
1689 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1690 run-time user, instead of root.
1692 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1693 Exim run-time user without dropping privileges.
1695 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1696 result string, instead of calling string_vformat() twice with the same
1699 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1700 for other users. Others should always drop root privileges if they use
1701 -C on the command line, even for a whitelisted configure file.
1703 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1705 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1711 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1712 $data_path, and $header_path variables; fixed documentation bugs and
1715 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1716 exipick to access non-standard spools, including the "frozen" queue
1719 NM/01 Bugzilla 965: Support mysql stored procedures.
1720 Patch from Alain Williams
1722 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1724 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1725 Patch from Andreas Metzler
1727 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1728 Patch from Kirill Miazine
1730 NM/05 Bugzilla 671: Added umask to procmail example.
1732 JJ/03 installed exipick 20100323.0, fixing doc bug
1734 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1735 directory. Notification and patch from Dan Rosenberg.
1737 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1739 TK/02 Improve log output when DKIM signing operation fails.
1741 MH/01 Treat the transport option dkim_domain as a colon separated
1742 list, not as a single string, and sign the message with each element,
1743 omitting multiple occurences of the same signer.
1745 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1746 Bugzilla 985, 986. Patch by Simon Arlott
1748 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1749 Patch by Simon Arlott
1751 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1752 MBX locking. Notification from Dan Rosenberg.
1758 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1760 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1762 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1764 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1766 NM/04 Bugzilla 915: Flag broken perl installation during build.
1772 TK/01 Added patch by Johannes Berg that expands the main option
1773 "spamd_address" if it starts with a dollar sign.
1775 TK/02 Write list of recipients to X-Envelope-Sender header when building
1776 the mbox-format spool file for content scanning (suggested by Jakob
1779 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1780 (http://www.dcc-servers.net/) support via dccifd. Activated by
1781 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1783 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1784 by Mark Daniel Reidel <mr@df.eu>.
1786 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1787 When building exim an external PCRE library is now needed -
1788 PCRE is a system library on the majority of modern systems.
1789 See entry on PCRE_LIBS in EDITME file.
1791 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1792 conversation. Added nologin parameter to request.
1793 Patch contributed by Kirill Miazine.
1795 TF/01 Do not log submission mode rewrites if they do not change the address.
1797 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1799 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1800 log files in place. Contributed by Roberto Lima.
1802 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1804 TF/03 Bugzilla 615: When checking the local_parts router precondition
1805 after a local_part_suffix or local_part_prefix option, Exim now
1806 does not use the address's named list lookup cache, since this
1807 contains cached lookups for the whole local part.
1809 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1810 Robert Millan. Documentation is in experimental-spec.txt.
1812 TF/04 Bugzilla 668: Fix parallel build (make -j).
1814 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1816 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1817 Patch provided by Jan Srzednicki.
1819 TF/05 Leading white space used to be stripped from $spam_report which
1820 wrecked the formatting. Now it is preserved.
1822 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1823 that they are available at delivery time.
1825 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1827 TF/08 TLS error reporting now respects the incoming_interface and
1828 incoming_port log selectors.
1830 TF/09 Produce a more useful error message if an SMTP transport's hosts
1831 setting expands to an empty string.
1833 NM/06 Bugzilla 744: EXPN did not work under TLS.
1834 Patch provided by Phil Pennock.
1836 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1837 Patch provided by Richard Godbee.
1839 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1840 acl_smtp_notquit, added index entry.
1842 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1843 Patch provided by Eugene Bujak.
1845 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1846 accept(). Patch provided by Maxim Dounin.
1848 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1849 Patch provided by Phil Pennock.
1851 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1853 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1854 Patch provided by Brad "anomie" Jorsch.
1856 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1857 Patch provided by Dean Brooks.
1859 TK/05 Add native DKIM support (does not depend on external libraries).
1861 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1862 Patch provided by Graeme Fowler.
1864 NM/16 Bugzilla 851: Documentation example syntax fix.
1866 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1868 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1871 NM/19 Bugzilla 745: TLS version reporting.
1872 Patch provided by Phil Pennock.
1874 NM/20 Bugzilla 167: bool: condition support.
1875 Patch provided by Phil Pennock.
1877 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1878 clients. Patch provided by Phil Pennock.
1880 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1881 Patch provided by Brad "anomie" Jorsch.
1883 NM/23 Bugzilla 687: Fix misparses in eximstats.
1884 Patch provided by Heiko Schlittermann.
1886 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1887 Patch provided by Heiko Schlittermann.
1889 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1890 plus update to original patch.
1892 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1894 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1895 Patch provided by David Brownlee.
1897 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1899 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1901 NM/30 Bugzilla 888: TLS documentation bugfixes.
1903 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1905 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1906 Unlike the original bugzilla I have changed all shell scripts in src tree.
1908 NM/33 Bugzilla 898: Transport filter timeout fix.
1909 Patch by Todd Rinaldo.
1911 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1912 Patch by Serge Demonchaux.
1914 NM/35 Bugzilla 39: Base64 decode bug fixes.
1915 Patch by Jakob Hirsch.
1917 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1919 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1921 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1923 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1929 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1930 ALT-N's libdkim that I have put here:
1931 http://duncanthrax.net/exim-experimental/
1933 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1934 were called 'true' and it seems that C99 defines that as a reserved
1935 keyword to be used with 'bool' variable types. That means you could
1936 not include C99-style headers which use bools without triggering
1937 build errors in sieve.c.
1939 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1940 as mailq or other aliases. Changed the --help handling significantly
1941 to do whats expected. exim_usage() emits usage/help information.
1943 SC/01 Added the -bylocaldomain option to eximstats.
1945 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1947 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1949 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1955 PH/01 Another patch from the Sieve maintainer.
1957 PH/02 When an IPv6 address is converted to a string for single-key lookup
1958 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1959 dots are used instead of colons so that keys in lsearch files need not
1960 contain colons. This was done some time before quoting was made available
1961 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1962 (notated using the quote facility) so as to distinguish them from IPv4
1963 keys. This meant that lookups for IP addresses in host lists did not work
1964 for iplsearch lookups.
1966 This has been fixed by arranging for IPv6 addresses to be expressed with
1967 colons if the lookup type is iplsearch. This is not incompatible, because
1968 previously such lookups could never work.
1970 The situation is now rather anomalous, since one *can* have colons in
1971 ordinary lsearch keys. However, making the change in all cases is
1972 incompatible and would probably break a number of configurations.
1974 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1977 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1978 conversion specification without a maximum field width, thereby enabling
1979 a rogue spamd server to cause a buffer overflow. While nobody in their
1980 right mind would setup Exim to query an untrusted spamd server, an
1981 attacker that gains access to a server running spamd could potentially
1982 exploit this vulnerability to run arbitrary code as the Exim user.
1984 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1985 $primary_hostname instead of what libspf2 thinks the hosts name is.
1987 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1988 a directory entry by the name of the lookup key. Previously, if a
1989 symlink pointed to a non-existing file or a file in a directory that
1990 Exim lacked permissions to read, a lookup for a key matching that
1991 symlink would fail. Now it is enough that a matching directory entry
1992 exists, symlink or not. (Bugzilla 503.)
1994 PH/03 The body_linecount and body_zerocount variables are now exported in the
1997 PH/04 Added the $dnslist_matched variable.
1999 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2000 This means they are set thereafter only if the connection becomes
2003 PH/06 Added the client_condition to authenticators so that some can be skipped
2004 by clients under certain conditions.
2006 PH/07 The error message for a badly-placed control=no_multiline_responses left
2007 "_responses" off the end of the name.
2009 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2011 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2012 (without spaces) instead of just copying the configuration text.
2014 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2016 PH/11 Added $max_received_linelength.
2018 PH/12 Added +ignore_defer and +include_defer to host lists.
2020 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2021 way in which PCRE > 7.0 is built.
2023 PH/14 Implemented queue_only_load_latch.
2025 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2026 MAIL command. The effect was to mangle the value on 64-bit systems.
2028 PH/16 Another patch from the Sieve maintainer.
2030 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2032 PH/18 If a system quota error occurred while trying to create the file for
2033 a maildir delivery, the message "Mailbox is full" was not appended to the
2034 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2035 only to a quota excession during the actual writing of the file.
2037 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2038 characters?) which causes problems in log lines. The DN values are now
2039 passed through string_printing() before being added to log lines.
2041 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2042 and InterBase are left for another time.)
2044 PH/21 Added message_body_newlines option.
2046 PH/22 Guard against possible overflow in moan_check_errorcopy().
2048 PH/23 POSIX allows open() to be a macro; guard against that.
2050 PH/24 If the recipient of an error message contained an @ in the local part
2051 (suitably quoted, of course), incorrect values were put in $domain and
2052 $local_part during the evaluation of errors_copy.
2058 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2059 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2062 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2063 issue a MAIL command.
2065 PH/02 In an ACL statement such as
2067 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2069 if a client was not listed at all, or was listed with a value other than
2070 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2071 the condition was not true (as it should be), so access was not denied.
2072 The bug was that the ! inversion was incorrectly passed on to the second
2073 item. This has been fixed.
2075 PH/03 Added additional dnslists conditions == and =& which are different from
2076 = and & when the dns lookup returns more than one IP address.
2078 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2079 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2081 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2082 FSYNC, which compiles an option called disable_fsync that allows for
2083 bypassing fsync(). The documentation is heavily laced with warnings.
2085 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2087 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2088 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2089 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2090 including adding "make clean"; (3) Added -fPIC when compiling the test
2091 dynamically loaded module, to get rid of a warning.
2093 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2094 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2095 = 0s. The bug is otherwise harmless.
2097 PH/07 There was a bug in the dovecot authenticator such that the value of
2098 $auth1 could be overwritten, and so not correctly preserved, after a
2099 successful authentication. This usually meant that the value preserved by
2100 the server_setid option was incorrect.
2102 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2104 PH/09 Installed PCRE release 7.0.
2106 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2107 run for batched SMTP input. It is now run at the start of every message
2108 in the batch. While fixing this I discovered that the process information
2109 (output by running exiwhat) was not always getting set for -bs and -bS
2110 input. This is fixed, and it now also says "batched" for BSMTP.
2112 PH/11 Added control=no_pipelining.
2114 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2115 patch, slightly modified), and move the expansion of helo_data till after
2116 the connection is made in the smtp transport (so it can use these
2119 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2121 PH/14 Added log_selector = +pid.
2123 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2125 PH/16 Add ${if forany and ${if forall.
2127 PH/17 Added dsn_from option to vary the From: line in DSNs.
2129 PH/18 Flush SMTP output before performing a callout, unless control =
2130 no_callout_flush is set.
2132 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2133 was true (the default) a successful delivery failed to delete the retry
2134 item, thus causing premature timeout of the address. The bug is now
2137 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2139 PH/21 Long custom messages for fakedefer and fakereject are now split up
2140 into multiline responses in the same way that messages for "deny" and
2141 other ACL rejections are.
2143 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2144 with slight modification.
2146 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2147 draft, changing the syntax and factoring some duplicate code".
2149 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2150 for deliveries of the second and subsequent messages over the same SMTP
2153 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2154 ${reduce, with only minor "tidies".
2156 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2158 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2159 expansion side effects.
2161 PH/27 When a message times out after an over-quota error from an Exim-imposed
2162 quota, the bounce message says "mailbox is full". This message was not
2163 being given when it was a system quota that was exceeded. It now should
2166 MH/03 Made $recipients available in local_scan(). local_scan() already has
2167 better access to the recipient list through recipients_list[], but
2168 $recipients can be useful in postmaster-provided expansion strings.
2170 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2171 in the case of a MAIL command with additional options following the
2172 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2173 were accidentally chopped off.
2175 PH/29 SMTP synchronization checks are implemented when a command is read -
2176 there is a check that no more input is waiting when there shouldn't be
2177 any. However, for some commands, a delay in an ACL can mean that it is
2178 some time before the response is written. In this time, more input might
2179 arrive, invalidly. So now there are extra checks after an ACL has run for
2180 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2181 pipelining has not been advertised.
2183 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2185 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2186 returned in the SMTP response when smtp_return_error_details was false.
2187 This has been fixed.
2189 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2190 instead of the C I/O that was originally supplied, because problems were
2191 reported on Solaris.
2193 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2194 Exim which did not show up earlier: it was assuming that a call to
2195 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2196 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2197 was a macro that became an assignment, so it seemed to work. This has
2198 changed to a proper function call with a void return, hence the compile
2199 error. Exim's code has been fixed.
2201 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2204 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2206 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2208 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2209 because it turns out to be too big or there is a timeout) there is no
2210 "Completed" line in the log. When some messages of this type were
2211 selected by exigrep, they were listed as "not completed". Others were
2212 picked up by some special patterns. I have improved the selection
2213 criteria to be more general.
2215 PH/38 The host_find_failed option in the manualroute router can now be set
2216 to "ignore", to completely ignore a host whose IP address cannot be
2217 found. If all hosts are ignored, the behaviour is controlled by the new
2218 host_all_ignored option.
2220 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2221 homing or because of multiple MX records with /mx) generated more than
2222 one IP address, and the following item turned out to be the local host,
2223 all the secondary addresses of the first item were incorrectly removed
2224 from the list, along with the local host and any following hosts (which
2225 is what is supposed to happen).
2227 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2228 whoever called Exim into the -H file. In the case of the daemon it was
2229 behaving confusingly. When first started, it used values for whoever
2230 started the daemon, but after a SIGHUP it used the Exim user (because it
2231 calls itself on a restart). I have changed the code so that it now always
2234 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2235 message are rejected with the same error (e.g. no authentication or bad
2236 sender address), and a DATA command is nevertheless sent (as can happen
2237 with PIPELINING or a stupid MUA), the error message that was given to the
2238 RCPT commands is included in the rejection of the DATA command. This is
2239 intended to be helpful for MUAs that show only the final error to their
2242 PH/42 Another patch from the Sieve maintainer.
2244 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2245 Eximstats - Fixed some broken HTML links and added missing column headers
2247 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2248 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2250 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2252 PH/43 Yet another patch from the Sieve maintainer.
2254 PH/44 I found a way to check for a TCP/IP connection going away before sending
2255 the response to the final '.' that terminates a message, but only in the
2256 case where the client has not sent further data following the '.'
2257 (unfortunately, this is allowed). However, in many cases there won't be
2258 any further data because there won't be any more messages to send. A call
2259 to select() can be used: if it shows that the input is "ready", there is
2260 either input waiting, or the socket has been closed. An attempt to read
2261 the next input character can distinguish the two cases. Previously, Exim
2262 would have sent an OK response which the client would never have see.
2263 This could lead to message repetition. This fix should cure that, at
2264 least in a lot of common cases.
2266 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2267 advertised in response to EHLO.
2273 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2274 fixed by 4.65/MH/01 (is this a record?) are fixed:
2276 (i) An empty string was always treated as zero by the numeric comparison
2277 operators. This behaviour has been restored.
2279 (ii) It is documented that the numeric comparison operators always treat
2280 their arguments as decimal numbers. This was broken in that numbers
2281 starting with 0 were being interpreted as octal.
2283 While fixing these problems I realized that there was another issue that
2284 hadn't been noticed. Values of message_size_limit (both the global option
2285 and the transport option) were treated as octal if they started with 0.
2286 The documentation was vague. These values are now always treated as
2287 decimal, and I will make that clear in the documentation.
2293 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2294 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2297 MH/01 Don't check that the operands of numeric comparison operators are
2298 integers when their expansion is in "skipping" mode (fixes bug
2299 introduced by 4.64-PH/07).
2301 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2302 child addresses, Exim now panics and dies. Previously, because the count
2303 is held in a short int, deliveries were likely to be lost. As such a
2304 large number of recipients for a single message is ridiculous
2305 (performance will be very, very poor), I have chosen to impose a limit
2306 rather than extend the field.
2312 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2313 leftover -K file (the existence of which was triggered by #402).
2314 While we were at it, introduced process PID as part of the -K
2315 filename. This should rule out race conditions when creating
2318 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2319 processing considerably. Previous code took too long for large mails,
2320 triggering a timeout which in turn triggers #401.
2322 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2323 in the DK code in transports.c. sendfile() is not really portable,
2324 hence the _LINUX specificness.
2326 TF/01 In the add_headers option to the mail command in an Exim filter,
2327 there was a bug that Exim would claim a syntax error in any
2328 header after the first one which had an odd number of characters
2331 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2332 callout verification, Exim cached a "reject" for the entire domain. This
2333 is correct for most verifications, but it is not correct for a recipient
2334 verification with use_sender or use_postmaster set, because in that case
2335 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2336 case of MAIL FROM:<> rejection from other early rejections (e.g.
2337 rejection of HELO). When verifying a recipient using a non-null MAIL
2338 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2339 Whatever the result of the callout, the value of the domain cache is
2340 left unchanged (for any other kind of callout, getting as far as trying
2341 RCPT means that the domain itself is ok).
2343 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2346 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2347 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2348 session, the daemon ignores ECONNECT errors and logs others; it now
2349 ignores EPIPE as well.
2351 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2352 (quoted-printable decoding).
2354 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2355 later the small subsequent patch to fix an introduced bug.
2357 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2359 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2361 PH/08 An error is now given if message_size_limit is specified negative.
2363 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2364 to be given (somewhat) arbitrary names.
2366 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2369 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2370 miscellaneous code fixes
2372 PH/10 Added the log_reject_target ACL modifier to specify where to log
2375 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2376 hostname. This is wrong, because it relates to the incoming message (and
2377 probably the interface on which it is arriving) and not to the outgoing
2378 callout (which could be using a different interface). This has been
2379 changed to use the value of the helo_data option from the smtp transport
2380 instead - this is what is used when a message is actually being sent. If
2381 there is no remote transport (possible with a router that sets up host
2382 addresses), $smtp_active_hostname is used.
2384 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2385 tweaks were necessary in order to get it to work (see also 21 below):
2386 (a) The code assumed that strncpy() returns a negative number on buffer
2387 overflow, which isn't the case. Replaced with Exim's string_format()
2389 (b) There were several signed/unsigned issues. I just did the minimum
2390 hacking in of casts. There is scope for a larger refactoring.
2391 (c) The code used strcasecmp() which is not a standard C function.
2392 Replaced with Exim's strcmpic() function.
2393 (d) The code set only $1; it now sets $auth1 as well.
2394 (e) A simple test gave the error "authentication client didn't specify
2395 service in request". It would seem that Dovecot has changed its
2396 interface. Fortunately there's a specification; I followed it and
2397 changed what the client sends and it appears to be working now.
2399 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2402 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2403 address A is aliased to B and C, where B exists and C does not. Without
2404 -v the output is "A verified" because verification stops after a
2405 successful redirection if more than one address is generated. However,
2406 with -v the child addresses are also verified. Exim was outputting "A
2407 failed to verify" and then showing the successful verification for C,
2408 with its parentage. It now outputs "B failed to verify", showing B's
2409 parentage before showing the successful verification of C.
2411 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2412 look up a TXT record in a specific list after matching in a combined
2415 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2416 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2417 they consult the DNS. I had assumed they would set it the way they
2418 wanted; and indeed my experiments on Linux seem to show that in some
2419 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2420 To be on the safe side, however, I have now made the interface to
2421 host_find_byname() similar to host_find_bydns(), with an argument
2422 containing the DNS resolver options. The host_find_byname() function now
2423 sets these options at its start, just as host_find_bydns() does. The smtp
2424 transport options dns_qualify_single and dns_search_parents are passed to
2425 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2426 of host_find_byname() use the default settings of RES_DEFNAMES
2427 (qualify_single) but not RES_DNSRCH (search_parents).
2429 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2430 spool_read_header() do less string testing, by means of a preliminary
2431 switch on the second character of optional "-foo" lines. (This is
2432 overdue, caused by the large number of possibilities that now exist.
2433 Originally there were few.) While I was there, I also converted the
2434 str(n)cmp tests so they don't re-test the leading "-" and the first
2435 character, in the hope this might squeeze out yet more improvement.
2437 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2438 flag allowing group syntax was set by the header_syntax check but not
2439 turned off, possible causing trouble later; (2) The flag was not being
2440 set at all for the header_verify test, causing "group"-style headers to
2441 be rejected. I have now set it in this case, and also caused header_
2442 verify to ignore an empty address taken from a group. While doing this, I
2443 came across some other cases where the code for allowing group syntax
2444 while scanning a header line wasn't quite right (mostly, not resetting
2445 the flag correctly in the right place). These bugs could have caused
2446 trouble for malformed header lines. I hope it is now all correct.
2448 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2449 with the "reply" argument non-NULL. The code, however (which originally
2450 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2451 but it didn't always do it. This confused somebody who was copying the
2452 code for some other use. I have removed all the tests.
2454 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2455 feature that was used to support insecure browsers during the U.S. crypto
2456 embargo. It requires special client support, and Exim is probably the
2457 only MTA that supported it -- and would never use it because real RSA is
2458 always available. This code has been removed, because it had the bad
2459 effect of slowing Exim down by computing (never used) parameters for the
2460 RSA_EXPORT functionality.
2462 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2463 authenticator to fail if there's a tab character in the incoming data
2464 (there should never be unless someone is messing about, as it's supposed
2465 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2466 if the connection is using TLS or if the remote IP is the same as the
2467 local IP, and the "valid-client-cert option" if a client certificate has
2470 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2471 authenticators. This can be used for authorization after authentication
2472 succeeds. (In the case of plaintext, it servers for both authentication
2475 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2476 if any retry times were supplied.
2478 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2479 connection, where there is no client IP address to check. In this
2480 situation, the verify now always succeeds.
2482 PH/25 Applied John Jetmore's -Mset patch.
2484 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2486 PH/27 In a string expansion for a processed (not raw) header when multiple
2487 headers of the same name were present, leading whitespace was being
2488 removed from all of them, but trailing whitespace was being removed only
2489 from the last one. Now trailing whitespace is removed from each header
2490 before concatenation. Completely empty headers in a concatenation (as
2491 before) are ignored.
2493 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2494 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2496 PH/29 [Removed. This was a change that I later backed out, and forgot to
2497 correct the ChangeLog entry (that I had efficiently created) before
2498 committing the later change.]
2500 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2501 address errors (4xx response to RCPT) over the same connection as other
2502 messages routed to the same hosts. Such deliveries are always "forced",
2503 so retry times are not inspected. This resulted in far too many retries
2504 for the affected addresses. The effect occurred only when there were more
2505 hosts than the hosts_max_try setting in the smtp transport when it had
2506 the 4xx errors. Those hosts that it had tried were not added to the list
2507 of hosts for which the message was waiting, so if all were tried, there
2508 was no problem. Two fixes have been applied:
2510 (i) If there are any address or message errors in an SMTP delivery, none
2511 of the hosts (tried or untried) are now added to the list of hosts
2512 for which the message is waiting, so the message should not be a
2513 candidate for sending over the same connection that was used for a
2514 successful delivery of some other message. This seems entirely
2515 reasonable: after all the message is NOT "waiting for some host".
2516 This is so "obvious" that I'm not sure why it wasn't done
2517 previously. Hope I haven't missed anything, but it can't do any
2518 harm, as the worst effect is to miss an optimization.
2520 (ii) If, despite (i), such a delivery is accidentally attempted, the
2521 routing retry time is respected, so at least it doesn't keep
2522 hammering the server.
2524 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2525 in ${readsocket because some servers need this prod.
2527 PH/32 Added some extra debug output when updating a wait-xxx database.
2529 PH/33 The hint "could be header name not terminated by colon", which has been
2530 given for certain expansion errors for a long time, was not being given
2531 for the ${if def:h_colon_omitted{... case.
2533 PH/34 The spec says: "With one important exception, whenever a domain list is
2534 being scanned, $domain contains the subject domain." There was at least
2535 one case where this was not true.
2537 PH/35 The error "getsockname() failed: connection reset by peer" was being
2538 written to the panic log as well as the main log, but it isn't really
2539 panic-worthy as it just means the connection died rather early on. I have
2540 removed the panic log writing for the ECONNRESET error when getsockname()
2543 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2544 runs only) independently of the message's sender address. This meant
2545 that, if the 4xx error was in fact related to the sender, a different
2546 message to the same recipient with a different sender could confuse
2547 things. In particular, this can happen when sending to a greylisting
2548 server, but other circumstances could also provoke similar problems.
2549 I have changed the default so that the retry time for these errors is now
2550 based a combination of the sender and recipient addresses. This change
2551 can be overridden by setting address_retry_include_sender=false in the
2554 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2555 remote server are returned as part of bounce messages. This was not
2556 happening for LMTP over a pipe (the lmtp transport), but now it is the
2557 same for both kinds of LMTP.
2559 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2560 in header lines that were in fact CNAMEs. This is no longer the case.
2562 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2563 and queue runs started by the daemon processed all messages. This has
2564 been fixed so that -R and -S can now usefully be given with -q<time>.
2566 PH/40 Import PCRE release 6.7 (fixes some bugs).
2568 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2570 PH/42 Give an error if -q is specified more than once.
2572 PH/43 Renamed the variables $interface_address and $interface_port as
2573 $received_ip_address and $received_port, to make it clear that these
2574 values apply to message reception, and not to the outgoing interface when
2575 a message is delivered. (The old names remain recognized, of course.)
2577 PH/44 There was no timeout on the connect() call when using a Unix domain
2578 socket in the ${readsocket expansion. There now is.
2580 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2581 be meaningful with "accept".
2583 SC/01 Eximstats V1.43
2584 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2586 SC/02 Eximstats V1.44
2587 Use a glob alias rather than an array ref in the generated
2588 parser. This improves both readability and performance.
2590 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2591 Collect SpamAssassin and rejection statistics.
2592 Don't display local sender or destination tables unless
2593 there is data to show.
2594 Added average volumes into the top table text output.
2596 SC/04 Eximstats V1.46
2597 Collect data on the number of addresses (recipients)
2598 as well as the number of messages.
2600 SC/05 Eximstats V1.47
2601 Added 'Message too big' to the list of mail rejection
2602 reasons (thanks to Marco Gaiarin).
2604 SC/06 Eximstats V1.48
2605 Mainlog lines which have GMT offsets and are too short to
2606 have a flag are now skipped.
2608 SC/07 Eximstats V1.49 (Alain Williams)
2609 Added the -emptyok flag.
2611 SC/08 Eximstats V1.50
2612 Fixes for obtaining the IP address from reject messages.
2614 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2615 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2616 whitespace changes from 4.64-PH/27
2618 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2621 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2622 are found, allow negative numbers in numeric criteria)
2624 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2626 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2627 to match changes made in 4.64-PH/43
2629 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2631 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2632 "domain" is actually an IP address, and force a failure. This locks out
2633 those revolvers/nameservers that support "A-for-A" lookups, in
2634 contravention of the specifications.
2636 PH/48 When a host name was looked up from an IP address, and the subsequent
2637 forward lookup of the name timed out, the host name was left in
2638 $sender_host_name, contrary to the specification.
2640 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2641 restricted to single-key lookups, Exim was not diagnosing an error if
2642 * or *@ was used with a query-style lookup.
2644 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2646 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2647 long ago, but noone interested enough thought of it. Let's just say that
2648 the "1.1" means that there are some new functions that weren't there at
2649 some point in the past.
2651 PH/51 Error processing for expansion failure of helo_data from an smtp
2652 transport during callout processing was broken.
2654 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2655 tested/used via the -bh/-bhc/-bs options.
2657 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2658 bug, fixed in subsequent PCRE releases).
2660 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2661 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2663 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2669 SC/01 Use a glob alias rather than an array ref in eximstats generated
2670 parser. This improves both readability and performance.
2672 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2673 Don't display local sender or destination tables in eximstats unless
2674 there is data to show.
2675 Added average volumes into the eximstats top table text output.
2677 SC/03 Collect data on the number of addresses (recipients) as well
2678 as the number of messages in eximstats.
2680 TF/01 Correct an error in the documentation for the redirect router. Exim
2681 does (usually) call initgroups() when daemonizing.
2683 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2684 with consistent privilege compared to when running as a daemon.
2686 TF/03 Note in the spec that $authenticated_id is not set for local
2687 submissions from trusted users.
2689 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2690 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2692 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2693 by adding some example configuration directives to the default
2694 configuration file. A little bit of work is required to uncomment the
2695 directives and define how usernames and passwords are checked, but
2696 there is now a framework to start from.
2698 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2699 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2700 without this. I don't know how relevant this is to other LDAP libraries.
2702 PH/02 Add the verb name to the "unknown ACL verb" error.
2704 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2706 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2708 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2709 directory not expanded when it should be if an expanded home directory
2710 was set for the address (which is overridden by the transport).
2712 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2715 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2716 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2717 because it is too late at that time, and has no effect.
2719 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2720 security issue with \' (bugzilla #107). I could not use the
2721 PQescapeStringConn() function, because it needs a PGconn value as one of
2724 PH/08 When testing addresses using -bt, indicate those final addresses that
2725 are duplicates that would not cause an additional delivery. At least one
2726 person was confused, thinking that -bt output corresponded to deliveries.
2727 (Suppressing duplicates isn't a good idea as you lose the information
2728 about possibly different redirections that led to the duplicates.)
2730 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2731 systems where poll() doesn't work, in particular OS X.
2733 PH/10 Added more information to debugging output for retry time not reached.
2735 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2736 operations in malware.c.
2738 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2741 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2742 syslog_duplication set false, log lines that would normally be written
2743 both the the main log and to the reject log were not written to syslog at
2746 PH/14 In the default configuration, change the use of "message" in ACL warn
2747 statements to "add_header".
2749 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2750 not followed by a command (e.g. "seen endif").
2752 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2753 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2756 PH/17 Added extra conditions to the default value of delay_warning_condition
2760 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2761 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2762 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2765 The Auto-Submitted: and various List- headers are standardised, whereas I
2766 don't think Precedence: ever was.
2768 PH/18 Refactored debugging code in route_finduser() to show more information,
2769 in particular, the error code if getpwnam() issues one.
2771 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2772 This is apparently needed in addition to the PH/07 change above to avoid
2773 any possible encoding problems.
2775 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2776 but not after initializing Perl.
2778 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2779 output them only if debugging. By default they are written stderr,
2780 apparently, which is not desirable.
2782 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2785 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2788 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2790 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2791 authenticated or an ident call has been made. Suppress the default
2792 values for $authenticated_id and $authenticated_sender (but permit -oMai
2793 and -oMas) when testing with -bh.
2795 PH/24 Re-jigged the order of the tests in the default configuration so that the
2796 tests for valid domains and recipients precede the DNS black list and CSA
2797 tests, on the grounds that those ones are more expensive.
2799 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2800 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2801 This bug exists in every version of Exim that I still have, right back to
2804 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2805 However, an attempt to turn on case-sensitivity in a regex key by
2806 including (?-i) didn't work because the subject string was already
2807 lowercased, and the effects were non-intuitive. It turns out that a
2808 one-line patch can be used to allow (?-i) to work as expected.
2814 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2815 other effects) broke the use of negated acl sub-conditions.
2817 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2820 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2821 "Deny" causes Exim to reject the incoming connection with a 554 error.
2822 Unfortunately, if there is a major crisis, such as a disk failure,
2823 tcp-wrappers gives "deny", whereas what one would like would be some
2824 kind of temporary error. A kludge has been added to help with this.
2825 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2826 554 error is used if errno is still zero or contains ENOENT (which occurs
2827 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2830 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2832 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2833 errors. Otherwise a message that provokes a temporary error (when other
2834 messages do not) can cause a whole host to time out.
2836 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2837 addresses were routed directly to files or pipes from a redirect router.
2838 File deliveries just didn't batch; pipe deliveries might have suffered
2841 PH/06 A failure to get a lock for a hints database would erroneously always say
2842 "Failed to get write lock", even when it was really a read lock.
2844 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2845 of 0600. This has been changed to use the value of the lockfile_mode
2846 option (which defaults to 0600).
2848 PH/08 Applied small patch from the Sieve maintainer.
2850 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2851 folder from quota calculations, a direct delivery into this folder messed
2852 up the contents of the maildirsize file. This was because the regex was
2853 used only to exclude .Trash (or whatever) when the size of the mailbox
2854 was calculated. There was no check that a delivery was happening into an
2855 excluded directory. This bug has been fixed by ignoring all quota
2856 processing for deliveries into excluded directories.
2858 PH/10 Added the maildirfolder_create_regex option to appendfile.
2864 PH/01 The code for finding all the local interface addresses on a FreeBSD
2865 system running IPv6 was broken. This may well have applied to all BSD
2866 systems, as well as to others that have similar system calls. The broken
2867 code found IPv4 interfaces correctly, but gave incorrect values for the
2868 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2869 that it would not match correctly against @[] and not recognize the IPv6
2872 PH/02 The ipliteral router was not recognizing addresses of the form user@
2873 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2875 PH/03 Added disable_ipv6.
2877 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2878 decoded form, because it is most often used to construct To: headers
2879 lines in autoreplies, and the decoded form may well be syntactically
2880 invalid. However, $reply_address has leading white space removed, and all
2881 newlines turned into spaces so that the autoreply transport does not
2884 PH/05 If group was specified without a user on a router, and no group or user
2885 was specified on a transport, the group from the router was ignored.
2887 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2888 for visible compile-time settings that can be used to change these
2889 numbers, for those that want even more. Backwards compatibility with old
2890 spool files has been maintained. However, going back to a previous Exim
2891 release will lost any variables that are in spool files.
2893 PH/07 Two small changes when running in the test harness: increase delay when
2894 passing a TCP/IP connection to a new process, in case the original
2895 process has to generate a bounce, and remove special handling of
2896 127.0.0.2 (sic), which is no longer necessary.
2898 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2899 be the same on different OS.
2901 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2904 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2905 whether --show-vars was specified or not
2907 JJ/02 exipick: Added support for new ACL variable spool format introduced
2910 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2911 syntactically invalid From: or Reply-to: line, and a filter used this to
2912 generate an autoreply, and therefore failed to obtain an address for the
2913 autoreply, Exim could try to deliver to a non-existent relative file
2914 name, causing unrelated and misleading errors. What now happens is that
2915 it logs this as a hard delivery error, but does not attempt to create a
2918 PH/11 The exinext utility has a -C option for testing purposes, but although
2919 the given file was scanned by exinext itself; it wasn't being passed on
2920 when Exim was called.
2922 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2923 an end-of-file indication when reading a command response.
2925 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2926 compiled. In many other places in Exim, IPv6 addresses are always
2927 recognized, so I have changed this. It also means that IPv4 domain
2928 literals of the form [IPV4:n.n.n.n] are now always recognized.
2930 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2931 used if the router is not running as root, for example, when verifying at
2932 ACL time, or when using -bh. The debugging output from this situation was
2933 non-existent - all you got was a failure to exec. I have made two
2936 (a) Failures to set uid/gid, the current directory, or a process leader
2937 in a subprocess such as that created by queryprogram now generate
2938 suitable debugging output when -d is set.
2940 (b) The queryprogram router detects when it is not running as root,
2941 outputs suitable debugging information if -d is set, and then runs
2942 the subprocess without attempting to change uid/gid.
2944 PH/15 Minor change to Makefile for building test_host (undocumented testing
2947 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2948 additional section of a DNS packet that returns MX or SRV records.
2949 Instead, it always explicitly searches for A/AAAA records. This avoids
2950 major problems that occur when a DNS server includes only records of one
2951 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2952 fixed another bug: if SRV records were looked up and the corresponding
2953 address records were *not* found in the additional section, the port
2954 values from the SRV records were lost.
2956 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2957 using the correct key (the original address) when searching the retry
2958 rules in order to find which one to use for generating the retry hint.
2960 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2961 adding the default one. Similarly, if it contains a Reply-To: header, the
2962 errors_reply_to option, if set, is not used.
2964 PH/19 When calculating a retry time, Exim used to measure the "time since
2965 failure" by looking at the "first failed" field in the retry record. Now
2966 it does not use this if it is later than than the arrival time of the
2967 message. Instead it uses the arrival time. This makes for better
2968 behaviour in cases where some deliveries succeed, thus re-setting the
2969 "first failed" field. An example is a quota failure for a huge message
2970 when small messages continue to be delivered. Without this change, the
2971 "time since failure" will always be short, possible causing more frequent
2972 delivery attempts for the huge message than are intended.
2973 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2975 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2976 $1, $2, $3) because the numerical variables can be reset during some
2977 expansion items (e.g. "match"), thereby losing the authentication data.
2979 PH/21 Make -bV show the size of off_t variables so that the test suite can
2980 decide whether to run tests for quotas > 2G.
2982 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2983 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2984 filecount value is greater than 2G or if a quota value is greater than 2G
2985 on a system where the size of off_t is not greater than 4, a panic error
2988 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2989 never match. The debug and -bh output now contains an explicit error
2990 message indicating a malformed IPv4 address or mask.
2992 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2993 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2994 PH/23 above applies.
2996 PH/25 Do not write to syslog when running in the test harness. The only
2997 occasion when this arises is a failure to open the main or panic logs
2998 (for which there is an explicit test).
3000 PH/26 Added the /no_tell option to "control=freeze".
3002 PH/27 If a host name lookup failed very early in a connection, for example, if
3003 the IP address matched host_lookup and the reverse lookup yielded a name
3004 that did not have a forward lookup, an error message of the form "no IP
3005 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3006 could be logged. Now it outputs the IP address instead of "NULL".
3008 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3009 allows the sender and the authenticated sender to be set when
3010 submitting a message from within Exim. Since child_open_exim() is
3011 documented for local_scan(), the new function should be too.
3013 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3014 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3015 results in an empty string is now treated as unset.
3017 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3019 PH/31 Added sender_verify_fail logging option.
3021 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3022 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3023 tidied the source and removed it altogether.
3025 PH/33 When a queue run was abandoned because the load average was too high, a
3026 log line was always written; now it is written only if the queue_run log
3027 selector is set. In addition, the log line for abandonment now contains
3028 information about the queue run such as the pid. This is always present
3029 in "start" and "stop" lines but was omitted from the "abandon" line.
3031 PH/34 Omit spaces between a header name and the colon in the error message that
3032 is given when verify = headers_syntax fails (if there are lots of them,
3033 the message gets confusing).
3035 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3036 names, as there are now some PTR records that contain slashes. This check
3037 is only to protect against broken name servers that fall over on strange
3038 characters, so the fact that it applies to all lookups doesn't matter.
3040 PH/36 Now that the new test suite is complete, we can remove some of the
3041 special code in Exim that was needed for the old test suite. For example,
3042 sorting DNS records because real resolvers return them in an arbitrary
3043 order. The new test suite's fake resolver always returns records in the
3046 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3047 bounces) except when queue_only is set, to avoid logging races between
3048 the different processes.
3050 PH/38 Panic-die if .include specifies a non-absolute path.
3052 PH/39 A tweak to the "H" retry rule from its user.
3054 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3055 a label. They prevented compilation on older perls.
3057 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3058 a warning to be raised on newish perls.
3060 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3061 on queue. Changes to match documented behaviour of showing count of
3062 messages matching specified criteria.
3064 PH/40 Changed the default ident timeout from 30s to 5s.
3066 PH/41 Added support for the use of login_cap features, on those BSD systems
3067 that have them, for controlling the resources used by pipe deliveries.
3069 PH/42 The content-scanning code uses fopen() to create files in which to put
3070 message data. Previously it was not paying any attention to the mode of
3071 the files. Exim runs with umask(0) because the rest of the code creates
3072 files with open(), and sets the required mode explicitly. Thus, these
3073 files were ending up world-writeable. This was not a big issue, because,
3074 being within the spool directory, they were not world-accessible. I have
3075 created a function called modefopen, which takes an additional mode
3076 argument. It sets umask(777), creates the file, chmods it to the required
3077 mode, then resets the umask. All the relevant calls to fopen() in the
3078 content scanning code have been changed to use this function.
3080 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3081 to 24 hours. This avoids potential overflow problems when processing G
3082 and H retry rules. I suspect nobody ever tinkers with this value.
3084 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3086 PH/45 When the plaintext authenticator is running as a client, the server's
3087 challenges are checked to ensure they are valid base64 strings. By
3088 default, the authentication attempt is cancelled if an invalid string is
3089 received. Setting client_ignore_invalid_base64 true ignores these errors.
3090 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3091 they are received. Thus, the responses can be made to depend on the
3092 challenges. If an invalid string is ignored, an empty string is placed in
3095 PH/46 Messages that are created by the autoreply transport now contains a
3096 References: header, in accordance with RFCs 2822 and 3834.
3098 PH/47 Added authenticated_sender_force to the smtp transport.
3100 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3102 PH/49 Installed latest patch from the Sieve maintainer.
3104 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3105 was also set, the appendfile transport was unnecessarily scanning a
3106 directory of message files (e.g. for maildir delivery) to find the count
3107 of files (along with the size), even though it did not need this
3108 information. It now does the scan only if it needs to find either the
3109 size of the count of files.
3111 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3113 PH/52 Two bugs concerned with error handling when the smtp transport is
3116 (i) Exim was not creating retry information for temporary errors given
3117 for individual recipients after the DATA command when the smtp transport
3118 was used in LMTP mode. This meant that they could be retried too
3119 frequently, and not timed out correctly.
3121 (ii) Exim was setting the flag that allows error details to be returned
3122 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3123 recipients that were returned after the DATA command.
3125 PH/53 This is related to PH/52, but is more general: for any failing address,
3126 when detailed error information was permitted to be returned to the
3127 sender, but the error was temporary, then after the final timeout, only
3128 "retry timeout exceeded" was returned. Now it returns the full error as
3129 well as "retry timeout exceeded".
3131 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3132 do this, and (what is worse) MTAs that accept it.
3134 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3135 will now be deprecated.
3137 PH/56 New os.c-cygwin from the Cygwin maintainer.
3139 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3140 formats (previously only available in exim formats via -bpr, -bpru,
3141 and -bpra. Now also available in native and exiqgrep formats)
3143 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3144 with very large, slow to parse queues
3146 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3148 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3150 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3151 responses to authentication challenges, though it was showing the
3152 challenges; (ii) I've removed the CR characters from the debug output for
3155 PH/58 Allow for the insertion of a newline as well as a space when a string
3156 is turned into more than one encoded-word during RFC 2047 encoding. The
3157 Sieve code now uses this.
3159 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3160 data_4xx, lost_connection, tls_required.
3162 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3163 message was being sent as an SMTP response.
3165 PH/61 Add -l and -k options to exicyclog.
3167 PH/62 When verifying, if an address was redirected to one new address, so that
3168 verification continued, and the new address failed or deferred after
3169 having set something in $address_data, the value of $address_data was not
3170 passed back to the ACL. This was different to the case when no
3171 redirection occurred. The value is now passed back in both cases.
3173 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3174 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3175 use login_cap.h, so on its own it isn't the distinguishing feature. The
3176 new name refers directly to the setclassresources() function.
3178 PH/65 Added configuration files for NetBSD3.
3180 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3182 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3183 is preferred over IPv4.
3185 PH/68 The bounce_return_message and bounce_return_body options were not being
3186 honoured for bounces generated during the reception of non-SMTP messages.
3187 In particular, this applied to messages rejected by the ACL. This bug has
3188 been fixed. However, if bounce_return_message is true and bounce_return_
3189 body is false, the headers that are returned for a non-SMTP message
3190 include only those that have been read before the error was detected.
3191 (In the case of an ACL rejection, they have all been read.)
3193 PH/69 The HTML version of the specification is now built in a directory called
3194 spec_html instead of spec.html, because the latter looks like a path with
3195 a MIME-type, and this confuses some software.
3197 PH/70 Catch two compiler warnings in sieve.c.
3199 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3200 function verify_get_ident() calls ip_connect() to connect a socket, but
3201 if the "connect()" function timed out, ip_connect() used to close the
3202 socket. However, verify_get_ident() also closes the socket later, and in
3203 between Exim writes to the log, which may get opened at this point. When
3204 the socket was closed in ip_connect(), the log could get the same file
3205 descriptor number as the socket. This naturally causes chaos. The fix is
3206 not to close the socket in ip_connect(); the socket should be closed by
3207 the function that creates it. There was only one place in the code where
3208 this was missing, in the iplookup router, which I don't think anybody now
3209 uses, but I've fixed it anyway.
3211 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3212 well as to direct DNS lookups. Otherwise the handling of names in host
3213 lists is inconsistent and therefore confusing.
3219 PH/01 Two changes to the default runtime configuration:
3221 (1) Move the checks for relay_from_hosts and authenticated clients from
3222 after to before the (commented out) DNS black list checks.
3224 (2) Add control=submission to the relay_from_hosts and authenticated
3225 clients checks, on the grounds that messages accepted by these
3226 statements are most likely to be submissions.
3228 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3230 (1) Generate an error if the third argument for the ${prvs expansion is
3233 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3236 (3) Reset the variables that are obtained from the first argument of
3237 ${prvscheck and used in the second argument before leaving the code,
3238 because their memory is reclaimed, so using them afterwards may do
3241 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3242 one (it's much easier than Tom thought :-).
3244 (5) Because of (4), we can now allow for the use of $prvscheck_result
3245 inside the third argument.
3247 PH/03 For some reason, the default setting of PATH when running a command from
3248 a pipe transport was just "/usr/bin". I have changed it to
3251 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3252 anything to be listed in the output from -bV.
3254 PH/05 When a filter generated an autoreply, the entire To: header line was
3255 quoted in the delivery log line, like this:
3257 => >A.N.Other <ano@some.domain> <original@ddress> ...
3259 This has been changed so that it extracts the operative address. There
3260 may be more than one such address. If so, they are comma-separated, like
3263 => >ano@some.domain,ona@other.domain <original@ddress> ...
3265 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3266 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3267 not being looked up in the rDNS to get a host name, Exim was showing the
3268 IP address twice in Received: lines, even though the IP addresses were
3269 identical. For example:
3271 Received: from [1.2.3.4] (helo=[1.2.3.4])
3273 However, if the real host name was known, it was omitting the HELO data
3274 if it matched the actual IP address. This has been tidied up so that it
3275 doesn't show the same IP address twice.
3277 PH/07 When both +timestamp and +memory debugging was on, the value given by
3278 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3279 called by the debug printing, thereby overwriting the timestamp buffer.
3280 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3282 PH/08 When the original message was included in an autoreply transport, it
3283 always said "this is a copy of the message, including all the headers",
3284 even if body_only or headers_only was set. It now gives an appropriate
3287 PH/09 Applied a patch from the Sieve maintainer which:
3289 o fixes some comments
3290 o adds the (disabled) notify extension core
3291 o adds some debug output for the result of if/elsif tests
3292 o points to the current vacation draft in the documentation
3293 and documents the missing references header update
3297 o fixes a bug in processing the envelope test (when testing
3298 multiple envelope elements, the last element determined the
3301 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3302 Electronic Mail") by including:
3304 Auto-submitted: auto-generated
3306 in the messages that it generates (bounce messages and others, such as
3307 warnings). In the case of bounce messages for non-SMTP messages, there was
3308 also a typo: it was using "Auto_submitted" (underscore instead of
3309 hyphen). Since every message generated by Exim is necessarily in response
3310 to another message, thes have all been changed to:
3312 Auto-Submitted: auto-replied
3314 in accordance with these statements in the RFC:
3316 The auto-replied keyword:
3318 - SHOULD be used on messages sent in direct response to another
3319 message by an automatic process,
3321 - MUST NOT be used on manually-generated messages,
3323 - MAY be used on Delivery Status Notifications (DSNs) and Message
3324 Disposition Notifications (MDNs),
3326 - MUST NOT be used on messages generated by automatic or periodic
3327 processes, except for messages which are automatic responses to
3330 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3331 to the default Received: header definition.
3333 PH/12 Added log selector acl_warn_skipped (default on).
3335 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3336 variables because (a) they are in the wrong storage pool and (b) even if
3337 they were copied, it wouldn't work properly because of the caching.
3339 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3340 checking when decoding. Apparently there are clients that generate
3341 overlong encoded strings. Why am I not surprised?
3343 PH/15 If the first argument of "${if match_address" was not empty, but did not
3344 contain an "@" character, Exim crashed. Now it writes a panic log message
3345 and treats the condition as false.
3347 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3349 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3350 extension "envelope-auth". The code is finished and in agreement with
3351 other implementations, but there is no documentation so far and in fact,
3352 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3353 not changing the active code.
3355 Print executed "if" and "elsif" statements when debugging is used. This
3356 helps a great deal to understand what a filter does.
3358 Document more things not specified clearly in RFC3028. I had all this
3359 sorted out, when out of a sudden new issues came to my mind. Oops."
3361 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3364 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3365 canonical form (as documented). However, after a host name lookup from
3366 the IP address, check_host() was doing a simple string comparison with
3367 addresses acquired from the DNS when checking that the found name did
3368 have the original IP as one of its addresses. Since any found IPv6
3369 addresses are likely to be in abbreviated form, the comparison could
3370 fail. Luckily, there already exists a function for doing the comparison
3371 by converting both addresses to binary, so now that is used instead of
3372 the text comparison.
3374 PH/20 There was another similar case to PH/19, when a complete host name was
3375 given in a host list; looking up its IP address could give an abbreviated
3376 form, whereas the current host's name might or might not be abbreviated.
3377 The same fix has been applied.
3383 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3384 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3387 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3388 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3390 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3392 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3393 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3394 Also, for "Auto-Submitted", treat anything other than "no" as
3395 non-personal, in accordance with RFC 3834. (Previously it treated
3396 anything starting "auto-" as non-personal.)
3398 TF/01 The control=submission/name=... option had a problem with syntax
3399 errors if the name included a slash character. The /name= option
3400 now slurps the rest of the string, so it can include any characters
3401 but it must come last in the list of options (after /sender_retain
3404 PH/05 Some modifications to the interface to the fake nameserver for the new
3412 TK/01 Added the "success_on_redirect" address verification option. See
3413 NewStuff for rationale and an example.
3415 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3417 PH/02 Patch to exigrep to allow it to work on syslog lines.
3419 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3420 fread() to skip over the body file's header line, because in Cygwin the
3421 header line is locked and is inaccessible.
3423 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3424 co-exist for some time) to make it clear that it is the Exim ID that is
3425 referenced, not the Message-ID: header line.
3427 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3428 string_format() function, because snprintf() does not exist on all
3431 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3432 ${stat: expansion item.
3434 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3435 protocol synchronization error", to keep the pedants happy.
3437 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3438 well as for IRIX systems, when gcc is being used. See the host.c source
3441 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3443 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3446 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3447 transport and to the smtp transport in LMTP mode.
3449 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3451 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3452 run a filter in a subprocess. This could lead to confusion in subsequent
3453 lookups in the parent process. There should also be a search_tidyup() at
3454 the end of the subprocess.
3456 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3457 only if the host matched helo_try_verify_hosts, which caused the
3458 verification to occur when the EHLO/HELO command was issued. The ACL just
3459 tested the remembered result. Now, if a previous verification attempt has
3460 not happened, "verify = helo" does it there and then.
3462 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3464 TK/03 Fix log output including CR from clamd.
3466 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3467 exist provoked a memory error which could cause a segfault.
3469 PH/15 Installed PCRE 6.2
3471 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3473 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3474 of the problem. Specifically, suggested +O2 rather than +O1 for the
3477 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3479 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3480 record, the port was not correctly logged when the outgoing_port log
3481 selector was set (it logged the transort's default port).
3483 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3484 fallback_hosts, and "hosts" in the smtp transport.
3486 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3487 host errors such as "Connection refused".
3489 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3490 authentication with radiusclient 0.4.9:
3492 - Error returned from rc_read_config was caught wrongly
3493 - Username/password not passed on to radius server due to wrong length.
3495 The presumption is that some radiusclient API changes for 4.51/PH/17
3496 were not taken care of correctly. The code is still untested by me (my
3497 Linux distribution still has 0.3.2 of radiusclient), but it was
3498 contributed by a Radius user.
3500 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3501 expanding the "port" option of the smtp transport.
3503 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3504 while reading a MIME header. Thanks to Tom Hughes for a patch.
3506 PH/24 Include config.h inside local_scan.h so that configuration settings are
3509 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3510 that in an ACL for RCPT (for example), you can examine exactly what was
3513 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3514 commands, but it was not correctly comparing the address with the actual
3515 client host address. Thus, it would show the EHLO address in Received:
3516 header lines when this was not necessary.
3518 PH/27 Added the % operator to ${eval:}.
3520 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3521 it should be ignoring failures (because with -C, for example, it has lost
3522 privilege). It wasn't ignoring creation failures other than "already
3525 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3528 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3529 because an input line was too long, either on its own, or by virtue of
3530 too many continuations, the temporary file was not being removed, and the
3531 return code was incorrect.
3533 PH/31 Missing "BOOL" in function definition in filtertest.c.
3535 PH/32 Applied Sieve patches from the maintainer.
3537 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3539 PH/33 Added "verify = not_blind".
3541 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3542 Local/Makefile (with some defaults set). These are used in built scripts
3543 such as exicyclog, but they have never been used in the exim_install
3544 script (though there are many overriding facilities there). I have
3545 arranged that the exim_install script now takes note of these two
3548 PH/35 Installed configuration files for Dragonfly.
3550 PH/36 When a locally submitted message by a trusted user did not contain a
3551 From: header, and the sender address was obtained from -f or from an SMTP
3552 MAIL command, and the trusted user did not use -F to supply a sender
3553 name, $originator_name was incorrectly used when constructing a From:
3554 header. Furthermore, $originator_name was used for submission mode
3555 messages from external hosts without From: headers in a similar way,
3556 which is clearly wrong.
3558 PH/37 Added control=suppress_local_fixups.
3560 PH/38 When log_selector = +received_sender was set, and the addition of the
3561 sender made the log line's construction buffer exactly full, or one byte
3562 less than full, an overflow happened when the terminating "\n" was
3565 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3566 when the result of a list match is failure because a DNS lookup failed.
3568 PH/40 RM_COMMAND is now used in the building process.
3570 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3571 the "build-* directories that it finds.
3573 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3574 address in a domain literal was a prefix of an interface address.
3576 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3577 when verifying a sender address, unless rewrite_headers is false.
3579 PH/44 Wrote a long comment about why errors_to addresses are verified as
3580 recipients, not senders.
3582 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3583 the ratelimit ACL was added.
3585 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3587 PH/46 Added extra information about PostgreSQL errors to the error string.
3589 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3590 avoiding the need to install special zones in a real server. This is
3591 backwards compatible; if it can't find the fake resolver, it drops back.
3592 Thus, both old and new test suites can be run.
3594 TF/02 Added util/ratelimit.pl
3596 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3597 clock is set back in time.
3599 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3600 Candler <B.Candler@pobox.com>.
3602 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3603 OK for addresses that are the result of redirections.
3605 PH/48 A number of further additions for the benefit of the new test suite,
3606 including a fake gethostbyname() that interfaces to the fake DNS resolver
3609 TF/06 The fix for widen_domains has also been applied to qualify_single and
3610 search_parents which are the other dnslookup options that can cause
3613 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3616 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3618 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3619 service" if it encountered an MX record with an empty target hostname.
3620 The message is now "an MX or SRV record indicated no SMTP service".
3622 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3623 if the DNS of the sending site is misconfigured. This is quite a
3624 common situation. This change restores the behaviour of treating a
3625 helo verification defer as a failure.
3627 PH/51 If self=fail was set on a router, the bounce message did not include the
3628 actual error message.
3634 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3636 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3637 command itself ended in error, the underlying message about the transport
3638 filter timeout was being overwritten with the pipe command error. Now the
3639 underlying error message should be appended to the second error message.
3641 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3643 PH/02 Reduce the amount of output that "make" produces by default. Full output
3644 can still be requested.
3646 PH/03 The warning log line about a condition test deferring for a "warn" verb
3647 was being output only once per connection, rather than after each
3648 occurrence (because it was using the same function as for successful
3649 "warn" verbs). This seems wrong, so I have changed it.
3651 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3652 it should not have, which might have caused a crash in the right
3653 circumstances, but probably never did.
3655 PH/04 Installed a modified version of Tony Finch's patch to make submission
3656 mode fix the return path as well as the Sender: header line, and to
3657 add a /name= option so that you can make the user's friendly name appear
3660 TF/03 Added the control = fakedefer ACL modifier.
3662 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3663 Mark Lowes for thorough testing.
3665 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3667 TK/03 Merged latest SRS patch from Miles Wilton.
3669 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3670 with the definition in sysexits.h (which is #included earlier).
3671 Fortunately, Exim does not actually use EX_OK. The code used to try to
3672 preserve the sysexits.h value, by assuming that macro definitions were
3673 scanned for macro replacements. I have been disabused of this notion,
3674 so now the code just undefines EX_OK before #including unistd.h.
3676 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3677 in the smtp transport. When a block could not be written in a single
3678 write() function, the timeout was being re-applied to each part-write.
3679 This seems wrong - if the receiver was accepting one byte at a time it
3680 would take for ever. The timeout is now adjusted when this happens. It
3681 doesn't have to be particularly precise.
3683 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3684 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3686 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3687 without a domain if the check to <postmaster@domain> fails.
3689 SC/01 Eximstats: added -xls and the ability to specify output files
3690 (patch written by Frank Heydlauf).
3692 SC/02 Eximstats: use FileHandles for outputting results.
3694 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3696 SC/04 Eximstats: fixed display of large numbers with -nvr option
3698 SC/05 Eximstats: fixed merging of reports with empty tables.
3700 SC/06 Eximstats: added the -include_original_destination flag
3702 SC/07 Eximstats: removed tabs and trailing whitespace.
3704 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3706 TK/06 MBOX spool code: Add real "From " MBOX separator line
3707 so the .eml file is really in mbox format (even though
3708 most programs do not really care). Patch from Alex Miller.
3710 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3711 The latter is generated from $received_to and is only set if the
3712 message has one envelope recipient. SA can use these headers,
3713 obviously out-of-the-box. Patch from Alex Miller.
3715 PH/08 The ${def test on a variable was returning false if the variable's
3716 value was "0", contrary to what the specification has always said!
3717 The result should be true unless the variable is empty.
3719 PH/09 The syntax error of a character other than { following "${if
3720 def:variable_name" (after optional whitespace) was not being diagnosed.
3721 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3722 accidental colon was present, for example, could give incorrect results.
3724 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3725 result is used (not including appendfile, where other changes are about
3728 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3729 This involved changing a lot of size variables from int to off_t. It
3730 should work with maildirs and everything.
3732 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3733 spamd dying while we are connected to it.
3735 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3738 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3739 function for BDB 4.3.
3741 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3743 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3744 the sources for PCRE have been re-arranged and more files are now
3747 PH/15 The code I had for printing potentially long long variables in PH/11
3748 above was not the best (it lost precision). The length of off_t variables
3749 is now inspected at build time, and an appropriate printing format (%ld
3750 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3751 to be "long long int" or "long int". This is needed for the internal
3752 formatting function string_vformat().
3754 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3755 the configuration file to be ":syslog", then the script "guesses" where
3756 the logs files are, rather than using the compiled in default. In our
3757 case the guess is not the same as the compiled default, so the script
3758 suddenly stopped working when I started to use syslog. The patch checks
3759 to see if log_file_path is "". If so, it attempts to read it from exim
3760 with no configuration file to get the compiled in version, before it
3761 falls back to the previous guessing code."
3763 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3764 implementing BATV in an Exim configuration. See NewStuff for the gory
3767 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3768 Makefile that are specific to HP-UX.
3770 PH/18 If the "use_postmaster" option was set for a recipient callout together
3771 with the "random" option, the postmaster address was used as the MAIL
3772 FROM address for the random test, but not for the subsequent recipient
3773 test. It is now used for both.
3775 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3776 patch removes a few documentation additions to RFC 3028, because the
3777 latest draft now contains them. It adds the new en;ascii-case comparator
3778 and a new error check for 8bit text in MIME parts. Comparator and
3779 require names are now matched exactly. I enabled the subaddress
3780 extension, but it is not well tested yet (read: it works for me)."
3782 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3783 rework some of the code of TK/09 above to avoid the hardwired use of
3784 "%lld" and "long long". Replaced the call to snprintf() with a call to
3787 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3788 records point to non-existent hosts", "retry timeout exceeded", and
3789 "retry time not reached for any host after a long failure period".
3791 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3792 experimental DomainKeys support:
3794 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3795 (2) On an error such as an illegally used "control", the wrong name for
3796 the control was given.
3798 These problems did NOT occur unless DomainKeys support was compiled.
3800 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3802 PH/24 Added ${if match_ip condition.
3804 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3805 databases so that it will be absolutely obvious if a crash occurs in the
3806 DB library. This is a regular occurrence (often caused by mis-matched
3809 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3810 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3811 picked up on a user's system that detects such things. There doesn't seem
3812 to be a gcc warning option for this - only an attribute that has to be
3813 put on the function's prototype. It seems that in Fedora Core 4 they have
3814 set this on a number of new functions. No doubt there will be more in due
3817 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3818 specify a transport. However, if an address that was verified by such a
3819 router was the subject of a callout, Exim crashed because it tried to
3820 read the rcpt_include_affixes from the non-existent transport. Now it
3821 just assumes that the setting of that option is false. This bug was
3822 introduced by 4.51/PH/31.
3824 PH/28 Changed -d+all to exclude +memory, because that information is very
3825 rarely of interest, but it makes the output a lot bigger. People tend to
3826 do -d+all out of habit.
3828 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3829 code in os-type was giving problems when libc.so lives in lib64, like on
3832 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3833 aren't the modern standard, and it seems that some systems' include files
3834 don't always have them. Exim was already checking for some of the newer
3835 ones like T_AAAA, and defining it itself. I've added checks for all the
3836 record types that Exim uses.
3838 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3839 not automatically generating a new one, as it is supposed to. This
3840 prevented TLS from working. If the file did exist, but contained invalid
3841 data, a new version was generated, as expected. It was only the case of a
3842 non-existent file that was broken.
3844 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3845 with a change in libdomainkeys > 0.64.
3847 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3848 from DNS. If the selector record carries the flag, it now has
3849 precedence over the domain-wide flag.
3851 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3853 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3854 the use of a port name that isn't defined in /etc/services) occurred, the
3855 message was deferred as in a normal delivery, and thus remained on the
3856 spool, instead of being failed because of the mua_wrapper setting. This
3857 is now fixed, and I tidied up some of the mua_wrapper messages at the
3860 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3861 the messages in a hash of arrays rather than using individual hashes.
3862 This is a bit cleaner and results in dramatic memory savings, albeit
3863 at a slight CPU cost.
3865 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3866 as requested by Marc Sherman.
3868 SC/10 Eximstats: added histograms for user specified patterns as requested
3871 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3873 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3874 fopen() in the content-scanning modules that did not already have it.
3880 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3881 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3883 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3885 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3887 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3888 to test Sieve filters that use "vacation".
3890 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3891 that changes the way the GnuTLS parameters are stored in the cache file.
3892 The new format can be generated externally. For backward compatibility,
3893 if the data in the cache doesn't make sense, Exim assumes it has read an
3894 old-format file, and it generates new data and writes a new file. This
3895 means that you can't go back to an older release without removing the
3898 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3899 work if there are any delivery delays because "one_time" forces the
3900 parent to be marked "delivered", so its unseen clone is never tried
3901 again. For this reason, Exim now forbids the simultaneous setting of
3904 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3905 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3906 entry if you want to know the details. The fix, however, neglected to
3907 consider the case when local delivery batching is involved. The test for
3908 "previously delivered" was not happening when checking to see if an
3909 address could be batched with a previous (undelivered) one; under
3910 certain circumstances this could lead to multiple deliveries to the same
3913 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3914 in its include files, and this causes problems building Exim.
3916 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3917 header_syntax) but Exim was just ignoring anything given after a slash.
3918 In particular, this caused confusion with an attempt to use "verify =
3919 reverse_host_lookup/defer_ok". An error is now given when options are
3920 supplied for verify items that do not have them. (Maybe reverse_host_
3921 lookup should have a defer_ok option, but that's a different point.)
3923 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3924 defined by RFC 821) to 2048, because there were problems with some AUTH
3925 commands, and RFC 1869 says the size should be increased for extended
3926 SMTP commands that take arguments.
3928 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3931 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3932 "unknown" error; now it says that the functionality isn't in the binary.
3934 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3935 an address' error message when a string expansion fails (syntax or
3936 whatever). Otherwise the password may appear in the log. Following change
3937 PH/42 below, there is no longer a chance of it appearing in a bounce
3940 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3942 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3943 of its addresses was ever tried. (Bugzilla bug #2.)
3945 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3946 the result incorrectly in the debug output. (It correctly added a newline
3947 to what was transported.)
3949 TF/01 Added $received_time.
3951 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3952 commented out examples of how to interface to a virus scanner and to
3953 SpamAssassin. Also added commented examples of av_scanner and
3954 spamd_address settings.
3956 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3957 and controls are allowed in which ACLs. There were a couple of minor
3958 errors. Some of the entries in the conditions table (which is a table of
3959 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3960 negation of where the condition IS allowed.
3962 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3964 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3965 header file does not have a version number, so I've had to invent a new
3966 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3967 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3968 radiusclient), but it was contributed by a Radius user.
3970 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3971 files or directories, for OpenSSL.
3973 PH/19 When an Exim process that is running as root has to create an Exim log
3974 file, it does so in a subprocess that runs as exim:exim so as to get the
3975 ownership right at creation (otherwise, other Exim processes might see
3976 the file with the wrong ownership). There was no test for failure of this
3977 fork() call, which would lead to the process getting stuck as it waited
3978 for a non-existent subprocess. Forks do occasionally fail when resources
3979 run out. I reviewed all the other calls to fork(); they all seem to check
3982 PH/20 When checking for unexpected SMTP input at connect time (before writing
3983 the banner), Exim was not dealing correctly with a non-positive return
3984 from the read() function. If the client had disconnected by this time,
3985 the result was a log entry for a synchronization error with an empty
3986 string after "input=" when read() returned zero. If read() returned -1
3987 (an event I could not check), uninitialized data bytes were printed.
3988 There were reports of junk text (parts of files, etc) appearing after
3991 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3993 PH/22 Added support for macro redefinition, and (re)definition in between
3994 driver and ACL definitions.
3996 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3997 forgetting to use the resulting value; it was using the unexpanded value.
3999 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4000 hadn't been configured. The fix is from Juergen Kreileder, who
4001 understands it better than I do:
4003 "Here's what I see happening with three configured cyrus_sasl
4004 authenticators configured (plain, login, cram-md5):
4006 On startup auth_cyrus_sasl_init() gets called for each of these.
4007 This means three calls to sasl_listmech() without a specified mech_list.
4008 => SASL tests which mechs of all available mechs actually work
4009 => three warnings about OTP not working
4010 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4012 With the patch, sasl_listmech() also gets called three times. But now
4013 SASL's mech_list option is set to the server_mech specified in the the
4014 authenticator. Or in other words, the answer from sasl_listmech()
4015 gets limited to just the mech you're testing for (which is different
4017 => the return list contains just 'plain' or 'login', 'cram-md5' or
4018 nothing depending on the value of ob->server_mech.
4020 I've just tested the patch: Authentication still works fine,
4021 unavailable mechs specified in the exim configuration are still
4022 caught, and the auth.log warnings about OTP are gone."
4024 PH/25 When debugging is enabled, the contents of the command line are added
4025 to the debugging output, even when log_selector=+arguments is not
4028 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4029 answer is "GNU", and only if the return is "GNU/something" is the answer
4032 PH/27 $acl_verify_message is now set immediately after the failure of a
4033 verification in an ACL, and so is available in subsequent modifiers. In
4034 particular, the message can be preserved by coding like this:
4036 warn !verify = sender
4037 set acl_m0 = $acl_verify_message
4039 Previously, $acl_verify_message was set only while expanding "message"
4040 and "log_message" when a very denied access.
4042 PH/28 Modified OS/os.c-Linux with
4044 -#ifndef OS_LOAD_AVERAGE
4045 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4047 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4048 nomenclature these days.)
4050 PH/29 Installed patch from the Sieve maintainer that adds the options
4051 sieve_useraddress and sieve_subaddress to the redirect router.
4053 PH/30 In these circumstances:
4054 . Two addresses routed to the same list of hosts;
4055 . First host does not offer TLS;
4056 . First host accepts first address;
4057 . First host gives temporary error to second address;
4058 . Second host offers TLS and a TLS session is established;
4059 . Second host accepts second address.
4060 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4061 and peerdn, if requested) that were in fact used only for the second
4064 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4065 attention to any local part prefix or suffix that was matched by the
4066 router that accepted the address. It now behaves in the same way as it
4067 does for delivery: the affixes are removed from the local part unless
4068 rcpt_include_affixes is set on the transport.
4070 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4071 timeout during the DATA phase of an incoming message.
4073 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4074 applied a patch sanctioned by the Sieve maintainer.
4076 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4077 the uid or gid is negative. A case of a negative gid caused this to be
4078 noticed. The fix allows for either to be negative.
4080 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4081 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4084 PH/36 The change PH/12 above was broken. Fixed it.
4086 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4087 the grounds that routing the same address twice would always produce the
4088 same answer. This might have been true once, but it is certainly no
4089 longer true now. Routing a child address may depend on the previous
4090 routing that produced that child. Some complicated redirection strategies
4091 went wrong when messages had multiple recipients, and made Exim's
4092 behaviour dependent on the order in which the addresses were given.
4094 I have moved the duplicate checking until after the routing is complete.
4095 Exim scans the addresses that are assigned to local and remote
4096 transports, and removes any duplicates. This means that more work will be
4097 done, as duplicates will always all be routed, but duplicates are
4098 presumably rare, so I don't expect this is of any significance.
4100 For deliveries to pipes, files, and autoreplies, the duplicate checking
4101 still happens during the routing process, since they are not going to be
4104 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4105 It corrects a timeout issue with spamd. This is Ian's comment: "The
4106 background is that sometimes spamd either never reads data from a
4107 connection it has accepted, or it never writes response data. The exiscan
4108 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4109 blindly assumes that writes won't block so it may never time out."
4111 PH/39 Allow G after quota size as well as K and M.
4113 PH/40 The value set for $authenticated_id in an authenticator may not contain
4114 binary zeroes or newlines because the value is written to log lines and
4115 to spool files. There was no check on this. Now the value is run through
4116 the string_printing() function so that such characters are converted to
4117 printable escape sequences.
4119 PH/41 $message_linecount is a new variable that contains the total number of
4120 lines in the message. Compare $body_linecount, which is the count for the
4123 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4124 bounce and delay warning messages, except in certain special cases, which
4127 (a) An SMTP error message from a remote host;
4128 (b) A message specified in a :fail: redirection;
4129 (c) A message specified in a "fail" command in a system filter;
4130 (d) A message specified in a FAIL return from the queryprogram router;
4131 (e) A message specified by the cannot_route_message router option.
4133 In these cases only, Exim does include the error details in bounce and
4134 warning messages. There are also a few cases where bland messages such
4135 as "unrouteable address" or "local delivery error" are given.
4137 PH/43 $value is now also set for the "else" part of a ${run expansion.
4139 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4140 being worked on, but at least Exim now implements the latest version to
4143 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4144 process to complete was treated as a delivery failure, a timeout while
4145 writing the message to the pipe was logged, but erroneously treated as a
4146 successful delivery. Such timeouts include transport filter timeouts. For
4147 consistency with the overall process timeout, these timeouts are now
4148 treated as errors, giving rise to delivery failures by default. However,
4149 there is now a new Boolean option for the pipe transport called
4150 timeout_defer, which, if set TRUE, converts the failures into defers for
4151 both kinds of timeout. A transport filter timeout is now identified in
4154 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4155 systems where "make" and "gmake" are different, calling "gmake" at top
4156 level broke things. I've arranged for the value of $(MAKE) to be passed
4157 from the Makefile to this script so that it can call the same version of
4161 A note about Exim versions 4.44 and 4.50
4162 ----------------------------------------
4164 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4165 changes of various kinds. As a consequence, a big documentation update was
4166 needed. This delayed the release for rather longer than seemed good, especially
4167 in the light of a couple of (minor) security issues. Therefore, the changes
4168 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4169 release. So 4.44 and 4.50 are in effect two different branches that both start
4172 I have left the 4.50 change log unchanged; it contains all the changes since
4173 4.43. The change log for 4.44 is below; many of its items are identical to
4174 those for 4.50. This seems to be the most sensible way to preserve the
4175 historical information.
4181 1. Minor wording change to the doc/README.SIEVE file.
4183 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4184 computation of the current number of files was incorrect.
4186 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4187 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4190 4. Give more explanation in the error message when the command for a transport
4191 filter fails to execute.
4193 5. There are several places where Exim runs a non-Exim command in a
4194 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4195 was being done only for the command run by the queryprogram router. It is
4196 now done for all such subprocesses. The other cases are: ${run, transport
4197 filters, and the commands run by the lmtp and pipe transports.
4199 6. Added CONFIGURE_GROUP build-time option.
4201 7. Some older OS have a limit of 256 on the maximum number of file
4202 descriptors. Exim was using setrlimit() to set 1000 as a large value
4203 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4204 systems. I've change it so that if it can't get 1000, it tries for 256.
4206 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4207 was an oversight, and furthermore, ever since the addition of extra
4208 controls (e.g. 4.43/32), the checks on when to allow different forms of
4209 "control" were broken. There should now be diagnostics for all cases when a
4210 control that does not make sense is encountered.
4212 9. Added the /retain_sender option to "control=submission".
4214 10. $recipients is now available in the predata ACL (oversight).
4216 11. Tidy the search cache before the fork to do a delivery from a message
4217 received from the command line. Otherwise the child will trigger a lookup
4218 failure and thereby defer the delivery if it tries to use (for example) a
4219 cached ldap connection that the parent has called unbind on.
4221 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4222 of $address_data from the recipient verification was clobbered by the
4223 sender verification.
4225 13. The value of address_data from a sender verification is now available in
4226 $sender_address_data in subsequent conditions in the ACL statement.
4228 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4230 15. Added a new option "connect=<time>" to callout options, to set a different
4233 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4234 was its contents. (It was OK if the option was not defined at all.)
4236 17. A "Completed" log line is now written for messages that are removed from
4237 the spool by the -Mrm option.
4239 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4240 information about exactly what failed.
4242 19. Added -dd to debug only the daemon process.
4244 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4245 handles timeouts, both on the server side and network timeouts. Renamed the
4246 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4248 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4249 It is now set to "smtps".
4251 22. $host_address is now set to the target address during the checking of
4252 ignore_target_hosts.
4254 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4255 being passed; this would have caused $sender_host_name to have been used if
4256 matching the list had actually called for a host name (not very likely,
4257 since this list is usually IP addresses). A host name is now passed as
4260 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4261 code that specifies a non-restarting handler (typically sigaction() in
4262 modern systems) in an attempt to fix a rare and obscure crash bug.
4264 25. Narrowed the window for a race in the daemon that could cause it to ignore
4265 SIGCHLD signals. This is not a major problem, because they are used only to
4266 wake it up if nothing else does.
4268 26. A malformed maildirsize file could cause Exim to calculate negative values
4269 for the mailbox size or file count. Odd effects could occur as a result.
4270 The maildirsize information is now recalculated if the size or filecount
4273 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4274 support for a long time. Removed HAVE_SYS_VFS_H.
4276 28. Installed the latest version of exipick from John Jetmore.
4278 29. In an address list, if the pattern was not a regular expression, an empty
4279 subject address (from a bounce message) matched only if the pattern was an
4280 empty string. Non-empty patterns were not even tested. This was the wrong
4281 because it is perfectly reasonable to use an empty address as part of a
4282 database query. An empty address is now tested by patterns that are
4283 lookups. However, all the other forms of pattern expect the subject to
4284 contain a local part and a domain, and therefore, for them, an empty
4285 address still always fails if the pattern is not itself empty.
4287 30. Exim went into a mad DNS loop when attempting to do a callout where the
4288 host was specified on an smtp transport, and looking it up yielded more
4289 than one IP address.
4291 31. Re-factored the code for checking spool and log partition space into a
4292 function that finds that data and another that does the check. The former
4293 is then used to implement four new variables: $spool_space, $log_space,
4294 $spool_inodes, and $log_inodes.
4296 32. The RFC2047 encoding function was originally intended for short strings
4297 such as real names; it was not keeping to the 75-character limit for
4298 encoded words that the RFC imposes. It now respects the limit, and
4299 generates multiple encoded words if necessary. To be on the safe side, I
4300 have increased the buffer size for the ${rfc2047: expansion operator from
4303 33. It is now permitted to omit both strings after an "if" condition; if the
4304 condition is true, the result is "true". As before, when the second string
4305 is omitted, a false condition yields an empty string. This makes it less
4306 cumbersome to write custom ACL and router conditions.
4308 34. Failure to deliver a bounce message always caused it to be frozen, even if
4309 there was an errors_to setting on the router. The errors_to setting is now
4312 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4313 canonical form (fully expanded) before being placed in
4314 $sender_host_address.
4316 36. The table in the code that translates DNS record types into text (T_A to
4317 "A" for instance) was missing entries for NS and CNAME. It is just possible
4318 that this could have caused confusion if both these types were looked up
4319 for the same domain, because the text type is used as part of Exim's
4320 per-process caching. But the chance of anyone hitting this buglet seems
4323 37. The dnsdb lookup has been extended in a number of ways.
4325 (1) There is a new type, "zns", which walks up the domain tree until it
4326 finds some nameserver records. It should be used with care.
4328 (2) There is a new type, "mxh", which is like "mx" except that it returns
4329 just the host names, not the priorities.
4331 (3) It is now possible to give a list of domains (or IP addresses) to be
4332 looked up. The behaviour when one of the lookups defers can be
4333 controlled by a keyword.
4335 (4) It is now possible to specify the separator character for use when
4336 multiple records are returned.
4338 38. The dnslists ACL condition has been extended: it is now possible to supply
4339 a list of IP addresses and/or domains to be looked up in a particular DNS
4342 39. Added log_selector=+queue_time_overall.
4344 40. When running the queue in the test harness, wait just a tad after forking a
4345 delivery process, to get repeatability of debugging output.
4347 41. Include certificate and key file names in error message when GnuTLS fails
4348 to set them up, because the GnuTLS error message doesn't include the name
4349 of the failing file when there is a problem reading it.
4351 42. Allow both -bf and -bF in the same test run.
4353 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4355 44. The "Exiscan patch" is now merged into the mainline Exim source.
4357 45. Sometimes the final signoff response after QUIT could fail to get
4358 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4359 < 0 before doing a fflush(). This bug looks as though it goes back to the
4360 introduction of TLS in release 3.20, but "sometimes" must have been rare
4361 because the tests only now provoked it.
4363 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4364 (this can affect the format of dates).
4366 47. exim_tidydb, when checking for the continued existence of a message for
4367 which it has found a message-specific retry record, was not finding
4368 messages that were in split spool directories. Consequently, it was
4369 deleting retry records that should have stayed in existence.
4371 48. Steve fixed some bugs in eximstats.
4373 49. The SPA authentication driver was not abandoning authentication and moving
4374 on to the next authenticator when an expansion was forced to fail,
4375 contradicting the general specification for all authenticators. Instead it
4376 was generating a temporary error. It now behaves as specified.
4378 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4379 (the order specifies the preference for clients). The order is now AES256,
4380 AES128, 3DES, ARCFOUR128.
4382 51. Small patch to Sieve code - explicitly set From: when generating an
4385 52. Exim crashed if a remote delivery caused a very long error message to be
4386 recorded - for instance if somebody sent an entire SpamAssassin report back
4387 as a large number of 550 error lines. This bug was coincidentally fixed by
4388 increasing the size of one of Exim's internal buffers (big_buffer) that
4389 happened as part of the Exiscan merge. However, to be on the safe side, I
4390 have made the code more robust (and fixed the comments that describe what
4393 53. Now that there can be additional text after "Completed" in log lines (if
4394 the queue_time_overall log selector is set), a one-byte patch to exigrep
4395 was needed to allow it to recognize "Completed" as not the last thing in
4398 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4399 patch that reportedly fixes this has been added. I am not expert enough to
4400 create a test for it. This is what the patch creator wrote:
4402 "I found a little strange behaviour of ldap code when working with
4403 Windows 2003 AD Domain, where users was placed in more than one
4404 Organization Units. When I tried to give exim partial DN, the exit code
4405 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4406 But simultaneously result of request was absolutely normal ldap result,
4407 so I produce this patch..."
4409 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4410 REFERENCE, so I have modified the code to exclude the patch when that macro
4413 55. Some experimental protocols are using DNS PTR records for new purposes. The
4414 keys for these records are domain names, not reversed IP addresses. The
4415 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4416 leaves it alone. Component reversal etc. now happens only for IP addresses.
4419 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4421 57. Double the size of the debug message buffer (to 2048) so that more of very
4422 long debug lines gets shown.
4424 58. The exicyclog utility now does better if the number of log files to keep
4425 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4427 59. Two changes related to the smtp_active_hostname option:
4429 (1) $smtp_active_hostname is now available as a variable.
4430 (2) The default for smtp_banner uses $smtp_active_hostname instead
4431 of $primary_hostname.
4433 60. The host_aton() function is supposed to be passed a string that is known
4434 to be a valid IP address. However, in the case of IPv6 addresses, it was
4435 not checking this. This is a hostage to fortune. Exim now panics and dies
4436 if the condition is not met. A case was found where this could be provoked
4437 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4438 components; fortuitously, this particular loophole had already been fixed
4439 by change 4.50/55 above.
4441 If there are any other similar loopholes, the new check in host_aton()
4442 itself should stop them being exploited. The report I received stated that
4443 data on the command line could provoke the exploit when Exim was running as
4444 exim, but did not say which command line option was involved. All I could
4445 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4446 running as the user.
4449 61. There was a buffer overflow vulnerability in the SPA authentication code
4450 (which came originally from the Samba project). I have added a test to the
4451 spa_base64_to_bits() function which I hope fixes it.
4454 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4455 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4457 63. The daemon start-up calls getloadavg() while still root for those OS that
4458 need the first call to be done as root, but it missed one case: when
4459 deliver_queue_load_max is set with deliver_drop_privilege. This is
4460 necessary for the benefit of the queue runner, because there is no re-exec
4461 when deliver_drop_privilege is set.
4463 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4464 This has been fixed.
4466 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4467 was in use, was not putting the data itself into the right store pool;
4468 consequently, it could be overwritten for a subsequent message in the same
4469 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4472 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4474 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4475 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4476 the function were treating the return as a boolean value, which happened to
4477 work because 0=false and not-0=true, but is not correct code.
4479 68. The host_aton() function was not handling scoped IPv6 addresses (those
4480 with, for example, "%eth0" on the end) correctly.
4482 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4483 negated items (that is, ~something) in unsigned ints. Some compilers
4484 apparently mutter when there is no cast.
4486 70. If an address verification called from an ACL failed, and did not produce a
4487 user-specific message (i.e. there was only a "system" message), nothing was
4488 put in $acl_verify_message. In this situation, it now puts the system
4491 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4492 session; change 4.31/43 added the unwanted input to the log line - except
4493 that it did not do this in the start of session case. It now does.
4495 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4496 This is wrong and can cause the other end to generate a synchronization
4497 error if it is another Exim or anything else that does the synchronization
4498 check. A QUIT command is no longer sent after a timeout.
4500 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4501 during host lookups.
4503 74. The defer_ok option of callout verification was not working if it was used
4504 when verifying addresses in header lines, that is, for this case:
4506 verify = header_sender/callout=defer_ok
4508 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4509 those file descriptors could be used for SMTP connections. If anything
4510 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4511 could be sent to the SMTP client, causing chaos. The daemon now opens
4512 stdin, stdout, and stderr to /dev/null when it puts itself into the
4515 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4516 log by default. The user can override this with suitable Perl magic.
4518 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4519 the log message when discard triggers, was not working for the DATA ACL or
4520 for the non-SMTP ACL.
4522 78. Error message wording change in sieve.c.
4524 79. If smtp_accept_max_per_host was set, the number of connections could be
4525 restricted to fewer than expected, because the daemon was trying to set up
4526 a new connection before checking whether the processes handling previous
4527 connections had finished. The check for completed processes is now done
4528 earlier. On busy systems, this bug wouldn't be noticed because something
4529 else would have woken the daemon, and it would have reaped the completed
4532 80. If a message was submitted locally by a user whose login name contained one
4533 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4534 It caused a spool format error. I have fixed the spool reading code. A
4535 related problem was that the "from" clause in the Received: line became
4536 illegal because of the space(s). It is now covered by ${quote_local_part.
4538 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4541 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4543 83. Patch from Sieve maintainer to fix unterminated string problem in
4544 "vacation" handling.
4546 84. Some minor changes to the Linux configuration files to help with other
4547 OS variants using glibc.
4549 85. One more patch for Sieve to update vacation handling to latest spec.
4552 ----------------------------------------------------
4553 See the note above about the 4.44 and 4.50 releases.
4554 ----------------------------------------------------
4560 1. Change 4.43/35 introduced a bug that caused file counts to be
4561 incorrectly computed when quota_filecount was set in an appendfile
4564 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4565 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4568 3. Give more explanation in the error message when the command for a transport
4569 filter fails to execute.
4571 4. There are several places where Exim runs a non-Exim command in a
4572 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4573 was being done only for the command run by the queryprogram router. It is
4574 now done for all such subprocesses. The other cases are: ${run, transport
4575 filters, and the commands run by the lmtp and pipe transports.
4577 5. Some older OS have a limit of 256 on the maximum number of file
4578 descriptors. Exim was using setrlimit() to set 1000 as a large value
4579 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4580 systems. I've change it so that if it can't get 1000, it tries for 256.
4582 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4583 was an oversight, and furthermore, ever since the addition of extra
4584 controls (e.g. 4.43/32), the checks on when to allow different forms of
4585 "control" were broken. There should now be diagnostics for all cases when a
4586 control that does not make sense is encountered.
4588 7. $recipients is now available in the predata ACL (oversight).
4590 8. Tidy the search cache before the fork to do a delivery from a message
4591 received from the command line. Otherwise the child will trigger a lookup
4592 failure and thereby defer the delivery if it tries to use (for example) a
4593 cached ldap connection that the parent has called unbind on.
4595 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4596 of $address_data from the recipient verification was clobbered by the
4597 sender verification.
4599 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4600 was its contents. (It was OK if the option was not defined at all.)
4602 11. A "Completed" log line is now written for messages that are removed from
4603 the spool by the -Mrm option.
4605 12. $host_address is now set to the target address during the checking of
4606 ignore_target_hosts.
4608 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4609 being passed; this would have caused $sender_host_name to have been used if
4610 matching the list had actually called for a host name (not very likely,
4611 since this list is usually IP addresses). A host name is now passed as
4614 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4615 code that specifies a non-restarting handler (typically sigaction() in
4616 modern systems) in an attempt to fix a rare and obscure crash bug.
4618 15. Narrowed the window for a race in the daemon that could cause it to ignore
4619 SIGCHLD signals. This is not a major problem, because they are used only to
4620 wake it up if nothing else does.
4622 16. A malformed maildirsize file could cause Exim to calculate negative values
4623 for the mailbox size or file count. Odd effects could occur as a result.
4624 The maildirsize information is now recalculated if the size or filecount
4627 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4628 support for a long time. Removed HAVE_SYS_VFS_H.
4630 18. Updated exipick to current release from John Jetmore.
4632 19. Allow an empty sender to be matched against a lookup in an address list.
4633 Previously the only cases considered were a regular expression, or an
4636 20. Exim went into a mad DNS lookup loop when doing a callout where the
4637 host was specified on the transport, if the DNS lookup yielded more than
4640 21. The RFC2047 encoding function was originally intended for short strings
4641 such as real names; it was not keeping to the 75-character limit for
4642 encoded words that the RFC imposes. It now respects the limit, and
4643 generates multiple encoded words if necessary. To be on the safe side, I
4644 have increased the buffer size for the ${rfc2047: expansion operator from
4647 22. Failure to deliver a bounce message always caused it to be frozen, even if
4648 there was an errors_to setting on the router. The errors_to setting is now
4651 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4652 canonical form (fully expanded) before being placed in
4653 $sender_host_address.
4655 24. Updated eximstats to version 1.33
4657 25. Include certificate and key file names in error message when GnuTLS fails
4658 to set them up, because the GnuTLS error message doesn't include the name
4659 of the failing file when there is a problem reading it.
4661 26. Expand error message when OpenSSL has problems setting up cert/key files.
4664 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4665 (this can affect the format of dates).
4667 28. exim_tidydb, when checking for the continued existence of a message for
4668 which it has found a message-specific retry record, was not finding
4669 messages that were in split spool directories. Consequently, it was
4670 deleting retry records that should have stayed in existence.
4672 29. eximstats updated to version 1.35
4673 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4674 1.35 - bugfix such that pie charts by volume are generated correctly
4676 30. The SPA authentication driver was not abandoning authentication and moving
4677 on to the next authenticator when an expansion was forced to fail,
4678 contradicting the general specification for all authenticators. Instead it
4679 was generating a temporary error. It now behaves as specified.
4681 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4682 (the order specifies the preference for clients). The order is now AES256,
4683 AES128, 3DES, ARCFOUR128.
4685 31. Small patch to Sieve code - explicitly set From: when generating an
4688 32. Exim crashed if a remote delivery caused a very long error message to be
4689 recorded - for instance if somebody sent an entire SpamAssassin report back
4690 as a large number of 550 error lines. This bug was coincidentally fixed by
4691 increasing the size of one of Exim's internal buffers (big_buffer) that
4692 happened as part of the Exiscan merge. However, to be on the safe side, I
4693 have made the code more robust (and fixed the comments that describe what
4696 33. Some experimental protocols are using DNS PTR records for new purposes. The
4697 keys for these records are domain names, not reversed IP addresses. The
4698 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4699 leaves it alone. Component reversal etc. now happens only for IP addresses.
4702 34. The host_aton() function is supposed to be passed a string that is known
4703 to be a valid IP address. However, in the case of IPv6 addresses, it was
4704 not checking this. This is a hostage to fortune. Exim now panics and dies
4705 if the condition is not met. A case was found where this could be provoked
4706 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4707 components; fortuitously, this particular loophole had already been fixed
4708 by change 4.50/55 or 4.44/33 above.
4710 If there are any other similar loopholes, the new check in host_aton()
4711 itself should stop them being exploited. The report I received stated that
4712 data on the command line could provoke the exploit when Exim was running as
4713 exim, but did not say which command line option was involved. All I could
4714 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4715 running as the user.
4718 35. There was a buffer overflow vulnerability in the SPA authentication code
4719 (which came originally from the Samba project). I have added a test to the
4720 spa_base64_to_bits() function which I hope fixes it.
4723 36. The daemon start-up calls getloadavg() while still root for those OS that
4724 need the first call to be done as root, but it missed one case: when
4725 deliver_queue_load_max is set with deliver_drop_privilege. This is
4726 necessary for the benefit of the queue runner, because there is no re-exec
4727 when deliver_drop_privilege is set.
4729 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4730 was in use, was not putting the data itself into the right store pool;
4731 consequently, it could be overwritten for a subsequent message in the same
4732 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4735 38. Sometimes the final signoff response after QUIT could fail to get
4736 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4737 < 0 before doing a fflush(). This bug looks as though it goes back to the
4738 introduction of TLS in release 3.20, but "sometimes" must have been rare
4739 because the tests only now provoked it.
4745 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4746 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4747 Now it returns an integer. A number of calls were still expecting a T/F
4748 return. Fortuitously, in all cases, the tests worked in OK situations,
4749 which is the norm. However, things would have gone wrong on any write
4750 failures on the smtp file descriptor. This function is used when sending
4751 messages over SMTP and also when doing verify callouts.
4753 2. When Exim is called to do synchronous delivery of a locally submitted
4754 message (the -odf or -odi options), it no longer closes stderr before doing
4757 3. Implemented the mua_wrapper option.
4759 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4761 5. Implemented the functions header_remove(), header_testname(),
4762 header_add_at_position(), and receive_remove_recipient(), and exported them
4765 6. If an ACL "warn" statement specified the addition of headers, Exim already
4766 inserted X-ACL-Warn: at the start if there was no header name. However, it
4767 was not making this test for the second and subsequent header lines if
4768 there were newlines in the string. This meant that an invalid header could
4769 be inserted if Exim was badly configured.
4771 7. Allow an ACL "warn" statement to add header lines at the start or after all
4772 the Received: headers, as well as at the end.
4774 8. Added the rcpt_4xx retry error code.
4776 9. Added postmaster_mailfrom=xxx to callout verification option.
4778 10. Added mailfrom=xxxx to the callout verification option, for verify=
4781 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4782 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4784 12. Inserted some casts to stop certain compilers warning when using pointer
4785 differences as field lengths or precisions in printf-type calls (mostly
4786 affecting debugging statements).
4788 13. Added optional readline() support for -be (dynamically loaded).
4790 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4791 same clock tick as a message's arrival, so that its received time was the
4792 same as the "first fail" time on the retry record, and that message
4793 remained on the queue past the ultimate address timeout, every queue runner
4794 would try a delivery (because it was past the ultimate address timeout) but
4795 after another failure, the ultimate address timeout, which should have then
4796 bounced the address, did not kick in. This was a "< instead of <=" error;
4797 in most cases the first failure would have been in the next clock tick
4798 after the received time, and all would be well.
4800 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4801 being recognized when the domain list was tested by the match_domain
4802 condition in an expansion string.
4804 16. Added the ${str2b64: operator.
4806 17. Exim was always calling setrlimit() to set a large limit for the number of
4807 processes, without checking whether the existing limit was already
4808 adequate. (It did check for the limit on file descriptors.) Furthermore,
4809 errors from getrlimit() and setrlimit() were being ignored. Now they are
4810 logged to the main and panic logs, but Exim does carry on, to try to do its
4811 job under whatever limits there are.
4813 18. Imported PCRE 5.0.
4815 19. Trivial typo in log message " temporarily refused connection" (the leading
4818 20. If the log selector return_path_on_delivery was set and an address was
4819 redirected to /dev/null, the delivery process crashed because it assumed
4820 that a return path would always be set for a "successful" delivery. In this
4821 case, the whole delivery is bypassed as an optimization, and therefore no
4824 21. Internal re-arrangement: the function for sending a challenge and reading
4825 a response while authentication was assuming a zero-terminated challenge
4826 string. It's now changed to take a pointer and a length, to allow for
4827 binary data in such strings.
4829 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4831 23. Exim was not respecting finduser_retries when seeking the login of the
4832 uid under which it was called; it was always trying 10 times. (The default
4833 setting of finduser_retries is zero.) Also, it was sleeping after the final
4834 failure, which is pointless.
4836 24. Implemented tls_on_connect_ports.
4838 25. Implemented acl_smtp_predata.
4840 26. If the domain in control=submission is set empty, Exim assumes that the
4841 authenticated id is a complete email address when it generates From: or
4842 Sender: header lines.
4844 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4845 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4846 chown and chgrp in /bin and hostname in /usr/bin.
4848 28. Exim was keeping the "process log" file open after each use, just as it
4849 does for the main log. This opens the possibility of it remaining open for
4850 long periods when the USR1 signal hits a daemon. Occasional processlog
4851 errors were reported, that could have been caused by this. Anyway, it seems
4852 much more sensible not to leave this file open at all, so that is what now
4855 29. The long-running daemon process does not normally write to the log once it
4856 has entered its main loop, and it closes the log before doing so. This is
4857 so that log files can straightforwardly be renamed and moved. However,
4858 there are a couple of unusual error situations where the daemon does write
4859 log entries, and I had neglected to close the log afterwards.
4861 30. The text of an SMTP error response that was received during a remote
4862 delivery was being truncated at 512 bytes. This is too short for some of
4863 the long messages that one sometimes sees. I've increased the limit to
4866 31. It is now possible to make retry rules that apply only when a message has a
4867 specific sender, in particular, an empty sender.
4869 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4870 it possible to be selective about when SMTP synchronization is enforced.
4872 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4874 32. Implemented hosts_connection_nolog.
4876 33. Added an ACL for QUIT.
4878 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4881 35. Added mailbox_size and mailbox_filecount to appendfile.
4883 36. Added control = no_multiline_responses to ACLs.
4885 37. There was a bug in the logic of the code that waits for the clock to tick
4886 in the case where the clock went backwards by a substantial amount such
4887 that the microsecond fraction of "now" was more than the microsecond
4888 fraction of "then" (but the whole seconds number was less).
4890 38. Added support for the libradius Radius client library this is found on
4891 FreeBSD (previously only the radiusclient library was supported).
4897 1. When certain lookups returned multiple values in the form name=value, the
4898 quoting of the values was not always being done properly. Specifically:
4899 (a) If the value started with a double quote, but contained no whitespace,
4901 (b) If the value contained whitespace other than a space character (i.e.
4902 tabs or newlines or carriage returns) it was not quoted.
4903 This fix has been applied to the mysql and pgsql lookups by writing a
4904 separate quoting function and calling it from the lookup code. The fix
4905 should probably also be applied to nisplus, ibase and oracle lookups, but
4906 since I cannot test any of those, I have not disturbed their existing code.
4908 2. A hit in the callout cache for a specific address caused a log line with no
4909 reason for rejecting RCPT. Now it says "Previous (cached) callout
4910 verification failure".
4912 3. There was an off-by-one bug in the queryprogram router. An over-long
4913 return line was truncated at 256 instead of 255 characters, thereby
4914 overflowing its buffer with the terminating zero. As well as fixing this, I
4915 have increased the buffer size to 1024 (and made a note to document this).
4917 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4918 when Exim is waiting for an SMTP response from a remote server, Exim
4919 restarts its select() call on the socket, thereby resetting its timeout.
4920 This is not a problem when such interrupts are rare. Somebody set up a cron
4921 job to run exiwhat every 2 minutes, which is less than the normal select()
4922 timeout (5 or 10 minutes). This meant that the select() timeout never
4923 kicked in because it was always reset. I have fixed this by comparing the
4924 time when an interrupt arrives with the time at the start of the first call
4925 to select(). If more time than the timeout has elapsed, the interrupt is
4926 treated as a timeout.
4928 5. Some internal re-factoring in preparation for the addition of Sieve
4929 extensions (by MH). In particular, the "personal" test is moved to a
4930 separate function, and given an option for scanning Cc: and Bcc: (which is
4931 not set for Exim filters).
4933 6. When Exim created an email address using the login of the caller as the
4934 local part (e.g. when creating a From: or Sender: header line), it was not
4935 quoting the local part when it contained special characters such as @.
4937 7. Installed new OpenBSD configuration files.
4939 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4940 try to make them clearer.
4942 9. Callout options, other than the timeout value, were being ignored when
4943 verifying sender addresses in header lines. For example, when using
4945 verify = header_sender/callout=no_cache
4947 the cache was (incorrectly) being used.
4949 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4950 only the Cygwin environment.
4952 11. When return_path_on_delivery was set as a log selector, if different remote
4953 addresses in the same message used different return paths and parallel
4954 remote delivery occurred, the wrong values would sometimes be logged.
4955 (Whenever a remote delivery process finished, the return path value from
4956 the most recently started remote delivery process was logged.)
4958 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4959 lines when AUTH and/or TLS are in use. This is the "received protocol"
4960 field. Exim used to use "asmtp" for authenticated SMTP, without any
4961 indication (in the protocol name) for TLS use. Now it follows the RFC and
4962 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4963 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4964 names appear in log lines as well as in Received: header lines.
4966 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4967 extensions, and comparison tests, and to fix some bugs.
4969 14. Changes to the "personal" filter test:
4971 (1) The test was buggy in that it was just doing the equivalent of
4972 "contains" tests on header lines. For example, if a user's address was
4973 anne@some.where, the "personal" test would incorrectly be true for
4975 To: susanne@some.where
4977 This test is now done by extracting each address from the header in turn,
4978 and checking the entire address. Other tests that are part of "personal"
4979 are now done using regular expressions (for example, to check local parts
4980 of addresses in From: header lines).
4982 (2) The list of non-personal local parts in From: addresses has been
4983 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4984 taken from the Sieve specification recommendations.
4986 (3) If the message contains any header line starting with "List-" it is
4987 treated as non-personal.
4989 (4) The test for "circular" in the Subject: header line has been removed
4990 because it now seems ill-conceived.
4992 15. Minor typos in src/EDITME comments corrected.
4994 16. Installed latest exipick from John Jetmore.
4996 17. If headers_add on a router specified a text string that was too long for
4997 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4998 of string_sprintf() is now avoided.
5000 18. $message_body_size was not set (it was always zero) when running the DATA
5001 ACL and the local_scan() function.
5003 19. For the "mail" command in an Exim filter, no default was being set for
5004 the once_repeat time, causing a random time value to be used if "once" was
5005 specified. (If the value happened to be <= 0, no repeat happened.) The
5006 default is now 0s, meaning "never repeat". The "vacation" command was OK
5007 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5008 (I found it when inspecting the code).
5010 20. There is now an overall timeout for performing a callout verification. It
5011 defaults to 4 times the callout timeout, which applies to individual SMTP
5012 commands during the callout. The overall timeout applies when there is more
5013 than one host that can be tried. The timeout is checked before trying the
5014 next host. This prevents very long delays if there are a large number of
5015 hosts and all are timing out (e.g. when the network connections are timing
5016 out). The value of the overall timeout can be changed by specifying an
5017 additional sub-option for "callout", called "maxwait". For example:
5019 verify = sender/callout=5s,maxwait=20s
5021 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5022 to the end before writing, but this should make it even safer).
5024 22. Exim was forgetting that it had advertised PIPELINING for the second and
5025 subsequent messages on an SMTP connection. It was also not resetting its
5026 memory on STARTTLS and an internal HELO.
5028 23. When Exim logs an SMTP synchronization error within a session, it now
5029 records whether PIPELINING has been advertised or not.
5031 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5032 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5033 rather than long int.
5035 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5037 26. Added the never_mail option to autoreply.
5043 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5044 crash if the getsockname() call failed; this can happen if a connection is
5045 closed very soon after it is established. The problem was simply in the
5046 order in which certain operations were done, causing Exim to try to write
5047 to the SMTP stream before it had set up the file descriptor. The bug has
5048 been fixed by making things happen in the correct order.
5054 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5055 before the connection was closed, thus losing the rejection response.
5057 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5058 some early Solaris releases, but causes trouble in current releases where
5059 socklen_t is defined.
5061 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5064 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5067 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5068 that releasing the top of it at the end releases what was used for sub-
5069 expansions (unless the block got too big). However, discard this block if
5070 the first thing is a variable or header, so that we can use its block when
5071 it is dynamic (useful for very large $message_headers, for example).
5073 6. Lookups now cache *every* query, not just the most recent. A new, separate
5074 store pool is used for this. It can be recovered when all lookup caches are
5075 flushed. Lookups now release memory at the end of their result strings.
5076 This has involved some general refactoring of the lookup sources.
5078 7. Some code has been added to the store_xxx() functions to reduce the amount
5079 of flapping under certain conditions.
5081 8. log_incoming_interface used to affect only the <= reception log lines. Now
5082 it causes the local interface and port to be added to several more SMTP log
5083 lines, for example "SMTP connection from", and rejection lines.
5085 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5087 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5089 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5090 could be overwritten at the end of the current message (or the start of a
5091 new message if it was set in a HELO ACL). The value is now preserved for
5092 the duration of the SMTP connection.
5094 12. If a transport had a headers_rewrite setting, and a matching header line
5095 contained an unqualified address, that address was qualified, even if it
5096 did not match any rewriting rules. The underlying bug was that the values
5097 of the flags that permit the existence of unqualified sender and recipient
5098 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5099 non-local messages, and by -bnq for local messages) were not being
5100 preserved with the message after it was received.
5102 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5103 "next input=" as part of the text comprising the host identity instead of
5104 the correct text. The code was using the same buffer for two different
5105 strings. However, depending on which order the printing function evaluated
5106 its arguments, the bug did not always show up. Under Linux, for example, my
5107 test suite worked just fine.
5109 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5110 doesn't work with some older versions of Perl. It has been changed to "my",
5111 which in any case is probably the better facility to use.
5113 15. A really picky compiler found some instances of statements for creating
5114 error messages that either had too many or two few arguments for the format
5117 16. The size of the buffer for calls to the DNS resolver has been increased
5118 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5119 for addresses that have a lot of PTR records. This alleviates a problem; it
5120 does not fully solve it.
5122 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5123 buffer now truncates the list and logs the incident, which is the same
5124 action as happens when Exim is looking up a host name and its aliases.
5125 Previously in this situation something unpredictable would happen;
5126 sometimes it was "internal error: store_reset failed".
5128 18. If a server dropped the connection unexpectedly when an Exim client was
5129 using GnuTLS and trying to read a response, the client delivery process
5130 crashed while trying to generate an error log message.
5132 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5133 string, for example:
5135 warn message = H1: something\nH2: something
5137 the text was added as a single header line from Exim's point of view
5138 though it ended up OK in the delivered message. However, searching for the
5139 second and subsequent header lines using $h_h2: did not work. This has been
5140 fixed. Similarly, if a system filter added multiple headers in this way,
5141 the routers could not see them.
5143 20. Expanded the error message when iplsearch is called with an invalid key to
5144 suggest using net-iplsearch in a host list.
5146 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5147 ACLs are no longer actually imposed (and a message to that effect is
5150 22. If a "gecos" field in a passwd entry contained escaped characters, in
5151 particular, if it contained a \" sequence, Exim got it wrong when building
5152 a From: or a Sender: header from that name. A second bug also caused
5153 incorrect handling when an unquoted " was present following a character
5154 that needed quoting.
5156 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5157 was not being matched caselessly.
5159 24. Arranged for all hyphens in the exim.8 source to be escaped with
5162 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5163 properly. Recipient callout cache records were still being keyed to include
5164 the sender, even when use_sender was set false. This led to far more
5165 callouts that were necessary. The sender is no longer included in the key
5166 when use_sender is false.
5168 26. Added "control = submission" modifier to ACLs.
5170 27. Added the ${base62d: operator to decode base 62 numbers.
5172 28. dnsdb lookups can now access SRV records.
5174 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5175 the configuration file.
5177 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5178 (-v) mode. This makes the output for a verbose queue run more intelligible.
5180 31. Added a use_postmaster feature to recipient callouts.
5182 32. Added the $body_zerocount variable, containing the number of binary zero
5183 bytes in the message body.
5185 33. The time of last modification of the "new" subdirectory is now used as the
5186 "mailbox time last read" when there is a quota error for a maildir
5189 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5191 35. Added +ignore_unknown as a special item in host lists.
5193 36. Code for decoding IPv6 addresses in host lists is now included, even if
5194 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5195 address was recognized as an IP address, but was then not correctly decoded
5196 into binary, causing unexpected and incorrect effects when compared with
5203 1. Very minor rewording of debugging text in manualroute to say "list of
5204 hosts" instead of "hostlist".
5206 2. If verify=header_syntax was set, and a header line with an unqualified
5207 address (no domain) and a large number of spaces between the end of the
5208 name and the colon was received, the reception process suffered a buffer
5209 overflow, and (when I tested it) crashed. This was caused by some obsolete
5210 code that should have been removed. The fix is to remove it!
5212 3. When running in the test harness, delay a bit after writing a bounce
5213 message to get a bit more predictability in the log output.
5215 4. Added a call to search_tidyup() just before forking a reception process. In
5216 theory, someone could use a lookup in the expansion of smtp_accept_max_
5217 per_host which, without the tidyup, could leave open a database connection.
5219 5. Added the variables $recipient_data and $sender_data which get set from a
5220 lookup success in an ACL "recipients" or "senders" condition, or a router
5221 "senders" option, similar to $domain_data and $local_part_data.
5223 6. Moved the writing of debug_print from before to after the "senders" test
5226 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5227 problems for message scanning, either using a data ACL, or using
5228 local_scan() because the Received: header was not generated till after they
5229 were called (in order to set the time as the time of reception completion).
5230 I have revised the way this works. The header is now generated after the
5231 body is received, but before the ACL or local_scan() are called. After they
5232 are run, the timestamp in the header is updated.
5238 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5239 before starting a queue runner without re-exec. This happened only when
5240 deliver_drop_privilege was set or when the Exim user was set to root. The
5241 effect of the bug was that timeouts during subsequent deliveries caused
5242 crashes instead of being properly handled. The handler is now left at its
5243 default (and expected) setting.
5245 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5246 message, again when deliver_drop_privilege is set or Exim is run as root.
5247 The bug described in (1) was not present in this case, but the tidying up
5248 of the other signals was missing. I have made the two cases consistent.
5250 3. The ignore_target_hosts setting on a manualroute router was being ignored
5251 for hosts that were looked up using the /MX notation.
5253 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5256 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5257 operated on the sender address. After changing the $sender_address to <>
5258 for the sender address verify, Exim was re-instated it as the original
5259 (before rewriting) address, but remembering that it had rewritten it, so it
5260 wasn't rewriting it again. This bug also had the effect of breaking the
5261 sender address verification caching when the sender address was rewritten.
5263 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5264 This has been changed so that if the ip literal address matches
5265 ignore_target_hosts, the router declines.
5267 7. Added expansion conditions match_domain, match_address, and match_local_
5268 part (NOT match_host).
5270 8. The placeholder for the Received: header didn't have a length field set.
5272 9. Added code to Exim itself and to exim_lock to test for a specific race
5273 condition that could lead to file corruption when using MBX delivery. The
5274 issue is with the lockfile that is created in /tmp. If this file is removed
5275 after a process has opened it but before that process has acquired a lock,
5276 there is the potential for a second process to recreate the file and also
5277 acquire a lock. This could lead to two Exim processes writing to the file
5278 at the same time. The added code performs the same test as UW imapd; it
5279 checks after acquiring the lock that its file descriptor still refers to
5280 the same named file.
5282 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5283 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5286 11. Added the smtp_active_hostname option. If used, this will typically be made
5287 to depend on the incoming interface address. Because $interface_address is
5288 not set up until the daemon has forked a reception process, error responses
5289 that can happen earlier (such as "too many connections") no longer contain
5292 12. If an expansion in a condition on a "warn" statement fails because a lookup
5293 defers, the "warn" statement is abandoned, and the next ACL statement is
5294 processed. Previously this caused the whole ACL to be aborted.
5296 13. Added the iplsearch lookup type.
5298 14. Added ident_timeout as a log selector.
5300 15. Added tls_certificate_verified as a log selector.
5302 16. Added a global option tls_require_ciphers (compare the smtp transport
5303 option of the same name). This controls incoming TLS connections.
5305 17. I finally figured out how to make tls_require_ciphers do a similar thing
5306 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5307 before starting the TLS session.
5309 18. Tabs are now shown as \t in -bP output.
5311 19. If the log selector return_path_on_delivery was set, Exim crashed when
5312 bouncing a message because it had too many Received: header lines.
5314 20. If two routers both had headers_remove settings, and the first one included
5315 a superfluous trailing colon, the final name in the first list and the
5316 first name in the second list were incorrectly joined into one item (with a
5317 colon in the middle).
5323 1. Added -C and -D options to the exinext utility, mainly to make it easier
5324 to include in the automated testing, but these could be helpful when
5325 multiple configurations are in use.
5327 2. The exinext utility was not formatting the output nicely when there was
5328 an alternate port involved in the retry record key, nor when there was a
5329 message id as well (for retries that were specific to a specific message
5330 and a specific host). It was also confused by IPv6 addresses, because of
5331 the additional colons they contain. I have fixed the IPv4 problem, and
5332 patched it up to do a reasonable job for IPv6.
5334 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5335 delivery, the log line now contains "pipelined" if PIPELINING was used.
5337 4. An SMTP transport process used to panic and die if the bind() call to set
5338 an explicit outgoing interface failed. This has been changed; it is now
5339 treated in the same way as a connect() failure.
5341 5. A reference to $sender_host_name in the part of a conditional expansion
5342 that was being skipped was still causing a DNS lookup. This no longer
5345 6. The def: expansion condition was not recognizing references to header lines
5346 that used bh_ and bheader_.
5348 7. Added the _cache feature to named lists.
5350 8. The code for checking quota_filecount in the appendfile transport was
5351 allowing one more file than it should have been.
5353 9. For compatibility with Sendmail, the command line option
5361 and sets the incoming protocol and host name (for trusted callers). The
5362 host name and its colon can be omitted when only the protocol is to be set.
5363 Note the Exim already has two private options, -pd and -ps, that refer to
5364 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5365 "s", but I don't think that's a major issue.
5367 10. A number of refactoring changes to the code, none of which should affect
5370 (a) The number of logging options was getting close to filling up the
5371 32-bit word that was used as a bit map. I have split them into two classes:
5372 those that are passed in the argument to log_write(), and those that are
5373 only ever tested independently outside of that function. These are now in
5374 separate 32-bit words, so there is plenty of room for expansion again.
5375 There is no change in the user interface or the logging behaviour.
5377 (b) When building, for example, log lines, the code previously used a
5378 macro that called string_cat() twice, in order to add two strings. This is
5379 not really sufficiently general. Furthermore, there was one instance where
5380 it was actually wrong because one of the argument was used twice, and in
5381 one call a function was used. (As it happened, calling the function twice
5382 did not affect the overall behaviour.) The macro has been replaced by a
5383 function that can join an arbitrary number of extra strings onto a growing
5386 (c) The code for expansion conditions now uses a table and a binary chop
5387 instead of a serial search (which was left over from when there were very
5388 few conditions). Also, it now recognizes conditions like "pam" even when
5389 the relevant support is not compiled in: a suitably worded error message is
5390 given if an attempt is made to use such a condition.
5392 11. Added ${time_interval:xxxxx}.
5394 12. A bug was causing one of the ddress fields not to be passed back correctly
5395 from remote delivery subprocesses. The field in question was not being
5396 subsequently used, so this caused to problems in practice.
5398 13. Added new log selectors queue_time and deliver_time.
5400 14. Might have fixed a bug in maildirsizefile handling that threw up
5401 "unexpected character" debug warnings, and recalculated the data
5402 unnecessarily. In any case, I expanded the warning message to give more
5405 15. Added the message "Restricted characters in address" to the statements in
5406 the default ACL that block characters like @ and % in local parts.
5408 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5409 Three changes have been made:
5411 (a) There was a serious bug; a negative response to MAIL caused the whole
5412 recipient domain to be cached as invalid, thereby blocking all messages
5413 to all local parts at the same domain, from all senders. This bug has
5414 been fixed. The domain is no longer cached after a negative response to
5415 MAIL if the sender used is not empty.
5417 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5420 (c) A new callout option, "use_sender" has been added for people who want
5421 the modified behaviour.
5427 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5430 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5431 indeed breaks things for older releases.
5433 3. Added additional logging to the case where there is a problem reading data
5434 from a filter that is running in a subprocess using a pipe, in order to
5435 try to track down a specific problem.
5437 4. Testing facility fudge: when running in the test harness and attempting
5438 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5439 getting "No route to host". Convert this to a timeout.
5441 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5444 6. Some OS don't have socklen_t but use size_t instead. This affects the
5445 fifth argument of getsockopt() amongst other things. This is now
5446 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5447 can be set for individual OS. I have set it for SunOS5, OSF1, and
5448 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5449 some earlier ones do not.
5451 7. Change 4.30/15 was not doing the test caselessly.
5453 8. The standard form for an IPv6 address literal was being rejected by address
5454 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5455 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5456 this, as well as the form without the "IPv6" on the front (but only when
5457 address literals are enabled, of course).
5459 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5461 10. Exim crashed if a message with an empty sender address specified by -f
5462 encountered a router with an errors_to setting. This could be provoked only
5463 by a command such as
5467 where an empty string was supplied; "<>" did not hit this bug.
5469 11. Installed PCRE release 4.5.
5471 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5472 remained set. It is now erased.
5474 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5475 times from message ids (which are base 36 rather than the normal 62).
5477 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5478 were being counted as actual protocol errors, and logged if the log
5479 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5480 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5481 and DATA following a set of rejected RCPTs do not count as protocol errors.
5482 In other words, Exim assumes they were pipelined, though this may not
5483 actually be the case. Of course, in all cases the client gets an
5484 appropriate error code.
5486 15. If a lookup fails in an ACL condition, a message about the failure may
5487 be available; it is used if testing the ACL cannot continue, because most
5488 such messages specify what the cause of the deferral is. However, some
5489 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5490 that caused an old message to be retained and used if a later statement
5491 caused a defer, replacing the real cause of the deferral.
5493 16. If an IP address had so many PTR records that the DNS lookup buffer
5494 was not large enough to hold them, Exim could crash while trying to process
5495 the truncated data. It now detects and logs this case.
5497 17. Further to 4.21/58, another change has been made: if (and only if) the
5498 first line of a message (the first header line) ends with CRLF, a bare LF
5499 in a subsequent header line has a space inserted after it, so as not to
5500 terminate the header.
5502 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5503 unnecessarily, used atoi() instead of strtol(), and didn't check the
5504 termination when getting file sizes from file names by regex.
5506 19. Completely re-implemented the support for maildirsize files, in the light
5507 of a number of problems with the previous contributed implementation
5508 (4.30/29). In particular:
5510 . If the quota is zero, the maildirsize file is maintained, but no quota is
5513 . If the maildir directory does not exist, it is created before any attempt
5514 to write a maildirsize file.
5516 . The quota value in the file is just a cache; if the quota is changed in
5517 the transport, the new value overrides.
5519 . A regular expression is available for excluding directories from the
5522 20. The autoreply transport checks the characters in options that define the
5523 message's headers; it allows continued headers, but it was checking with
5524 isspace() after an embedded newline instead of explicitly looking for a
5527 21. If all the "regular" hosts to which an address was routed had passed their
5528 expiry times, and had not reached their retry times, the address was
5529 bounced, even if fallback hosts were defined. Now Exim should go on to try
5532 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5533 equivalent code in the SMTP transport. Some hosts send humungous responses
5534 to HELO/EHLO, more than 1024 it seems.
5536 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5537 gives compiler warnings in some environments. I've now done it "properly",
5540 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5541 (because of problems with the built-in one) was declared to return uschar *
5542 instead of char *, causing compiler failure.
5544 25. Fixed a file descriptor leak when processing alias/forward files.
5546 26. Fixed a minor format string issue in dbfn.c.
5548 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5550 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5551 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5554 29. When a sender address is verified, it is cached, to save repeating the test
5555 when there is more than one recipient in a message. However, when the
5556 verification involves a callout, it is possible for different callout
5557 options to be set for different recipients. It is too complicated to keep
5558 track of this in the cache, so now Exim always runs a verification when a
5559 callout is required, relying on the callout cache for the optimization.
5560 The overhead is duplication of the address routing, but this should not be
5563 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5564 to be verified with callout=postmaster, and the main callout worked but the
5565 postmaster check failed, the verification correctly failed. However, if a
5566 subsequent RCPT command asked for sender verification *without* the
5567 postmaster check, incorrect caching caused this verification also to fail,
5570 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5571 it was not caching the DNS options (qualify_single, search_parents) that
5572 were used when the lookup failed. A subsequent lookup with different
5573 options therefore always gave the same answer, though there were cases
5574 where it should not have. (Example: a "domains = !$mx_any" option on a
5575 dnslookup router: the "domains" option is always processed without any
5576 widening, but the router might have qualify_single set.) Now Exim uses the
5577 cached value only when the same options are set.
5579 32. Added John Jetmore's "exipick" utility to the distribution.
5581 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5582 than a timeout (e.g. a certificate is required, and is not provided), an
5583 Exim server now closes the connection immediately. Previously it waited for
5584 the client to close - but if the client is SSL, it seems that they each
5585 wait for each other, leading to a delay before one of them times out.
5587 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5588 maintained 0.8.x compatibility because I don't think many are using it, and
5589 it is clearly obsolete.
5591 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5594 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5595 client certificate was expired. A simple patch fixes this, though I don't
5596 understand the full logic of why the verify callback is called multiple
5599 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5600 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5601 which causes problems with some clients (such as the Certicom SSL Plus
5602 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5603 disables the coutermeasure allowing Eudora to connect."
5605 38. Exim was not checking that a write() to a log file succeeded. This could
5606 lead to Bad Things if a log got too big, in particular if it hit a file
5607 size limit. Exim now panics and dies if it cannot write to a log file, just
5608 as it does if it cannot open a log file.
5610 39. Modified OS/Makefile-Linux so that it now contains
5612 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5614 The two -D definitions ensure that Exim is compiled with large file
5615 support, which makes it possible to handle log files that are bigger than
5618 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5619 instance) a domain was checked against a named list that involved a lookup,
5620 causing $domain_data to be set, then another domain was checked against the
5621 same list, then the first domain was re-checked, the value of $domain_data
5622 after the final check could be wrong. In particular, if the second check
5623 failed, it could be set empty. This bug probably also applied to
5626 41. The strip_trailing_dot option was not being applied to the address given
5627 with the -f command-line option.
5629 42. The code for reading a message's header from the spool was incrementing
5630 $received_count, but never initializing it. This meant that the value was
5631 incorrect (doubled) while delivering a message in the same process in which
5632 it was received. In the most common configuration of Exim, this never
5633 happens - a fresh exec is done - but it can happen when
5634 deliver_drop_privilege is set.
5636 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5637 it now includes up to 150 characters of the unexpected data in the log
5640 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5641 and building data strings. The size of both of these buffers was 10 000
5642 bytes - far larger than anybody would *ever* want, thought I. Needless to
5643 say, somebody hit the limit. I have increased the maximum line length to
5644 20 000 and the maximum data length of concatenated lines to 100 000. I have
5645 also fixed two bugs, because there was no checking on these buffers. Tsk,
5646 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5647 buffer is too small.
5649 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5650 lsearch lookups. Now it does.
5652 46. When parsing a route_list item in a manualroute router, a fixed-length
5653 buffer was used for the list of hosts. I made this 1024 bytes long,
5654 thinking that nobody would ever have a list of hosts that long. Wrong.
5655 Somebody had a whole pile of complicated expansion conditions, and the
5656 string was silently truncated, leading to an expansion error. It turns out
5657 that it is easier to change to an unlimited length (owing to other changes
5658 that have happened since this code was originally written) than to build
5659 structure for giving a limitation error. The length of the item that
5660 expands into the list of hosts is now unlimited.
5662 47. The lsearch lookup could not handle data where the length of text line was
5663 more than 4095 characters. Such lines were truncated, leading to shortened
5664 data being returned. It should now handle lines of any length.
5666 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5667 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5670 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5671 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5672 no longer happens in comment lines. A list of replacements is now placed
5673 at the head of all of the source files, except those whose only change is
5674 to replace PERL_COMMAND in the very first #! line.
5676 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5677 messages on the queue, with a bottom-up merge sort, using code contributed
5678 by Michael Haardt. This should make operations like -bp somewhat faster on
5679 large queues. It won't affect queue runners, except when queue_run_in_order
5682 51. Installed eximstats 1.31 in the distribution.
5684 52. Added support for SRV lookups to the dnslookup router.
5686 53. If an ACL referred to $message_body or $message_body_end, the value was not
5687 reset for any messages that followed in the same SMTP session.
5689 54. The store-handling optimization for building very long strings was not
5690 differentiating between the different store pools. I don't think this
5691 actually made any difference in practice, but I've tidied it.
5693 55. While running the routers to verify a sender address, $sender_address
5694 was still set to the sender address. This is wrong, because when routing to
5695 send a bounce to the sender, it would be empty. Therefore, I have changed
5696 it so that, while verifying a sender address, $sender_address is set to <>.
5697 (There is no change to what happens when verifying a recipient address.)
5699 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5700 target A or AAAA records (if not already returned) without resetting the
5701 qualify_single or search_parents options of the DNS resolver. These are
5702 inappropriate in this case because the targets of MX and SRV records must
5703 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5704 target that, when qualified, matched something in the local domain. These
5705 two options are now turned off when doing these lookups.
5707 57. It seems that at least some releases of Reiserfs (which does not have the
5708 concept of a fixed number of inodes) returns zero and not -1 for the
5709 number of available inodes. This interacted badly with check_spool_inodes,
5710 which assumed that -1 was the "no such thing" setting. What I have done is
5711 to check that the total number of inodes is greater than zero before doing
5712 the test of how many are available.
5714 58. When a "warn" ACL statement has a log_message modifier, the message is
5715 remembered, and not repeated. This is to avoid a lot of repetition when a
5716 message has many recipients that cause the same warning to be written.
5717 However, Exim was preserving the list of already written lines for an
5718 entire SMTP session, which doesn't seem right. The memory is now reset if a
5719 new message is started.
5721 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5722 debugging output unless log_rewrite was also set.
5724 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5725 of (int)(handle) when we know that handle contains (void *)(-1).
5727 61. The Exim daemon panic-logs an error return when it closes the incoming
5728 connection. However "connection reset by peer" seems to be common, and
5729 isn't really an error worthy of noting specially, so that particular error
5732 62. When Exim is trying to find all the local interfaces, it used to panic and
5733 die if the ioctl to get the interface flags failed. However, it seems that
5734 on at least one OS (Solaris 9) it is possible to have an interface that is
5735 included in the list of interfaces, but for which you get a failure error
5736 for this call. This happens when the interface is not "plumbed" into a
5737 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5738 failure of the "get flags" call assumes that the interface is down.
5740 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5741 makes life easier for people who are doing arithmetic on fields extracted
5742 from dates, where you often get leading zeros that should not be
5743 interpreted as octal.
5745 64. Added qualify_domain to the redirect router, to override the global
5748 65. If a pathologically long header line contained very many addresses (the
5749 report of this problem mentioned 10 000) and each of them was rewritten,
5750 Exim could use up a very large amount of memory. (It kept on making new
5751 copies of the header line as it rewrote, and never released the old ones.)
5752 At the expense of a bit more processing, the header rewriting function has
5753 been changed so that it no longer eats memory in this way.
5755 66. The generation of the Received: header has been moved from the time that a
5756 message starts to be received, to the time that it finishes. The timestamp
5757 in the Received: header should now be very close to that of the <= log
5758 line. There are two side-effects of this change:
5760 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5761 logged header lines no longer include the local Received: line, because
5762 it has not yet been created. The same applies to a copy of the message
5763 that is returned to a non-SMTP sender when a message is rejected.
5765 (b) When a filter file is tested using -bf, no additional Received: header
5766 is added to the test message. After some thought, I decided that this
5769 This change does not affect the value of $received_for. It is still set
5770 after address rewriting, but before local_scan() is called.
5772 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5774 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5775 gave an unhelpful panic error message, and a defer error. I have managed to
5776 change this behaviour so that it now rejects any supplied certificate,
5777 which seems right, as the list of acceptable certificates is empty.
5779 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5780 gave an unhelpful defer error. I have not managed to make this reject any
5781 supplied certificates, but the error message it gives is "no certificate
5782 supplied", which is not helpful.
5784 70. exigrep's output now also includes lines that are not associated with any
5785 message, but which match the given pattern. Implemented by a patch from
5786 Martin Sluka, which also tidied up the Perl a bit.
5788 71. Recipient callout verification, like sender verification, was using <> in
5789 the MAIL FROM command. This isn't really the right thing, since the actual
5790 sender may affect whether the remote host accepts the recipient or not. I
5791 have changed it to use the actual sender in the callout; this means that
5792 the cache record is now keyed on a recipient/sender pair, not just the
5793 recipient address. There doesn't seem to be a real danger of callout loops,
5794 since a callout by the remote host to check the sender would use <>.
5795 [SEE ABOVE: changed after hitting problems.]
5797 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5798 temporary errors. However, in the case of such a code being given after
5799 the end of a data transmission (i.e. after ".") Exim was failing to write
5800 a retry record for the message. (Yes, there was some broken host that was
5801 actually sending 8xx at this point.)
5803 73. An unknown lookup type in a host list could cause Exim to panic-die when
5804 the list was checked. (An example that provoked this was putting <; in the
5805 middle of a list instead of at the start.) If this happened during a DATA
5806 ACL check, a -D file could be left lying around. This kind of configuration
5807 error no longer causes Exim to die; instead it causes a defer error. The
5808 incident is still logged to the main and panic logs.
5810 74. Buglet left over from Exim 3 conversion. The message "too many messages
5811 in one connection" was written to the rejectlog but not the mainlog, except
5812 when address rewriting (yes!) was being logged.
5814 75. Added write_rejectlog option.
5816 76. When a system filter was run not as root (that is, when system_filter_user
5817 was set), the values of the $n variables were not being returned to the
5818 main process; thus, they were not subsequently available in the $sn
5821 77. Added +return_path_on_delivery log selector.
5823 78. A connection timeout was being treated differently from recipients deferred
5824 when testing hosts_max_try with a message that was older than the host's
5825 retry timeout. (The host should not be counted, thus allowing all hosts to
5826 be tried at least once before bouncing.) This may have been the cause of an
5827 occasionally reported bug whereby a message would remain on the queue
5828 longer than the retry timeout, but would be bounced if a delivery was
5829 forced. I say "may" because I never totally pinned down the problem;
5830 setting up timeout/retry tests is difficult. See also the next item.
5832 79. The ultimate address timeout was not being applied to errors that involved
5833 a combination of host plus message (for example, a timeout on a MAIL
5834 command). When an address resolved to a number of possible hosts, and they
5835 were not all tried for each delivery (e.g. because of hosts_max_try), a
5836 message could remain on the queue longer than the retry timeout.
5838 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5841 81. Fixed an obscure SMTP outgoing bug which required at least the following
5842 conditions: (a) there was another message waiting for the same server;
5843 (b) the server returned 5xx to all RCPT commands in the first message so
5844 that the message was not completed; (c) the server dropped the connection
5845 or gave a negative response to the RSET that Exim sends to abort the
5846 transaction. The observed case was a dropped connection after DATA that had
5847 been sent in pipelining mode. That is, the server had advertised PIPELINING
5848 but was not implementing it correctly. The effect of the bug was incorrect
5849 behaviour, such as trying another host, and this could lead to a crash.
5855 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5856 and daemon.c were passed as pointers to ints; they should have been
5857 pointers to socklen_t variables (which are typically unsigned ints).
5859 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5862 3. Fixed a really odd bug that affected only the testing scheme; patching a
5863 certain fixed string in the binary changed the value of another string that
5864 happened to be identical to the end of the original first string.
5866 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5867 name", it returns that address as the IP address. On some operating
5868 systems (e.g. Solaris), it also passes back the IP address string as the
5869 "host name". However, on others (e.g. Linux), it passes back an empty
5870 string. Exim wasn't checking for this, and was changing the host name to an
5871 empty string, assuming it had been canonicalized.
5873 5. Although rare, it is permitted to have more than one PTR record for a given
5874 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5875 all the names associated with an address, because they do in Solaris.
5876 However, it seems that they do not in Linux for data that comes from the
5877 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5878 I found this out when I moved to a new Linux workstation and tried to run
5879 the Exim test suite.
5881 To get round this problem I have changed the code so that it now does its
5882 own call to the DNS to look up PTR records when searching for a host name.
5883 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5884 addresses that are only in /etc/hosts are still found.
5886 This behaviour is, however, controlled by an option called host_lookup_
5887 order, which defaults to "bydns:byaddr". If people want to use the other
5888 order, or indeed, just use one or the other means of lookup, they can
5889 specify it in this variable.
5891 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5892 some operating systems, this comes back from gethostbyaddr() as an empty
5893 string, and this is what Exim used to test for. However, it seems that in
5894 other systems, "." is yielded. Exim now tests for this case too.
5896 7. The values of check_spool_space and check_log_space are now held internally
5897 as a number of kilobytes instead of an absolute number of bytes. If a
5898 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5899 kilobyte. This means that much larger values can be stored.
5901 8. Exim monitor: an attempt to get the action menu when not actually pointing
5902 at a message produces an empty menu entitled "No message selected". This
5903 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5904 no entries in it ("Shell widget menu has zero width and/or height"). So I
5905 have added a single, blank menu entry in this case.
5907 9. Added ${quote_local_part.
5909 10. MIME decoding is now applied to the contents of Subject: header lines when
5912 11. Now that a reference to $sender_host_address automatically causes a reverse
5913 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5914 host lookup before query-style lookups in lists that might use this
5915 variable. This has therefore been abolished, and the "net-" prefix is no
5916 longer necessary for query-style lookups.
5918 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5919 have been a typo for LFLAGS, so it has been changed.
5921 13. The install script calls Exim with "-C /dev/null" in order to find the
5922 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5923 to be output. However, since Exim outputs its version number before the
5924 error, it didn't break the script. It just looked ugly. I fixed this by
5925 always allowing "-C /dev/null" if the caller is root.
5927 14. Ignore overlarge ACL variable number when reading spool file - insurance
5928 against a later release with more variables having written the file.
5930 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5931 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5932 this, as well as the form without the "IPv6" on the front.
5934 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5935 OS/Makefile-Darwin file.
5937 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5938 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5940 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5941 variable. However, some OS use ints for inodes. I've added cast to long int
5942 to get rid of the compiler warning.
5944 19. I had forgotten to lock out "/../" in configuration file names when
5945 ALT_CONFIG_PREFIX was set.
5947 20. Routers used for verification do not need to specify transports. However,
5948 if such a router generated a host list, and callout was configured, Exim
5949 crashed, because it could not find a port number from the (non-existent)
5950 transport. It now assumes port 25 in this circumstance.
5952 21. Added the -t option to exigrep.
5954 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5955 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5956 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5957 them is set, LOOKUP_LSEARCH is forced.
5959 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5961 24. Added sender and host information to the "rejected by local_scan()" log
5962 line; previously there was no indication of these.
5964 25. Added .include_if_exists.
5966 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5967 Linux. It turns out that not all file systems support this. Apparently some
5968 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5969 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5970 not supported on the file descriptor, is now ignored when Exim is trying to
5971 sync a directory. This applies only to Linux.
5973 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5975 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5976 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5979 29. Added support for maildirsize files from supplied patch (modified a bit).
5981 30. The use of :fail: followed by an empty string could lead Exim to respond to
5982 sender verification failures with (e.g.):
5984 550 Verification failed for <xxx>
5985 550 Sender verify failed
5987 where the first response line was missing the '-' that indicates it is not
5988 the final line of the response.
5990 31. The loop for finding the name of the user that called Exim had a hardwired
5991 limit of 10; it now uses the value of finduser_retries, which is used for
5992 all other user lookups.
5994 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5997 33. Exim was neglecting to zero errno before one call of strtol() when
5998 expanding a string and expecting an integer value. On some systems this
5999 resulted in spurious "integer overflow" errors. Also, it was casting the
6000 result into an int without checking.
6002 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6003 did not work. The code looks as if it has *never* worked, though it appears
6004 to have been documented since at least release 1.62. I have made it work.
6006 35. The "timeout_DNS" error in retry rules, also documented since at least
6007 1.62, also never worked. As it isn't clear exactly what this means, and
6008 clearly it isn't a major issue, I have abolished the feature by treating it
6009 as "timeout", and writing a warning to the main and panic logs.
6011 36. The display of retry rules for -brt wasn't always showing the error code
6014 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6015 timeout_connect_A, timeout_connect_MX.
6017 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6018 to the empty sender.
6020 39. The daemon was not analysing the content of -oX till after it had closed
6021 stderr and disconnected from the controlling terminal. This meant that any
6022 syntax errors were only noted on the panic log, and the return code from
6023 the command was 0. By re-arranging the code a little, I've made the
6024 decoding happen first, so such errors now appear on stderr, and the return
6025 code is 1. However, the actual setting up of the sockets still happens in
6026 the disconnected process, so errors there are still only recorded on the
6029 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6030 connections (as happens on some IP stacks) was logged at start up time as
6031 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6032 it from "IPv6 and IPv4", which means that two separate sockets are being
6035 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6036 says whether AF_INET or AF_INET6 was passed as an argument.
6038 42. Exiwhat output was messed up when time zones were included in log
6041 43. Exiwhat now gives more information about the daemon's listening ports,
6042 and whether -tls-on-connect was used.
6044 44. The "port" option of the smtp transport is now expanded.
6046 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6047 silently ignored. Now an error message is written to the main and panic
6050 46. There's a new ACL modifier called "logwrite" which writes to a log file
6051 as soon as it is encountered.
6053 47. Added $local_user_uid and $local_user_gid at routing time.
6055 48. Exim crashed when trying to verify a sender address that was being
6058 49. Exim was recognizing only a space character after ".include". It now also
6059 recognizes a tab character.
6061 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6062 extracting the relevant information from the specification. The man page no
6063 longer contains scrambled data for the -d option, and I've added a section
6064 at the front about calling Exim under different names.
6066 51. Added "extra_headers" argument to the "mail" command in filter files.
6068 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6071 53. Installed eximstats 1.29.
6073 54. Added transport_filter_timeout as a generic transport option.
6075 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6076 Cc: header lines. This was required by RFC 822, but it not required by RFC
6079 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6080 incoming messages that did not have them. Now it does so only if the
6081 message originates locally, that is, if there is no associated remote host
6082 address. When Resent- header lines are present, this applies to the Resent-
6083 lines rather than the non-Resent- lines.
6085 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6086 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6088 58. Messages for configuration errors now include the name of the main
6089 configuration file - useful now that there may be more than one file in a
6090 list (.included file names were always shown).
6092 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6093 for those rare installations that do not start the daemon as root or run it
6094 setuid root. I've cut out the call to initgroups() if the daemon is not
6097 60. The Exim user and group can now be bound into the binary as text strings
6098 that are looked up at the start of Exim's processing.
6100 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6102 62. Added $mailstore_basename variable.
6104 63. Installed patch to sieve.c from Michael Haardt.
6106 64. When Exim failed to open the panic log after failing to open the main log,
6107 the original message it was trying to log was written to stderr and debug
6108 output, but if they were not available (the usual case in production), it
6109 was lost. Now it is written to syslog before the two lines that record the
6110 failures to open the logs.
6112 65. Users' Exim filters run in subprocesses under the user's uid. It is
6113 possible for a "deliver" command or an alias in a "personal" command to
6114 provoke an address rewrite. If logging of address rewriting is configured,
6115 this fails because the process is not running as root or exim. There may be
6116 a better way of dealing with this, but for the moment (because 4.30 needs
6117 to be released), I have disabled address rewrite logging when running a
6118 filter in a non-root, non-exim process.
6124 1. The buildconfig auxiliary program wasn't quoting the value set for
6125 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6126 not defined. This bug was masked in 4.22 by the effect that was fixed in
6129 2. Some messages that were rejected after a message id was allocated were
6130 shown as "incomplete" by exigrep. It no longer does this for messages that
6131 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6133 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6134 have allow_domain_literals set, the ID did not get logged in the <= line.
6135 Domain literals are now always recognized in Message-ID: header lines.
6137 4. The first argument for a ${extract expansion item is the key name or field
6138 number. Leading and trailing spaces in this item were not being ignored,
6139 causing some misleading effects.
6141 5. When deliver_drop_privilege was set, single queue runner processes started
6142 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6143 same command in the process it spins off) were not dropping privilege.
6145 6. When the daemon running as "exim" started a queue runner, it always
6146 re-executed Exim in the spun-off process. This is a waste of effort when
6147 deliver_drop_privilege is set. The new process now just calls the
6148 queue-runner function directly.
6154 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6157 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6158 ignored. Though the use of -oP was forcing the writing of a pid file, it
6159 was always written to the default place.
6161 3. If the message "no IP address found for host xxxx" is generated during
6162 incoming verification, it is now followed by identification of the incoming
6163 connection (so you can more easily find what provoked it).
6165 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6167 5. Added some features to "harden" Exim a bit more against certain attacks:
6169 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6170 be put in Local/Makefile. This is like the never_users runtime option,
6171 but it cannot be overridden. The default setting is "root".
6173 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6174 prefix string with which any file named in a -C command line option
6177 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6178 is retained for -C and -D only if the caller of Exim is root. Without
6179 it, the exim user may also use -C and -D and retain privilege.
6181 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6182 command line option is disabled.
6184 6. Macro names set by the -D option must start with an upper case letter, just
6185 like macro names defined in the configuration file.
6187 7. Added "dereference=" facility to LDAP.
6189 8. Two instances of the typo "uknown" in the source files are fixed.
6191 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6192 the Configure-Makefile script screwed up while processing it.
6194 10. Incorporated PCRE 4.4.
6196 11. The SMTP synchronization check was not operating right at the start of an
6197 SMTP session. For example, it could not catch a HELO sent before the client
6198 waited for the greeting. There is now a check for outstanding input at the
6199 point when the greeting is written. Because of the duplex, asynchronous
6200 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6201 way, but not yet received, when the check is performed.
6203 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6204 on TCP/IP sockets, because this apparently causes some broken clients to
6207 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6208 unchanged) from the Cygwin maintainer.
6210 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6211 was supported instead of testing for mbx. Effectively a typo.
6213 15. The spa authenticator server code was not checking that the input it
6214 received was valid base64.
6216 16. The debug output line for the "set" modifier in ACLs was not showing the
6217 name of the variable that was being set.
6219 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6221 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6222 lookup on its own. Something else was needed to trigger the lookup. For
6223 example, a match in host_lookup or the need for a host name in a host list.
6224 Now, if $sender_host_name is referenced and the host name has not yet been
6225 looked up, a lookup is performed. If the lookup fails, the variable remains
6226 empty, and $host_lookup_failed is set to "1".
6228 19. Added "eqi" as a case-independent comparison operator.
6230 20. The saslauthd authentication condition could segfault if neither service
6231 nor realm was specified.
6233 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6234 error message that was logged was misleading, and incoming SMTP
6235 connections were dropped. The message is now more accurate, and temporary
6236 errors are given to SMTP connections.
6238 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6239 (except RSET) with a 421 error, until QUIT is received. However, it was
6240 failing to send a response to QUIT.
6242 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6243 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6244 ACL is now run after the helo_try_verify_hosts code.
6246 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6247 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6248 case-independent, so other case variants are also recognized). Apparently
6249 some systems use these upper case variants.
6251 25. If more than two messages were waiting for the same host, and a transport
6252 filter was specified for the transport, Exim sent two messages over the
6253 same TCP/IP connection, and then failed with "socket operation on non-
6254 socket" when it tried to send the third.
6256 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6258 27. The extern definition of crypt16() in expand.c was not being excluded when
6259 the OS had its own crypt16() function.
6261 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6262 as a preferred synonym for return_size_limit, both as an option and as an
6265 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6267 30. Changed the default configuration ACL to relax the local part checking rule
6268 for addresses that are not in any local domains. For these addresses,
6269 slashes and pipe symbols are allowed within local parts, but the sequence
6270 /../ is explicitly forbidden.
6272 31. SPA server authentication was not clearing the challenge buffer before
6275 32. log_message in a "warn" ACL statement was writing to the reject log as
6276 well as to the main log, which contradicts the documentation and doesn't
6277 seem right (because no rejection is happening). So I have stopped it.
6279 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6280 However, I am unable to do any testing of this.
6282 34. Fixed an infelicity in the appendfile transport. When checking directories
6283 for a mailbox, to see if any needed to be created, it was accidentally
6284 using path names with one or more superfluous leading slashes; tracing
6285 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6287 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6288 contents are added to the log line that is written for every discarded
6289 recipient. (Previously a log_message setting was ignored.)
6291 36. The ${quote: operator now quotes the string if it is empty.
6293 37. The install script runs exim in order to find its version number. If for
6294 some reason other than non-existence or emptiness, which it checks, it
6295 could not run './exim', it was installing it with an empty version number,
6296 i.e. as "exim-". This error state is now caught, and the installation is
6299 38. An argument was missing from the function that creates an error message
6300 when Exim fails to connect to the socket for saslauthd authentication.
6301 This could cause Exim to crash, or give a corrupted message.
6303 39. Added isip, isip4, and isip6 to ${if conditions.
6305 40. The ACL variables $acl_xx are now saved with the message, and can be
6306 accessed later in routers, transports, and filters.
6308 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6309 strings in the file are not string-expanded.
6311 42. If a MAIL command specified a SIZE value that was too large to fit into an
6312 int variable, the check against message_size_limit failed. Such values are
6313 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6314 day this will have to be increased, but I don't think I want to be around
6315 when emails are that large.
6322 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6323 iconv() is not standard in FreeBSD.
6325 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6326 IPv6 enabled. The observed symptom was a segmentation fault on return from
6327 the function os_common_find_running_interfaces() in src/os.c.
6329 3. In the check_special_case() function in daemon.c I had used "errno" as an
6330 argument name, which causes warnings on some systems. This was basically a
6331 typo, since it was named "eno" in the comments!
6333 4. The code that waits for the clock to tick (at a resolution of some fraction
6334 of a second) so as to ensure message-id uniqueness was always waiting for
6335 at least one whole tick, when it could have waited for less. [This is
6336 almost certainly not relevant at current processor speeds, where it is
6337 unlikely to ever wait at all. But we try to future-proof.]
6339 5. The function that sleeps for a time interval that includes fractions of a
6340 second contained a race. It did not block SIGALRM between setting the
6341 timer, and suspending (a couple of lines later). If the interval was short
6342 and the sigsuspend() was delayed until after it had expired, the suspension
6343 never ended. On busy systems this could lead to processes getting stuck for
6346 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6347 process, before it forks any delivery processes. The open lookup caching
6348 mechanism meant that the open file or database connection was passed into
6349 the delivery process. The problem was that delivery processes always tidy
6350 up cached lookup data. This could cause a problem for the next delivery
6351 process started by the queue runner, because the external queue runner
6352 process does not know about the closure. So the next delivery process
6353 still has data in the lookup cache. In the case of a file lookup, there was
6354 no problem because closing a file descriptor in a subprocess doesn't affect
6355 the parent. However, if the lookup was caching a connection to a database,
6356 the connection was closed, and the second delivery process was likely to
6357 see errors such as "PGSQL: query failed: server closed the connection
6358 unexpectedly". The problem has been fixed by closing all cached lookups
6359 in a queue runner before running a delivery process.
6361 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6362 seem to have the "const" qualifier which it has on other OS. I've
6365 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6366 *of the same type* (client or server) with the same public name that an
6367 error should be diagnosed.
6369 9. When Exim looked up a host name for an IP address, but failed to find the
6370 original IP address when looking up the host name (a safety check), it
6371 output the message "<ip address> does not match any IP for NULL", which was
6372 confusing, to say the least. The bug was that the host name should have
6373 appeared instead of "NULL".
6375 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6376 user that is built into the binary, and the -C or -D options is used, root
6377 privilege is dropped before the configuration file is read. In addition,
6378 logging is switched to stderr instead of the normal log files. If the
6379 configuration then re-defines the Exim user, the unprivileged environment
6380 is probably not what is expected, so Exim logs a panic warning message (but
6383 However, if deliver_drop_privilege is set, the unprivileged state may well
6384 be exactly what is intended, so the warning has been cut out in that case,
6385 and Exim is allowed to try to write to its normal log files.
6391 1. smtp_return_error_details was not giving details for temporary sender
6392 or receiver verification errors.
6394 2. Diagnose a configuration error if two authenticators have the same public
6397 3. Exim used not to create the message log file for a message until the first
6398 delivery attempt. This could be confusing when incoming messages were held
6399 for policy or load reasons. The message log file is now created at the time
6400 the message is received, and an initial "Received" line is written to it.
6402 4. The automatically generated man page for command line options had a minor
6403 bug that caused no ill effects; however, a more serious problem was that
6404 the procedure for building the man page automatically didn't always
6405 operate. Consequently, release 4.20 contains an out-of-date version. This
6406 shouldn't happen again.
6408 5. When building Exim with embedded Perl support, the script that builds the
6409 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6410 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6412 6. The freeze_tell option was not being used for messages that were frozen on
6413 arrival, either by an ACL or by local_scan().
6415 7. Added the smtp_incomplete_transaction log selector.
6417 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6418 was accepting AUTH without a new EHLO.
6420 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6421 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6424 10. Logging of TCP/IP connections (when configured) now happens in the main
6425 daemon process instead of the child process, so that the TCP/IP connection
6426 count is more accurate (but it can never be perfect).
6428 11. The use of "drop" in a nested ACL was not being handled correctly in the
6429 outer ACL. Now, if condition failure induced by the nested "drop" causes
6430 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6431 or "require"), the connection is dropped.
6433 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6434 that yield "discard" can now be used with an "accept" or a "discard" verb,
6435 but an error is generated for any others (because I can't see a useful way
6436 to define what should happen).
6438 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6439 are now processed in the same way as lines in the Exim configuration file.
6440 In particular, continuation lines are supported.
6442 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6444 15. Added -ti meaning -t -i.
6446 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6447 domains, and warn by logging if others are found.
6449 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6450 returned from GIFCONF when Exim is trying to find the list of interfaces on
6451 a host. The code in os.c has been modified to copy each ifreq to an aligned
6452 structure in all cases.
6454 Also, in some cases, the returned ifreq's were being copied to a 'struct
6455 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6456 means the last couple of bytes of an IPv6 address could be chopped if the
6457 ifreq contained only a normal sockaddr (14 bytes storage).
6459 18. Named domain lists were not supported in the hosts_treat_as_local option.
6460 An entry such as +xxxx was not recognized, and was treated as a literal
6463 19. Ensure that header lines added by a DATA ACL are included in the reject log
6464 if the ACL subsequently rejects the message.
6466 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6467 MD5 (which is deprecated).
6469 21. When testing a filter file using -bf, Exim was writing a message when it
6470 took the sender from a "From " line in the message, but it was not doing so
6471 when it took $return_path from a Return-Path: header line. It now does.
6473 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6474 with a valid header line field name (a series of printing characters
6475 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6477 23. Changed "disc" in the source to "disk" to conform to the documentation and
6478 the book and for uniformity.
6480 24. Ignore Sendmail's -Ooption=value command line item.
6482 25. When execve() failed while trying to run a command in a pipe transport,
6483 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6484 could be confused with a return value of 69 from the command itself. This
6485 has been changed to 127, the value the shell returns if it is asked to run
6486 a non-existent command. The wording for the related log line suggests a
6487 non-existent command as the problem.
6489 26. If received_header_text expands to an empty string, do not add a Received:
6490 header line to the message. (Well, it adds a token one on the spool, but
6491 marks it "old" so that it doesn't get used or transmitted.)
6493 27. Installed eximstats 1.28 (addition of -nt option).
6495 28. There was no check for failure on the call to getsockname() in the daemon
6496 code. This can fail if there is a shortage of resources on the system, with
6497 ENOMEM, for example. A temporary error is now given on failure.
6499 29. Contrary to the C standard, it seems that in some environments, the
6500 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6501 program. Exim now does this explicitly; it affects the formatting of
6502 timestamps using strftime().
6504 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6505 complaints. I've now initialized all the variables, to avoid this.
6507 32. Header lines added by a system filter were not being "seen" during
6508 transport-time rewrites.
6510 33. The info_callback() function passed to OpenSSL is set up with type void
6511 (*)(SSL *, int, int), as described somewhere. However, when calling the
6512 function (actually a macro) that sets it up, the type void(*)() is
6513 expected. I've put in a cast to prevent warnings from picky compilers.
6515 34. If a DNS black list lookup found a CNAME record, but there were no A
6516 records associated with the domain it pointed at, Exim crashed.
6518 35. If a DNS black list lookup returned more than one A record, Exim ignored
6519 all but the first. It now scans all returned addresses if a particular IP
6520 value is being sought. In this situation, the contents of the
6521 $dnslist_value variable are a list of all the addresses, separated by a
6524 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6525 to accept a host name and all its aliases if the forward lookup for any of
6526 them yielded the IP address of the incoming connection. Now it accepts only
6527 those names whose forward lookup yields the correct IP address. Any other
6528 names are discarded. This closes a loophole whereby a rogue DNS
6529 administrator could create reverse DNS records to break through a
6530 wildcarded host restriction in an ACL.
6532 37. If a user filter or a system filter that ran in a subprocess used any of
6533 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6534 the wrong values were passed to the pipe command ($thisaddress had the
6535 value of $0, $0 had the value of $1, etc). This bug was introduced by
6536 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6538 38. Improved the line breaking for long SMTP error messages from ACLs.
6539 Previously, if there was no break point between 40 and 75 characters, Exim
6540 left the rest of the message alone. Two changes have been made: (a) I've
6541 reduced the minimum length to 35 characters; (b) if it can't find a break
6542 point between 35 and 75 characters, it looks ahead and uses the first one
6543 that it finds. This may give the occasional overlong line, but at least the
6544 remaining text gets split now.
6546 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6547 file descriptors might be low, and that setting 1000 would always raise it.
6548 It turns out that in some environments, the limit is already over 1000 and
6549 that lowering it causes trouble. So now Exim takes care not to decrease it.
6551 40. When delivering a message, the value of $return_path is set to $sender_
6552 address at the start of routing (routers may change the value). By an
6553 oversight, this default was not being set up when an address was tested by
6554 -bt or -bv, which affected the outcome if any router or filter referred to
6557 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6558 log only when "message" or "log_message" are set. However, if one of the
6559 conditions was an address verification, or a call to a nested ACL, the
6560 messages generated by the underlying test were being passed through. This
6561 no longer happens. The underlying message is available in $acl_verify_
6562 message for both "message" and "log_message" expansions, so it can be
6563 passed through if needed.
6565 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6566 new expansion $bh_ to give the encoded byte string without charset
6567 translation. Translation happens only if iconv() is available; HAVE_ICONV
6568 indicates this at build time. HEADERS_CHARSET gives the charset to
6569 translate to; headers_charset can change it in the configuration, and
6570 "headers charset" can change it in an individual filter file.
6572 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6573 that creates RFC 2047 encoded "words" labels them as that charset instead
6574 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6575 expansion operator; (ii) when Exim creates a From: line for a local
6576 message; (iii) when a header line is rewritten to include a "phrase" part.
6578 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6579 buggy, causing it to skip the first lines of messages whose message ID
6580 ended in 'D'. This would not have bitten before Exim release 4.14, because
6581 message IDs were unlikely to end in 'D' before then. The effect was to have
6582 incorrect size information for certain domains.
6584 45. #include "config.h" was missing at the start of the crypt16.c module. This
6585 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6588 46. If there was a timeout during a "random" callout check, Exim treated it as
6589 a failure of the random address, and carried on sending RSET and the real
6590 address. If the delay was just some slowness somewhere, the response to the
6591 original RCPT would be taken as a response to RSET and so on, causing
6592 mayhem of various kinds.
6594 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6595 when I implemented it. It didn't allow for the fact that some option values
6596 may legitimately be negative (e.g. size_addition), and it didn't even do
6597 the right test for positive values.
6599 48. Domain names in DNS records are case-independent. Exim always looks them up
6600 in lower case. Some resolvers return domain names in exactly the case they
6601 appear in the zone file, that is, they may contain uppercase letters. Not
6602 all resolvers do this - some return always lower case. Exim was treating a
6603 change of case by a resolver as a change of domain, similar to a widening
6604 of a domain abbreviation. This triggered its re-routing code and so it was
6605 trying to route what was effectively the same domain again. This normally
6606 caused routing to fail (because the router wouldn't handle the domain
6607 twice). Now Exim checks for this case specially, and just changes the
6608 casing of the domain that it ultimately uses when it transmits the message
6611 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6614 50. If a filter generated a file delivery with a non-absolute name (possible if
6615 no home directory exists for the router), the forbid_file option was not
6618 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6619 the existing equality matching.
6621 52. Exim was using ints instead of ino_t variables in some places where it was
6622 dealing with inode numbers.
6624 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6625 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6626 called TMPDIR, and if it finds it is different, it changes its value.
6628 54. The smtp_printf() function is now made available to local_scan() so
6629 additional output lines can be written before returning. There is also an
6630 smtp_fflush() function to enable the detection of a dropped connection.
6631 The variables smtp_input and smtp_batched_input are exported to
6634 55. Changed the default runtime configuration: the message "Unknown user"
6635 has been removed from the ACL, and instead placed on the localuser router,
6636 using the cannot_route_message feature. This means that any verification
6637 failures that generate their own messages won't get overridden. Similarly,
6638 the "Unrouteable address" message that was in the ACL for unverifiable
6639 relay addresses has also been removed.
6641 56. Added hosts_avoid_esmtp to the smtp transport.
6643 57. The exicyclog script was not checking for the esoteric option
6644 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6645 will work only if exicyclog is run under the appropriate euid.
6647 58. Following a discussion on the list, the rules by which Exim recognises line
6648 endings on incoming messages have been changed. The -dropcr and drop_cr
6649 options are now no-ops, retained only for backwards compatibility. The
6650 following line terminators are recognized: LF CRLF CR. However, special
6651 processing applies to CR:
6653 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6654 nor a local message in the state where . is a terminator.
6656 (ii) If a bare CR is encountered in a header line, an extra space is added
6657 after the line terminator so as not to end the header. The reasoning
6658 behind this is that bare CRs in header lines are most likely either
6659 to be mistakes, or people trying to play silly games.
6661 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6662 was being incorrectly given as 18 bytes larger than it should have been.
6663 This is a VOB (very old bug).
6665 60. This may never have affected anything current, but just in case it has:
6666 When the local host is found other than at the start of a list of hosts,
6667 the local host, those with the same MX, and any that follow, are discarded.
6668 When the list in question was part of a longer list of hosts, the following
6669 hosts (not currently being processed) were also being discarded. This no
6670 longer happens. I'm not sure if this situation could ever has previously
6673 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6676 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6677 header. This is something that is recommended in a new Internet Draft, and
6678 is something that is documented as being done by Sendmail. There are two
6679 possible values. For messages generated by the autoreply transport, Exim
6682 Auto-Submitted: auto-replied
6684 whereas for all other generated messages (e.g. bounces) it adds
6686 Auto-Submitted: auto-generated
6688 63. The "personal" condition in filters now includes a test for the
6689 Auto-Submitted: header. If it contains the string "auto-" the message it
6690 not considered personal.
6692 64. Added rcpt_include_affixes as a generic transport option.
6694 65. Added queue_only_override (default true).
6696 66. Added the syslog_duplication option.
6698 67. If what should have been the first header line of a message consisted of
6699 a space followed by a colon, Exim was mis-interpreting it as a header line.
6700 It isn't of course - it is syntactically invalid and should therefore be
6701 treated as the start of the message body. The misbehaviour could have
6702 caused a number of strange effects, including loss of data in subsequent
6703 header lines, and spool format errors.
6705 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6706 client host had authenticated. This control can now be exercised by an ACL
6707 for more flexibility.
6709 69. By default, callouts do not happen when testing with -bh. There is now a
6710 variant, -bhc, which does actually run the callout code, including
6711 consulting and updating the callout cache.
6713 70. Added support for saslauthd authentication, courtesy of Alexander
6716 71. If statvfs() failed on the spool or log directories while checking their
6717 size for availability, Exim confusingly gave the error "space shortage".
6718 Furthermore, in debugging mode it crashed with a floating point exception.
6719 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6720 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6721 serious problem, Exim now writes to the main and panic logs when this
6722 happens, with details of the failure. It then refuses to accept the
6723 incoming message, giving the message "spool directory problem" or "log
6724 directory problem" with a 421 code for SMTP messages.
6726 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6727 0, 1, and 2 exist, because some OS complain for execs without them (see
6728 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6729 descriptors. However, the code omitted to check that the open succeeded,
6730 causing mysterious errors if for some reason the permissions on /dev/null
6731 got screwed. Now Exim writes a message to the main and panic logs, and
6732 bombs out if it can't open /dev/null.
6734 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6735 interact so that it is all more flexible. It is supposed to remain
6736 backwards compatible. Also added extra_local_interfaces.
6738 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6739 to bomb out with an assertion failure - to the client this appears as a
6740 connection drop. This problem occurs in the part of the code that was taken
6741 from the Samba project. Fortunately, the assertion is in a very simple
6742 function, so I have fixed this by reproducing the function inline in the
6743 one place where it is called, and arranging for authentication to fail
6744 instead of killing the process with assert().
6746 75. The SPA client code was not working when the server requested OEM rather
6747 than Unicode encoding.
6749 76. Added code to make require_files with a specific uid setting more usable in
6750 the case where statting the file as root fails - usually a non-root-mounted
6751 NFS file system. When this happens and the failure is EACCES, Exim now
6752 forks a subprocess and does the per-uid checking as the relevant uid.
6754 77. Added process_log_path.
6756 78. If log_file_path was not explicitly set, a setting of check_log_space or
6757 check_log_inodes was ignored.
6759 79. If a space check for the spool or log partitions fails, the incident is now
6760 logged. Of course, in the latter case the data may get lost...
6762 80. Added the %p formatting code to string_format() so that it can be used to
6763 print addresses in debug_print(). Adjusted all the address printing in the
6764 debugging in store.c to use %p rather than %d.
6766 81. There was a concern that a line of code in smtp_in.c could overflow a
6767 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6768 initially expressed, the concern was not well-founded, because trailing
6769 spaces are removed early. However, if the trailing spaces were followed by
6770 a NULL, they did not get removed, so the overflow was possible. Two fixes
6773 (a) I re-wrote the offending code in a cleaner fashion.
6774 (b) If an incoming SMTP command contains a NULL character, it is rejected
6777 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6778 runs initgroups(), so that if the Exim user is in any additional groups,
6779 they will be used during message reception.
6785 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob