1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
40 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
41 Disable the session-cache too, which might reduce our load. Since we
42 currrectly use a new context for every connection, both as server and
43 client, there is no benefit for these.
44 GnuTLS appears to not support tickets server-side by default (we don't
45 call gnutls_session_ticket_enable_server()) but client side is enabled
46 by default on recent versions (3.1.3 +) unless the PFS priority string
49 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
50 <https://reproducible-builds.org/specs/source-date-epoch/>.
52 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
53 the check for any unsuccessful recipients did not notice the limit, and
54 erroneously found still-pending ones.
56 JH/08 Pipeline CHUNKING command and data together, on kernels that support
57 MSG_MORE. Only in-clear (not on TLS connections).
59 JH/09 Avoid using a temporary file during transport using dkim. Unless a
60 transport-filter is involved we can buffer the headers in memory for
61 creating the signature, and read the spool data file once for the
62 signature and again for transmission.
64 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
65 4.77 as the kernel support then wasn't solid, having issues in 64bit
66 mode. Now, it's been long enough. Add support for FreeBSD also.
68 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
69 case where the routing stage had gathered several addresses to send to
70 a host before calling the transport for the first, we previously failed
71 to close down TLS in the old transport process before passing the TCP
72 connection to the new process. The new one sent a STARTTLS command
73 which naturally failed, giving a failed delivery and bloating the retry
74 database. Investigation and fix prototype from Wolfgang Breyha.
76 JH/12 Fix check on SMTP command input synchronisation. Previously there were
77 false-negatives in the check that the sender had not preempted a response
78 or prompt from Exim (running as a server), due to that code's lack of
79 awareness of the SMTP input buffering.
81 PP/04 Add commandline_checks_require_admin option.
82 Exim drops privileges sanely, various checks such as -be aren't a
83 security problem, as long as you trust local users with access to their
84 own account. When invoked by services which pass untrusted data to
85 Exim, this might be an issue. Set this option in main configuration
86 AND make fixes to the calling application, such as using `--` to stop
89 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
90 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
91 responses to those, into a single TLS record each way (this usually means
92 a single packet). As a side issue, smtp_enforce_sync now works on TLS
95 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
96 affects you only if you're dancing at the edge of the param size limits.
97 If you are, and this message makes sense to you, then: raise the
98 configured limit or use OpenSSL 1.1. Nothing we can do for older
101 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
102 cmdline element to get the documented default one. Previously it was
109 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
110 than -2003 did; needs libidn2 in addition to libidn.
112 JH/02 The path option on a pipe transport is now expanded before use.
114 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
115 Patch provided by "Björn", documentation fix added too.
117 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
118 missing a wire-to-host endian conversion.
120 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
121 close after a BDAT command line could be taken as a following command,
122 giving a synch failure. Fix by only checking for synch immediately
123 before acknowledging the chunk.
125 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
126 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
127 macro. Patches provided by Josh Soref.
129 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
130 Previously we did not; the RFC seems ambiguous and VRFY is not listed
131 by IANA as a service extension. However, John Klensin suggests that we
134 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
135 the dkim code may be unix-mode line endings rather than smtp wire-format
136 CRLF, so prepend a CR to any bare LF.
138 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
139 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
141 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
142 the first were themselves being wrongly included in the feed into dkim
143 processing; with most chunk sizes in use this resulted in an incorrect
144 body hash calculated value.
146 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
147 DKIM signature block, for verification. Although advised against by
148 standards it is specifically not ruled illegal.
150 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
152 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
153 missing a body hash (the bh= tag).
155 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
156 It seems that HAProxy sends the Proxy Protocol information in clear and
157 only then does a TLS startup, so do the same.
159 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
160 TCP connections (such as for Spamd) unless the daemon successfully set
161 Fast Open mode on its listening sockets. This fixes breakage seen on
162 too-old kernels or those not configured for Fast Open, at the cost of
163 requiring both directions being enabled for TFO, and TFO never being used
164 by non-daemon-related Exim processes.
166 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
167 endings, at least on the first header line. Try to canonify any that get
168 past that check, despite the cost.
170 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
171 now limited to an arbitrary five deep, while parsing addresses with the
172 strip_excess_angle_brackets option enabled.
174 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
175 instead leave the unprompted TLS handshake in socket buffer for the
176 TLS library to consume.
178 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
180 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
182 JH/16 Drop variables when they go out of scope. Memory management drops a whole
183 region in one operation, for speed, and this leaves assigned pointers
184 dangling. Add checks run only under the testsuite which checks all
185 variables at a store-reset and panics on a dangling pointer; add code
186 explicitly nulling out all the variables discovered. Fixes one known
187 bug: a transport crash, where a dangling pointer for $sending_ip_address
188 originally assigned in a verify callout, is re-used.
190 PP/06 Drop '.' from @INC in various Perl scripts.
192 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
194 PP/08 Reduce a number of compilation warnings under clang; building with
195 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
196 should be warning-free.
198 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
200 HS/01 Fix portability problems introduced by PP/08 for platforms where
201 realloc(NULL) is not equivalent to malloc() [SunOS et al].
203 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
204 chunk. This allows us to accept broken chunked messages. We need a more
205 general solution here.
207 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
208 already-broken messages in the queue.
210 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
212 JH/19 Fix reference counting bug in routing-generated-address tracking.
218 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
219 supports it and a size is available (ie. the sending peer gave us one).
221 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
222 years of being deprecated). The replacements are the ACLs
223 acl_smtp_mime and acl_not_smtp_mime.
225 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
226 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
227 or even an in-clear connection were permitted. Now, if the host lookup
228 was dnssec and dane was requested then the host is only used if the
229 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
230 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
231 if one fails this test.
232 This means that a poorly-configured remote DNS will make it incommunicado;
233 but it protects against a DNS-interception attack on it.
235 JH/04 Bug 1810: make continued-use of an open smtp transport connection
236 non-noisy when a race steals the message being considered.
238 JH/05 If main configuration option tls_certificate is unset, generate a
239 self-signed certificate for inbound TLS connections.
241 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
242 in rewrites and routers.
244 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
245 and logged a warning sing 4.83; now they are a configuration file error.
247 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
248 (lacking @domain). Apply the same qualification processing as RCPT.
250 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
252 JH/10 Support ${sha256:} applied to a string (as well as the previous
255 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
256 a cutthrough deliver is pending, as we always want to make a connection.
257 This also avoids re-routing the message when later placing the cutthrough
258 connection after a verify cache hit.
259 Do not update it with the verify result either.
261 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
262 when routing results in more than one destination address.
264 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
265 signing (which inhibits the cutthrough capability). Previously only
266 the presence of an option was tested; now an expansion evaluating as
267 empty is permissible (obviously it should depend only on data available
268 when the cutthrough connection is made).
270 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
271 the relevant preceding SMTP command did not note the pipelining mode.
273 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
274 Previously they were not counted.
276 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
277 as one having no matching records. Previously we deferred the message
278 that needed the lookup.
280 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
281 distinguished as "(=".
283 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
284 for missing MX records. Previously it only worked for missing A records.
286 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
288 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
289 after the data-go-ahead and data-ack. Patch from Jason Betts.
291 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
292 even for a "none" policy. Patch from Tony Meyer.
294 JH/22 Fix continued use of a connection for further deliveries. If a port was
295 specified by a router, it must also match for the delivery to be
298 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
299 When one of the recipients of a message was unsuitable for the connection
300 (has no matching addresses), we lost track of needing to mark it
301 deferred. As a result mail would be lost.
303 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
305 JH/25 Decoding ACL controls is now done using a binary search; the source code
306 takes up less space and should be simpler to maintain. Merge the ACL
307 condition decode tables also, with similar effect.
309 JH/26 Fix problem with one_time used on a redirect router which returned the
310 parent address unchanged. A retry would see the parent address marked as
311 delivered, so not attempt the (identical) child. As a result mail would
314 JH/27 Fix a possible security hole, wherein a process operating with the Exim
315 UID can gain a root shell. Credit to http://www.halfdog.net/ for
316 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
319 JH/28 Enable {spool,log} filesystem space and inode checks as default.
320 Main config options check_{log,spool}_{inodes,space} are now
321 100 inodes, 10MB unless set otherwise in the configuration.
323 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
324 Previously it only applied to the main-section connection policy
327 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
329 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
330 by me. Added RFC7919 DH primes as an alternative.
332 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
333 are not in the system include path.
335 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
336 GnuTLS, when a session startup failed (eg because the client disconnected)
337 Exim did stdio operations after fclose. This was exposed by a recent
338 change which nulled out the file handle after the fclose.
340 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
341 signed directly by the cert-signing cert, rather than an intermediate
342 OCSP-signing cert. This is the model used by LetsEncrypt.
344 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
346 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
347 an incoming connection.
349 HS/02 Bug 1802: Do not half-close the connection after sending a request
352 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
353 fallback to "prime256v1".
355 JH/34 SECURITY: Use proper copy of DATA command in error message.
356 Could leak key material. Remotely exploitable. CVE-2016-9963.
362 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
363 and 3.4.4 - once the server is enabled to respond to an OCSP request
364 it does even when not requested, resulting in a stapling non-aware
365 client dropping the TLS connection.
367 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
368 support variable-length bit vectors. No functional change.
370 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
371 The I= interface field on outgoing lines is now after the H= remote
372 host field, same as incoming lines. There is a separate
373 outgoing_interface log selector which allows you to disable the
376 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
377 If not running log_selector +smtp_connection the mainlog would be held
378 open indefinitely after a "too many connections" event, including to a
379 deleted file after a log rotate. Leave the per net connection logging
380 leaving it open for efficiency as that will be quickly detected by the
381 check on the next write.
383 HS/01 Bug 1671: Fix post transport crash.
384 Processing the wait-<transport> messages could crash the delivery
385 process if the message IDs didn't exist for some reason. When
386 using 'split_spool_directory=yes' the construction of the spool
387 file name failed already, exposing the same netto behaviour.
389 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
390 mime_regex ACL conditions.
392 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
393 to DSN fail messages (bounces): remote IP, remote greeting, remote response
394 to HELO, local diagnostic string.
396 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
397 log line to debug. Even when configured with a tls authenticator many
398 client connections are expected to not authenticate in this way, so
399 an authenticate fail is not an error.
401 HS/02 Add the Exim version string to the process info. This way exiwhat
402 gives some more detail about the running daemon.
404 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
405 matter for fast-change records such as DNSBLs.
407 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
408 retry record, even if constant. There may be multiple transports with
409 different interface settings and the retry behaviour needs to be kept
412 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
414 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
416 JH/10 Bug 840: fix log_defer_output option of pipe transport
418 JH/11 Bug 830: use same host for all RCPTS of a message, even under
419 hosts_randomize. This matters a lot when combined with mua_wrapper.
421 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
422 ${quote_pgsql:<string>} operator.
424 JH/13 Bug 1708: avoid misaligned access in cached lookup.
426 JH/14 Change header file name for freeradius-client. Relevant if compiling
427 with Radius support; from the Gentoo tree and checked under Fedora.
429 JH/15 Bug 1712: Introduce $prdr_requested flag variable
431 JH/16 Bug 1714: Permit an empty string as expansion result for transport
432 option transport_filter, meaning no filtering.
434 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
436 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
437 defaults to "*" (all hosts). The variable is now available when not built
438 with TLS, default unset, mainly to enable keeping the testsuite sane.
439 If a server certificate is not supplied (via tls_certificate) an error is
440 logged, and clients will find TLS connections fail on startup. Presumably
441 they will retry in-clear.
442 Packagers of Exim are strongly encouraged to create a server certificate
443 at installation time.
445 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
446 with the $config_file variable.
448 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
449 in transport context, after the attempt, and per-recipient. The latter type
450 is per host attempted. The event data is the error message, and the errno
451 information encodes the lookup type (A vs. MX) used for the (first) host,
452 and the trailing two digits of the smtp 4xx response.
454 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
455 to write to mainlog (or rejectlog, paniclog) in the window between file
456 creation and permissions/ownership being changed. Particularly affects
457 installations where exicyclog is run as root, rather than exim user;
458 result is that the running daemon panics and dies.
460 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
462 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
463 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
464 "pri" and "weight". Note that the previous implicit priority given by the
465 list order is no longer honoured.
467 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
470 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
471 by defining SUPPORT_SOCKS.
473 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
474 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
475 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
476 variables are renamed to proxy_{local,external}_{address,port}.
478 JH/27 Move Internationalisation support from Experimental to mainline, enabled
479 for a build by defining SUPPORT_I18N
481 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
482 of the query string, and make ${quote_redis:} do that quoting.
484 JH/29 Move Events support from Experimental to mainline, enabled by default
485 and removable for a build by defining DISABLE_EVENT.
487 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
489 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
490 cached by the daemon.
492 JH/32 Move Redis support from Experimental to mainline, enabled for a build
493 by defining LOOKUP_REDIS. The libhiredis library is required.
495 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
496 keys are given for lookup.
498 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
499 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
500 only supported when built with TLS support. The PolarSSL SHA routines
501 are still used when the TLS library is too old for convenient support.
503 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
504 openssl_options), for security. OpenSSL forces this from version 1.1.0
505 server-side so match that on older versions.
507 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
508 allocation for $value could be released as the expansion processing
509 concluded, but leaving the global pointer active for it.
511 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
512 and to use the domains and local_parts ACL conditions.
514 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
515 incorrectly not doubled on cutthrough transmission, hence seen as a
516 body-termination at the receiving system - resulting in truncated mails.
517 Commonly the sender saw a TCP-level error, and retransmitted the message
518 via the normal store-and-forward channel. This could result in duplicates
519 received - but deduplicating mailstores were liable to retain only the
520 initial truncated version.
522 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
524 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
526 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
527 we're in there, support oversigning also; bug 1309.
529 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
531 HS/04 Add support for keep_environment and add_environment options.
533 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
534 either intentional arithmetic overflow during PRNG, or testing config-
537 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
538 delivery resulted in actual delivery. Cancel cutthrough before DATA
541 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
542 rejects a recipient: pass the reject to the originator.
544 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
545 Many were false-positives and ignorable, but it's worth fixing the
548 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
549 for the new environment-manipulation done at startup. Move the routines
550 from being local to tls.c to being global via the os.c file.
552 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
553 an extract embedded as result-arg for a map, the first arg for extract
554 is unavailable so we cannot tell if this is a numbered or keyed
555 extraction. Accept either.
561 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
564 JH/02 The smtp transport option "multi_domain" is now expanded.
566 JH/03 The smtp transport now requests PRDR by default, if the server offers
569 JH/04 Certificate name checking on server certificates, when exim is a client,
570 is now done by default. The transport option tls_verify_cert_hostnames
571 can be used to disable this per-host. The build option
572 EXPERIMENTAL_CERTNAMES is withdrawn.
574 JH/05 The value of the tls_verify_certificates smtp transport and main options
575 default to the word "system" to access the system default CA bundle.
576 For GnuTLS, only version 3.0.20 or later.
578 JH/06 Verification of the server certificate for a TLS connection is now tried
579 (but not required) by default. The verification status is now logged by
580 default, for both outbound TLS and client-certificate supplying inbound
583 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
586 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
587 Status Notification (bounce) messages are now MIME format per RFC 3464.
588 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
589 under the control of the dsn_advertise_hosts option, and routers may
590 have a dsn_lasthop option.
592 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
593 default, modifiable by a malware= option. The list separator for
594 the options can now be changed in the usual way. Bug 68.
596 JH/10 The smtp_receive_timeout main option is now expanded before use.
598 JH/11 The incoming_interface log option now also enables logging of the
599 local interface on delivery outgoing connections.
601 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
602 if the interface and destination host and port all match.
604 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
607 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
608 Patch from Andrew Lewis.
610 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
611 now supports optional time-restrictions, weighting, and priority
612 modifiers per server. Patch originally by <rommer@active.by>.
614 JH/16 The spamd_address main option now supports a mixed list of local
615 and remote servers. Remote servers can be IPv6 addresses, and
616 specify a port-range.
618 JH/17 Bug 68: The spamd_address main option now supports an optional
619 timeout value per server.
621 JH/18 Bug 1581: Router and transport options headers_add/remove can
622 now have the list separator specified.
624 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
627 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
630 JH/21 Support for the A6 type of dns record is withdrawn.
632 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
633 rather than the verbs used.
635 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
636 from 255 to 1024 chars.
638 JH/24 Verification callouts now attempt to use TLS by default.
640 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
641 are generic router options now. The defaults didn't change.
643 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
644 Original patch from Alexander Shikoff, worked over by JH.
646 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
647 files if ZCAT_COMMAND is not executable.
649 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
651 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
653 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
654 Normally benign, it bites when the pair was led to by a CNAME;
655 modern usage is to not canonicalize the domain to a CNAME target
656 (and we were inconsistent anyway for A-only vs AAAA+A).
658 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
660 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
661 when evaluating $sender_host_dnssec.
663 JH/31 Check the HELO verification lookup for DNSSEC, adding new
664 $sender_helo_dnssec variable.
666 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
668 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
670 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
672 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
673 documented as working, but never had. Support all but $spam_report.
675 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
676 added for tls authenticator.
678 HS/03 Add perl_taintmode main config option
684 TL/01 When running the test suite, the README says that variables such as
685 no_msglog_check are global and can be placed anywhere in a specific
686 test's script, however it was observed that placement needed to be near
687 the beginning for it to behave that way. Changed the runtest perl
688 script to read through the entire script once to detect and set these
689 variables, reset to the beginning of the script, and then run through
690 the script parsing/test process like normal.
692 TL/02 The BSD's have an arc4random API. One of the functions to induce
693 adding randomness was arc4random_stir(), but it has been removed in
694 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
695 function when detected.
697 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
698 cause callback expansion.
700 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
701 syntax errors in an expansion can be treated as a string instead of
702 logging or causing an error, due to the internal use of bool_lax
703 instead of bool when processing it.
705 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
706 server certificates when making smtp deliveries.
708 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
710 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
712 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
714 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
715 Merged patch from Sebastian Wiedenroth.
717 JH/05 Fix results-pipe from transport process. Several recipients, combined
718 with certificate use, exposed issues where response data items split
719 over buffer boundaries were not parsed properly. This eventually
720 resulted in duplicates being sent. This issue only became common enough
721 to notice due to the introduction of connection certificate information,
722 the item size being so much larger. Found and fixed by Wolfgang Breyha.
724 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
725 size buffer was used, resulting in syntax errors when an expansion
728 JH/07 Add support for directories of certificates when compiled with a GnuTLS
729 version 3.3.6 or later.
731 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
732 is EXPERIMENTAL_EVENT, the main-configuration and transport options
733 both become "event_action", the variables become $event_name, $event_data
734 and $event_defer_errno. There is a new variable $verify_mode, usable in
735 routers, transports and related events. The tls:cert event is now also
736 raised for inbound connections, if the main configuration event_action
739 TL/06 In test suite, disable OCSP for old versions of openssl which contained
740 early OCSP support, but no stapling (appears to be less than 1.0.0).
742 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
743 server certificate names available under the smtp transport option
744 "tls_verify_cert_hostname" now do not permit multi-component wildcard
747 JH/10 Time-related extraction expansions from certificates now use the main
748 option "timezone" setting for output formatting, and are consistent
749 between OpenSSL and GnuTLS compilations. Bug 1541.
751 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
752 encoded parameter in the incoming message. Bug 1558.
754 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
755 include certificate info, eximon was claiming there were spoolfile
758 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
760 JH/14 Log delivery-related information more consistently, using the sequence
761 "H=<name> [<ip>]" wherever possible.
763 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
764 are problematic for Debian distribution, omit them from the release
767 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
769 JH/16 Fix string representation of time values on 64bit time_t architectures.
772 JH/17 Fix a null-indirection in certextract expansions when a nondefault
773 output list separator was used.
778 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
779 checkers that were complaining about end of non-void function with no
782 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
783 This was a regression introduced in 4.83 by another bugfix.
785 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
787 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
788 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
794 TF/01 Correctly close the server side of TLS when forking for delivery.
796 When a message was received over SMTP with TLS, Exim failed to clear up
797 the incoming connection properly after forking off the child process to
798 deliver the message. In some situations the subsequent outgoing
799 delivery connection happened to have the same fd number as the incoming
800 connection previously had. Exim would try to use TLS and fail, logging
801 a "Bad file descriptor" error.
803 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
804 utilities have not been installed.
806 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
807 temporary space as the ACL may create new global variables.
809 TL/01 LDAP support uses per connection or global context settings, depending
810 upon the detected version of the libraries at build time.
812 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
813 to extract and use the src ip:port in logging and expansions as if it
814 were a direct connection from the outside internet. PPv2 support was
815 updated based on HAProxy spec change in May 2014.
817 JH/02 Add ${listextract {number}{list}{success}{fail}}.
819 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
820 Properly escape header and check for NULL return.
822 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
825 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
827 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
828 characters in header names, implemented as a verify condition.
829 Contributed by Michael Fischer v. Mollard.
831 TL/05 Rename SPF condition results err_perm and err_temp to standardized
832 results permerror and temperror. Previous values are deprecated but
833 still accepted. In a future release, err_perm and err_temp will be
834 completely removed, which will be a backward incompatibility if the
835 ACL tests for either of these two old results. Patch contributed by
836 user bes-internal on the mailing list.
838 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
840 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
841 selectors, in both main and reject logs.
843 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
846 JH/07 Add malware type "sock" for talking to simple daemon.
848 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
850 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
851 routers/transports under cutthrough routing.
853 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
854 numbers. Touch up "bool" conditional to keep the same definition.
856 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
858 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
860 JH/12 Expand items in router/transport headers_add or headers_remove lists
861 individually rather than the list as a whole. Bug 1452.
863 Required for reasonable handling of multiple headers_ options when
864 they may be empty; requires that headers_remove items with embedded
865 colons must have them doubled (or the list-separator changed).
867 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
868 view the policy declared in the DMARC record. Currently, $dmarc_status
869 is a combined value of both the record presence and the result of the
872 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
874 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
875 dnslookup router and the smtp transport (applying to the forward
878 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
879 of ldap servers used for a specific lookup. Patch provided by Heiko
882 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
883 New variable $lookup_dnssec_authenticated for observability.
885 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
886 Patch submitted by Lars Timman.
888 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
890 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
891 Requires trusted mode and valid format message id, aborts otherwise.
892 Patch contributed by Heiko Schlichting.
894 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
895 certextract with support for various fields. Bug 1358.
897 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
898 is requested by default, modifiable by smtp transport option
901 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
902 operate on certificate variables to give certificate fingerprints
903 Also new ${sha256:cert_variable}.
905 JH/23 The PRDR feature is moved from being Experimental into the mainline.
907 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
908 Christian Aistleitner.
910 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
912 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
913 file. Patch from Wolfgang Breyha.
915 JH/25 Expand the coverage of the delivery $host and $host_address to
916 client authenticators run in verify callout. Bug 1476.
918 JH/26 Port service names are now accepted for tls_on_connect_ports, to
919 align with daemon_smtp_ports. Bug 72.
921 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
922 support and error reporting did not work properly.
924 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
925 and is readable. Patch from Andrew Colin Kissa.
927 TL/14 Enhance documentation of ${run expansion and how it parses the
928 commandline after expansion, particularly in the case when an
929 unquoted variable expansion results in an empty value.
931 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
933 PP/02 Fix internal collision of T_APL on systems which support RFC3123
934 by renaming away from it. Addresses GH issue 15, reported by
937 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
939 TL/15 SECURITY: prevent double expansion in math comparison functions
940 (can expand unsanitized data). Not remotely exploitable.
947 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
949 PP/02 Make -n do something, by making it not do something.
950 When combined with -bP, the name of an option is not output.
952 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
955 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
956 $sender_host_name and config options to manage this, and basic check
959 PP/05 DSCP support for outbound connections and control modifier for inbound.
961 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
962 (Only plugin which currently uses this is kerberos4, which nobody should
963 be using, but we should make it available and other future plugins might
964 conceivably use it, even though it would break NAT; stuff *should* be
965 using channel bindings instead).
967 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
968 name; added for Sendmail compatibility; requires admin caller.
969 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
970 just ignore it); requires trusted caller.
971 Also parse but ignore: -Ac -Am -X<logfile>
974 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
976 TL/02 Add +smtp_confirmation as a default logging option.
978 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
979 Patch by Magnus Holmgren from 2007-02-20.
981 TL/04 Bugzilla 1281 - Spec typo.
982 Bugzilla 1283 - Spec typo.
983 Bugzilla 1290 - Spec grammar fixes.
985 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
987 TL/06 Add Experimental DMARC support using libopendmarc libraries.
989 TL/07 Fix an out of order global option causing a segfault. Reported to dev
990 mailing list by by Dmitry Isaikin.
992 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
994 JH/02 Support "G" suffix to numbers in ${if comparisons.
996 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
998 NM/01 Bugzilla 1197 - Spec typo
999 Bugzilla 1196 - Spec examples corrections
1001 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1003 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1004 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1007 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1008 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1010 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1011 "acl {{name}{arg}...}", and optional args on acl condition
1014 JH/05 Permit multiple router/transport headers_add/remove lines.
1016 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1018 JH/07 Avoid using a waiting database for a single-message-only transport.
1019 Performance patch from Paul Fisher. Bugzilla 1262.
1021 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1024 JH/09 Add $headers_added variable, with content from use of ACL modifier
1025 add_header (but not yet added to the message). Bugzilla 199.
1027 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1028 Pulled from Bugzilla 817 by Wolfgang Breyha.
1030 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1032 (nb: this is the same fix as in Exim 4.80.1)
1034 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1037 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1040 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1042 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1043 advertise SMTP AUTH mechanism to us, instead of a generic
1044 protocol violation error. Also, make Exim more robust to bad
1045 data from the Dovecot auth socket.
1047 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1049 When a queue runner is handling a message, Exim first routes the
1050 recipient addresses, during which it prunes them based on the retry
1051 hints database. After that it attempts to deliver the message to
1052 any remaining recipients. It then updates the hints database using
1055 So if a recipient address works intermittently, it can get repeatedly
1056 deferred at routing time. The retry hints record remains fresh so the
1057 address never reaches the final cutoff time.
1059 This is a fairly common occurrence when a user is bumping up against
1060 their storage quota. Exim had some logic in its local delivery code
1061 to deal with this. However it did not apply to per-recipient defers
1062 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1064 This change adds a proper retry rule check during routing so that the
1065 final cutoff time is checked against the message's age. We only do
1066 this check if there is an address retry record and there is not a
1067 domain retry record; this implies that previous attempts to handle
1068 the address had the retry_use_local_parts option turned on. We use
1069 this as an approximation for the destination being like a local
1070 delivery, as in LMTP.
1072 I suspect this new check makes the old local delivery cutoff check
1073 redundant, but I have not verified this so I left the code in place.
1075 TF/02 Correct gecos expansion when From: is a prefix of the username.
1077 Test 0254 submits a message to Exim with the header
1081 When I ran the test suite under the user fanf2, Exim expanded
1082 the header to contain my full name, whereas it should have added
1083 a Resent-Sender: header. It erroneously treats any prefix of the
1084 username as equal to the username.
1086 This change corrects that bug.
1088 GF/01 DCC debug and logging tidyup
1089 Error conditions log to paniclog rather than rejectlog.
1090 Debug lines prefixed by "DCC: " to remove any ambiguity.
1092 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1094 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1095 Bug spotted by Jeremy Harris; was flawed since initial commit.
1096 Would have resulted in OCSP responses post-SNI triggering an Exim
1097 NULL dereference and crash.
1099 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1101 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1102 Bug detection, analysis and fix by Samuel Thibault.
1103 Bugzilla 1331, Debian bug #698092.
1105 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1107 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1108 Server implementation by Todd Lyons, client by JH.
1109 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1110 config variable "prdr_enable" controls whether the server
1111 advertises the facility. If the client requests PRDR a new
1112 acl_data_smtp_prdr ACL is called once for each recipient, after
1113 the body content is received and before the acl_smtp_data ACL.
1114 The client is controlled by both of: a hosts_try_prdr option
1115 on the smtp transport, and the server advertisement.
1116 Default client logging of deliveries and rejections involving
1117 PRDR are flagged with the string "PRDR".
1119 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1120 fclose(). Diagnosis by Todd Lyons.
1122 PP/17 Update configure.default to handle IPv6 localhost better.
1123 Patch by Alain Williams (plus minor tweaks).
1126 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1127 This is now consistent with GnuTLS, and is now documented: the
1128 previous undocumented portable approach to treating the option as
1129 unset was to force an expansion failure. That still works, and
1130 an empty string is now equivalent.
1132 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1133 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1134 not performing validation itself.
1136 PP/20 Added force_command boolean option to pipe transport.
1137 Patch from Nick Koston, of cPanel Inc.
1139 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1142 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1144 PP/21 Fix eximon continuous updating with timestamped log-files.
1145 Broken in a format-string cleanup in 4.80, missed when I repaired the
1146 other false fix of the same issue.
1147 Report and fix from Heiko Schlichting.
1150 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1151 Report from Prashanth Katuri.
1153 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1154 It's SecureTransport, so affects any MacOS clients which use the
1155 system-integrated TLS libraries, including email clients.
1157 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1158 using a MIME ACL for non-SMTP local injection.
1159 Report and assistance in diagnosis by Warren Baker.
1161 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1163 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1165 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1166 last id that failed so it may be referenced in subsequent ACL's.
1168 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1171 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1172 ldap library initialization, allowing self-signed CA's to be
1173 used. Also properly sets require_cert option later in code by
1174 using NULL (global ldap config) instead of ldap handle (per
1175 session). Bug diagnosis and testing by alxgomz.
1177 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1178 the src/util/ subdirectory.
1180 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1181 renamed to Transport Post Delivery Action by Jeremy Harris, as
1184 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1185 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1186 redis_servers = needs to be configured which will be used by the redis
1187 lookup. Patch from Warren Baker, of The Packet Hub.
1189 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1191 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1192 hostname or reverse DNS when processing a host list. Used suggestions
1193 from multiple comments on this bug.
1195 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1197 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1198 Missed a few lines, added it to make the runtest require no keyboard
1201 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1202 contains upper case chars. Make router use caseful_local_part.
1204 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1205 support when GnuTLS has been built with p11-kit.
1211 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1213 This, or similar/improved, will also be change PP/11 of 4.82.
1219 PP/01 Handle short writes when writing local log-files.
1220 In practice, only affects FreeBSD (8 onwards).
1221 Bugzilla 1053, with thanks to Dmitry Isaikin.
1223 NM/01 Bugzilla 949 - Documentation tweak
1225 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1228 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1230 PP/02 Implemented gsasl authenticator.
1232 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1234 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1235 `pkg-config foo` for cflags/libs.
1237 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1238 with rest of GSASL and with heimdal_gssapi.
1240 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1241 `pkg-config foo` for cflags/libs for the TLS implementation.
1243 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1244 properties get this fed in as external SSF. A number of robustness
1245 and debugging improvements to the cyrus_sasl authenticator.
1247 PP/08 cyrus_sasl server now expands the server_realm option.
1249 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1250 Patch by Jeremy Harris.
1252 PP/10 Added dbmjz lookup type.
1254 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1256 PP/12 MAIL args handles TAB as well as SP, for better interop with
1257 non-compliant senders.
1258 Analysis and variant patch by Todd Lyons.
1260 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1261 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1262 Patch from Dirk Mueller <dmueller@suse.com>
1264 PP/13 tls_peerdn now print-escaped for spool files.
1265 Observed some $tls_peerdn in wild which contained \n, which resulted
1266 in spool file corruption.
1268 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1269 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1270 or write after TLS renegotiation, which otherwise led to messages
1273 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1274 as a tracking header (ie: a signed header comes before the signature).
1275 Patch from Wolfgang Breyha.
1277 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1278 comma-sep list; embedded commas doubled.
1280 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1282 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1284 Report and patch from Dmitry Banschikov.
1286 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1287 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1288 failures. We appear to now support TLS1.1+ with Exim.
1290 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1291 lets Exim select keys and certificates based upon TLS SNI from client.
1292 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1293 before an outbound SMTP session. New log_selector, +tls_sni.
1295 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1296 NULL dereference. Report and patch from Alun Jones.
1298 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1299 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1300 Not seeing resolver debug output on NetBSD, but suspect this is a
1301 resolver implementation change.
1303 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1304 Left warnings. Added "eximon gdb" invocation mode.
1306 PP/21 Defaulting "accept_8bitmime" to true, not false.
1308 PP/22 Added -bw for inetd wait mode support.
1310 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1311 locate the relevant includes and libraries. Made this the default.
1313 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1314 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1316 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1317 This may cause build issues on older platforms.
1319 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1320 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1321 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1322 Added SNI support via GnuTLS too.
1323 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1325 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1327 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1328 Applied second patch from Janne, implementing suggestion to default
1329 multiple-strings-in-record handling to match SPF spec.
1331 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1333 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1334 read-only, out of scope).
1335 Patch from Wolfgang Breyha, report from Stuart Northfield.
1337 PP/29 Fix three issues highlighted by clang analyser static analysis.
1338 Only crash-plausible issue would require the Cambridge-specific
1339 iplookup router and a misconfiguration.
1340 Report from Marcin Mirosław.
1342 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1344 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1345 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1346 As part of this, removing so much warning spew let me fix some minor
1347 real issues in debug logging.
1349 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1350 assignment on my part. Fixed.
1352 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1353 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1354 Janne Snabb (who went above and beyond: thank you).
1356 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1357 string otherwise requires a connection and a bunch more work and it's
1358 relatively easy to get wrong. Should also expose TLS library linkage
1361 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1362 64-bit ${eval} (JH/03).
1364 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1365 GNU libc to support some of the 64-bit stuff, should not lead to
1366 conflicts. Defined before os.h is pulled in, so if a given platform
1367 needs to override this, it can.
1369 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1370 protection layer was required, which is not implemented.
1371 Bugzilla 1254, patch from Wolfgang Breyha.
1373 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1374 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1375 tls_dhparam take prime identifiers. Also unbreak combination of
1376 OpenSSL+DH_params+TLSSNI.
1378 PP/39 Disable SSLv2 by default in OpenSSL support.
1384 PP/01 Solaris build fix for Oracle's LDAP libraries.
1385 Bugzilla 1109, patch from Stephen Usher.
1387 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1389 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1392 TF/02 Fix a couple more cases where we did not log the error message
1393 when unlink() failed. See also change 4.74-TF/03.
1395 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1396 lock up or crash if it happened to be inside a call to libc when it
1397 got a SIGUSR1 from exiwhat.
1399 The SIGUSR1 handler appends the current process status to the process
1400 log which is later printed by exiwhat. It used to use the general
1401 purpose logging code to do this, but several functions it calls are
1402 not safe for signals.
1404 The new output code in the SIGUSR1 handler is specific to the process
1405 log, and simple enough that it's easy to inspect for signal safety.
1406 Removing some special cases also simplifies the general logging code.
1407 Removing the spurious timestamps from the process log simplifies
1410 TF/04 Improved ratelimit ACL condition.
1412 The /noupdate option has been deprecated in favour of /readonly which
1413 has clearer semantics. The /leaky, /strict, and /readonly update modes
1414 are mutually exclusive. The update mode is no longer included in the
1415 database key; it just determines when the database is updated. (This
1416 means that when you upgrade Exim will forget old rate measurements.)
1418 Exim now checks that the per_* options are used with an update mode that
1419 makes sense for the current ACL. For example, when Exim is processing a
1420 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1421 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1422 must specify per_mail/readonly. If you omit the update mode it defaults to
1423 /leaky where that makes sense (as before) or /readonly where required.
1425 The /noupdate option is now undocumented but still supported for
1426 backwards compatibility. It is equivalent to /readonly except that in
1427 ACLs where /readonly is required you may specify /leaky/noupdate or
1428 /strict/noupdate which are treated the same as /readonly.
1430 A useful new feature is the /count= option. This is a generalization
1431 of the per_byte option, so that you can measure the throughput of other
1432 aggregate values. For example, the per_byte option is now equivalent
1433 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1435 The per_rcpt option has been generalized using the /count= mechanism
1436 (though it's more complicated than the per_byte equivalence). When it is
1437 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1438 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1439 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1440 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1441 Note that using per_rcpt with a non-readonly update mode in more than
1442 one ACL will cause the recipients to be double-counted. (The per_mail
1443 and per_byte options don't have this problem.)
1445 The handling of very low rates has changed slightly. If the computed rate
1446 is less than the event's count (usually one) then this event is the first
1447 after a long gap. In this case the rate is set to the same as this event's
1448 count, so that the first message of a spam run is counted properly.
1450 The major new feature is a mechanism for counting the rate of unique
1451 events. The new per_addr option counts the number of different
1452 recipients that someone has sent messages to in the last time period. It
1453 behaves like per_rcpt if all the recipient addresses are different, but
1454 duplicate recipient addresses do not increase the measured rate. Like
1455 the /count= option this is a general mechanism, so the per_addr option
1456 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1457 example, measure the rate that a client uses different sender addresses
1458 with the options per_mail/unique=$sender_address. There are further
1459 details in the main documentation.
1461 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1463 TF/06 Removed a few PCRE remnants.
1465 TF/07 Automatically extract Exim's version number from tags in the git
1466 repository when doing development or release builds.
1468 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1469 Bugzilla 879. Patch from Paul Fisher.
1471 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1472 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1475 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1477 PP/05 Variable $av_failed, true if the AV scanner deferred.
1478 Bugzilla 1078. Patch from John Horne.
1480 PP/06 Stop make process more reliably on build failure.
1481 Bugzilla 1087. Patch from Heiko Schlittermann.
1483 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1484 Bugzilla 1089. Patch from Heiko Schlittermann.
1486 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1487 Bugzilla 1131. Patch from Holger Weiß.
1489 PP/09 Handle IPv6 addresses with SPF.
1490 Bugzilla 860. Patch from Wolfgang Breyha.
1492 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1494 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1497 PP/11 match_* no longer expand right-hand-side by default.
1498 New compile-time build option, EXPAND_LISTMATCH_RHS.
1499 New expansion conditions, "inlist", "inlisti".
1501 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1503 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1505 PP/14 fix log_write() format string regression from TF/03.
1506 Bugzilla 1152. Patch from Dmitry Isaikin.
1512 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1514 PP/02 Harmonised TLS library version reporting; only show if debugging.
1515 Layout now matches that introduced for other libraries in 4.74 PP/03.
1517 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1519 PP/04 New "dns_use_edns0" global option.
1521 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1524 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1525 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1527 TK/01 Updated PolarSSL code to 0.14.2.
1528 Bugzilla 1097. Patch from Andreas Metzler.
1530 PP/07 Catch divide-by-zero in ${eval:...}.
1531 Fixes bugzilla 1102.
1533 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1536 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1537 format-string attack -- SECURITY: remote arbitrary code execution.
1539 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1540 time unintentionally subject to list matching rules, letting the header
1541 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1542 arbitrary string expansion). This allowed for information disclosure.
1544 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1545 INT_MIN/-1 -- value coerced to INT_MAX.
1551 NM/01 Workaround for PCRE version dependency in version reporting
1554 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1555 This fixes portability to compilers other than gcc, notably
1556 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1558 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1559 makefiles for portability to HP-UX and POSIX correctness.
1561 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1562 Also via indented variable definition in the Makefile.
1563 (Debugging by Oliver Heesakkers).
1565 PP/02 Restore caching of spamd results with expanded spamd_address.
1566 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1568 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1569 Improves build reliability. Fix from: Frank Elsner
1571 NM/02 Fix wide character breakage in the rfc2047 coding
1572 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1574 NM/03 Allow underscore in dnslist lookups
1575 Fixes bug 1026. Patch from Graeme Fowler
1577 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1578 Code patches from Adam Ciarcinski of NetBSD.
1580 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1583 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1584 is logged, to avoid truncation. Patch from John Horne.
1586 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1587 Patch from Jakob Hirsch.
1589 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1590 SQL string expansion failure details.
1591 Patch from Andrey Oktyabrski.
1593 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1594 Patch from Simon Arlott.
1596 PP/09 New lookups functionality failed to compile on old gcc which rejects
1597 extern declarations in function scope.
1598 Patch from Oliver Fleischmann
1600 PP/10 Use sig_atomic_t for flags set from signal handlers.
1601 Check getgroups() return and improve debugging.
1602 Fixed developed for diagnosis in bug 927 (which turned out to be
1605 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1606 Patch from Mark Zealey.
1608 PP/12 Bugzilla 1056: Improved spamd server selection.
1609 Patch from Mark Zealey.
1611 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1612 Based on patch from Heiko Schlittermann.
1614 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1615 Patch from Uwe Doering, sign-off by Michael Haardt.
1617 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1618 variable declaration deep within a block. Bug and patch from
1621 PP/15 lookups-Makefile IRIX compatibility coercion.
1623 PP/16 Make DISABLE_DKIM build knob functional.
1625 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1626 Patch by Simon Arlott
1628 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1629 variable argument macros. Our copy now differs from upstream.
1635 TF/01 Failure to get a lock on a hints database can have serious
1636 consequences so log it to the panic log.
1638 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1639 controlled using the smtp_confirmation log selector.
1641 TF/03 Include the error message when we fail to unlink a spool file.
1643 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1644 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1645 for maintaining out-of-tree patches for some time.
1647 PP/01 Bugzilla 139: Documentation and portability issues.
1648 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1649 Handle per-OS dynamic-module compilation flags.
1651 PP/02 Let /dev/null have normal permissions.
1652 The 4.73 fixes were a little too stringent and complained about the
1653 permissions on /dev/null. Exempt it from some checks.
1654 Reported by Andreas M. Kirchwitz.
1656 PP/03 Report version information for many libraries, including
1657 Exim version information for dynamically loaded libraries. Created
1658 version.h, now support a version extension string for distributors
1659 who patch heavily. Dynamic module ABI change.
1661 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1662 privilege escalation vulnerability whereby the Exim run-time user
1663 can cause root to append content of the attacker's choosing to
1666 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1669 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1670 If dropping privileges for untrusted macros, we disabled normal logging
1671 on the basis that it would fail; for the Exim run-time user, this is not
1672 the case, and it resulted in successful deliveries going unlogged.
1673 Fixed. Reported by Andreas Metzler.
1679 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1680 only prepend for the Resent-* case. Fixes regression introduced in
1681 Exim 4.70 by NM/22 for Bugzilla 607.
1683 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1684 increasing numbers of administrators be bitten by this.
1686 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1688 PP/03 Bugzilla 994: added openssl_options main configuration option.
1690 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1692 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1694 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1696 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1697 without a peer certificate, leading to a segfault because of an
1698 assumption that peers always have certificates. Be a little more
1699 paranoid. Problem reported by Martin Tscholak.
1701 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1702 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1703 NB: ClamAV planning to remove STREAM in "middle of 2010".
1704 CL also introduces -bmalware, various -d+acl logging additions and
1705 more caution in buffer sizes.
1707 PP/09 Implemented reverse_ip expansion operator.
1709 PP/10 Bugzilla 937: provide a "debug" ACL control.
1711 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1713 PP/12 Bugzilla 973: Implement --version.
1715 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1717 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1719 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1721 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1722 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1723 ignore trailing whitespace.
1725 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1727 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1730 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1731 Notification from Dr Andrew Aitchison.
1733 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1734 ExtendedDetectionInfo response format.
1735 Notification from John Horne.
1737 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1740 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1741 XSL and documented dependency on system catalogs, with examples of how
1744 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1747 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1748 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1749 configuration file which is writeable by the Exim user or group.
1751 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1752 of configuration files to cover files specified with the -C option if
1753 they are going to be used with root privileges, not just the default
1756 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1757 option (effectively making it always true).
1759 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1760 files to be used while preserving root privileges.
1762 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1763 that rogue child processes cannot use them.
1765 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1766 run-time user, instead of root.
1768 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1769 Exim run-time user without dropping privileges.
1771 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1772 result string, instead of calling string_vformat() twice with the same
1775 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1776 for other users. Others should always drop root privileges if they use
1777 -C on the command line, even for a whitelisted configure file.
1779 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1781 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1787 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1788 $data_path, and $header_path variables; fixed documentation bugs and
1791 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1792 exipick to access non-standard spools, including the "frozen" queue
1795 NM/01 Bugzilla 965: Support mysql stored procedures.
1796 Patch from Alain Williams
1798 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1800 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1801 Patch from Andreas Metzler
1803 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1804 Patch from Kirill Miazine
1806 NM/05 Bugzilla 671: Added umask to procmail example.
1808 JJ/03 installed exipick 20100323.0, fixing doc bug
1810 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1811 directory. Notification and patch from Dan Rosenberg.
1813 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1815 TK/02 Improve log output when DKIM signing operation fails.
1817 MH/01 Treat the transport option dkim_domain as a colon separated
1818 list, not as a single string, and sign the message with each element,
1819 omitting multiple occurences of the same signer.
1821 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1822 Bugzilla 985, 986. Patch by Simon Arlott
1824 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1825 Patch by Simon Arlott
1827 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1828 MBX locking. Notification from Dan Rosenberg.
1834 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1836 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1838 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1840 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1842 NM/04 Bugzilla 915: Flag broken perl installation during build.
1848 TK/01 Added patch by Johannes Berg that expands the main option
1849 "spamd_address" if it starts with a dollar sign.
1851 TK/02 Write list of recipients to X-Envelope-Sender header when building
1852 the mbox-format spool file for content scanning (suggested by Jakob
1855 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1856 (http://www.dcc-servers.net/) support via dccifd. Activated by
1857 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1859 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1860 by Mark Daniel Reidel <mr@df.eu>.
1862 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1863 When building exim an external PCRE library is now needed -
1864 PCRE is a system library on the majority of modern systems.
1865 See entry on PCRE_LIBS in EDITME file.
1867 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1868 conversation. Added nologin parameter to request.
1869 Patch contributed by Kirill Miazine.
1871 TF/01 Do not log submission mode rewrites if they do not change the address.
1873 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1875 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1876 log files in place. Contributed by Roberto Lima.
1878 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1880 TF/03 Bugzilla 615: When checking the local_parts router precondition
1881 after a local_part_suffix or local_part_prefix option, Exim now
1882 does not use the address's named list lookup cache, since this
1883 contains cached lookups for the whole local part.
1885 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1886 Robert Millan. Documentation is in experimental-spec.txt.
1888 TF/04 Bugzilla 668: Fix parallel build (make -j).
1890 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1892 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1893 Patch provided by Jan Srzednicki.
1895 TF/05 Leading white space used to be stripped from $spam_report which
1896 wrecked the formatting. Now it is preserved.
1898 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1899 that they are available at delivery time.
1901 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1903 TF/08 TLS error reporting now respects the incoming_interface and
1904 incoming_port log selectors.
1906 TF/09 Produce a more useful error message if an SMTP transport's hosts
1907 setting expands to an empty string.
1909 NM/06 Bugzilla 744: EXPN did not work under TLS.
1910 Patch provided by Phil Pennock.
1912 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1913 Patch provided by Richard Godbee.
1915 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1916 acl_smtp_notquit, added index entry.
1918 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1919 Patch provided by Eugene Bujak.
1921 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1922 accept(). Patch provided by Maxim Dounin.
1924 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1925 Patch provided by Phil Pennock.
1927 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1929 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1930 Patch provided by Brad "anomie" Jorsch.
1932 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1933 Patch provided by Dean Brooks.
1935 TK/05 Add native DKIM support (does not depend on external libraries).
1937 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1938 Patch provided by Graeme Fowler.
1940 NM/16 Bugzilla 851: Documentation example syntax fix.
1942 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1944 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1947 NM/19 Bugzilla 745: TLS version reporting.
1948 Patch provided by Phil Pennock.
1950 NM/20 Bugzilla 167: bool: condition support.
1951 Patch provided by Phil Pennock.
1953 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1954 clients. Patch provided by Phil Pennock.
1956 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1957 Patch provided by Brad "anomie" Jorsch.
1959 NM/23 Bugzilla 687: Fix misparses in eximstats.
1960 Patch provided by Heiko Schlittermann.
1962 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1963 Patch provided by Heiko Schlittermann.
1965 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1966 plus update to original patch.
1968 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1970 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1971 Patch provided by David Brownlee.
1973 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1975 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1977 NM/30 Bugzilla 888: TLS documentation bugfixes.
1979 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1981 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1982 Unlike the original bugzilla I have changed all shell scripts in src tree.
1984 NM/33 Bugzilla 898: Transport filter timeout fix.
1985 Patch by Todd Rinaldo.
1987 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1988 Patch by Serge Demonchaux.
1990 NM/35 Bugzilla 39: Base64 decode bug fixes.
1991 Patch by Jakob Hirsch.
1993 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1995 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1997 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1999 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2005 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2006 ALT-N's libdkim that I have put here:
2007 http://duncanthrax.net/exim-experimental/
2009 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2010 were called 'true' and it seems that C99 defines that as a reserved
2011 keyword to be used with 'bool' variable types. That means you could
2012 not include C99-style headers which use bools without triggering
2013 build errors in sieve.c.
2015 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2016 as mailq or other aliases. Changed the --help handling significantly
2017 to do whats expected. exim_usage() emits usage/help information.
2019 SC/01 Added the -bylocaldomain option to eximstats.
2021 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2023 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2025 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2031 PH/01 Another patch from the Sieve maintainer.
2033 PH/02 When an IPv6 address is converted to a string for single-key lookup
2034 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2035 dots are used instead of colons so that keys in lsearch files need not
2036 contain colons. This was done some time before quoting was made available
2037 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2038 (notated using the quote facility) so as to distinguish them from IPv4
2039 keys. This meant that lookups for IP addresses in host lists did not work
2040 for iplsearch lookups.
2042 This has been fixed by arranging for IPv6 addresses to be expressed with
2043 colons if the lookup type is iplsearch. This is not incompatible, because
2044 previously such lookups could never work.
2046 The situation is now rather anomalous, since one *can* have colons in
2047 ordinary lsearch keys. However, making the change in all cases is
2048 incompatible and would probably break a number of configurations.
2050 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2053 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2054 conversion specification without a maximum field width, thereby enabling
2055 a rogue spamd server to cause a buffer overflow. While nobody in their
2056 right mind would setup Exim to query an untrusted spamd server, an
2057 attacker that gains access to a server running spamd could potentially
2058 exploit this vulnerability to run arbitrary code as the Exim user.
2060 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2061 $primary_hostname instead of what libspf2 thinks the hosts name is.
2063 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2064 a directory entry by the name of the lookup key. Previously, if a
2065 symlink pointed to a non-existing file or a file in a directory that
2066 Exim lacked permissions to read, a lookup for a key matching that
2067 symlink would fail. Now it is enough that a matching directory entry
2068 exists, symlink or not. (Bugzilla 503.)
2070 PH/03 The body_linecount and body_zerocount variables are now exported in the
2073 PH/04 Added the $dnslist_matched variable.
2075 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2076 This means they are set thereafter only if the connection becomes
2079 PH/06 Added the client_condition to authenticators so that some can be skipped
2080 by clients under certain conditions.
2082 PH/07 The error message for a badly-placed control=no_multiline_responses left
2083 "_responses" off the end of the name.
2085 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2087 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2088 (without spaces) instead of just copying the configuration text.
2090 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2092 PH/11 Added $max_received_linelength.
2094 PH/12 Added +ignore_defer and +include_defer to host lists.
2096 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2097 way in which PCRE > 7.0 is built.
2099 PH/14 Implemented queue_only_load_latch.
2101 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2102 MAIL command. The effect was to mangle the value on 64-bit systems.
2104 PH/16 Another patch from the Sieve maintainer.
2106 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2108 PH/18 If a system quota error occurred while trying to create the file for
2109 a maildir delivery, the message "Mailbox is full" was not appended to the
2110 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2111 only to a quota excession during the actual writing of the file.
2113 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2114 characters?) which causes problems in log lines. The DN values are now
2115 passed through string_printing() before being added to log lines.
2117 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2118 and InterBase are left for another time.)
2120 PH/21 Added message_body_newlines option.
2122 PH/22 Guard against possible overflow in moan_check_errorcopy().
2124 PH/23 POSIX allows open() to be a macro; guard against that.
2126 PH/24 If the recipient of an error message contained an @ in the local part
2127 (suitably quoted, of course), incorrect values were put in $domain and
2128 $local_part during the evaluation of errors_copy.
2134 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2135 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2138 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2139 issue a MAIL command.
2141 PH/02 In an ACL statement such as
2143 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2145 if a client was not listed at all, or was listed with a value other than
2146 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2147 the condition was not true (as it should be), so access was not denied.
2148 The bug was that the ! inversion was incorrectly passed on to the second
2149 item. This has been fixed.
2151 PH/03 Added additional dnslists conditions == and =& which are different from
2152 = and & when the dns lookup returns more than one IP address.
2154 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2155 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2157 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2158 FSYNC, which compiles an option called disable_fsync that allows for
2159 bypassing fsync(). The documentation is heavily laced with warnings.
2161 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2163 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2164 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2165 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2166 including adding "make clean"; (3) Added -fPIC when compiling the test
2167 dynamically loaded module, to get rid of a warning.
2169 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2170 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2171 = 0s. The bug is otherwise harmless.
2173 PH/07 There was a bug in the dovecot authenticator such that the value of
2174 $auth1 could be overwritten, and so not correctly preserved, after a
2175 successful authentication. This usually meant that the value preserved by
2176 the server_setid option was incorrect.
2178 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2180 PH/09 Installed PCRE release 7.0.
2182 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2183 run for batched SMTP input. It is now run at the start of every message
2184 in the batch. While fixing this I discovered that the process information
2185 (output by running exiwhat) was not always getting set for -bs and -bS
2186 input. This is fixed, and it now also says "batched" for BSMTP.
2188 PH/11 Added control=no_pipelining.
2190 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2191 patch, slightly modified), and move the expansion of helo_data till after
2192 the connection is made in the smtp transport (so it can use these
2195 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2197 PH/14 Added log_selector = +pid.
2199 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2201 PH/16 Add ${if forany and ${if forall.
2203 PH/17 Added dsn_from option to vary the From: line in DSNs.
2205 PH/18 Flush SMTP output before performing a callout, unless control =
2206 no_callout_flush is set.
2208 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2209 was true (the default) a successful delivery failed to delete the retry
2210 item, thus causing premature timeout of the address. The bug is now
2213 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2215 PH/21 Long custom messages for fakedefer and fakereject are now split up
2216 into multiline responses in the same way that messages for "deny" and
2217 other ACL rejections are.
2219 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2220 with slight modification.
2222 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2223 draft, changing the syntax and factoring some duplicate code".
2225 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2226 for deliveries of the second and subsequent messages over the same SMTP
2229 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2230 ${reduce, with only minor "tidies".
2232 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2234 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2235 expansion side effects.
2237 PH/27 When a message times out after an over-quota error from an Exim-imposed
2238 quota, the bounce message says "mailbox is full". This message was not
2239 being given when it was a system quota that was exceeded. It now should
2242 MH/03 Made $recipients available in local_scan(). local_scan() already has
2243 better access to the recipient list through recipients_list[], but
2244 $recipients can be useful in postmaster-provided expansion strings.
2246 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2247 in the case of a MAIL command with additional options following the
2248 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2249 were accidentally chopped off.
2251 PH/29 SMTP synchronization checks are implemented when a command is read -
2252 there is a check that no more input is waiting when there shouldn't be
2253 any. However, for some commands, a delay in an ACL can mean that it is
2254 some time before the response is written. In this time, more input might
2255 arrive, invalidly. So now there are extra checks after an ACL has run for
2256 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2257 pipelining has not been advertised.
2259 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2261 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2262 returned in the SMTP response when smtp_return_error_details was false.
2263 This has been fixed.
2265 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2266 instead of the C I/O that was originally supplied, because problems were
2267 reported on Solaris.
2269 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2270 Exim which did not show up earlier: it was assuming that a call to
2271 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2272 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2273 was a macro that became an assignment, so it seemed to work. This has
2274 changed to a proper function call with a void return, hence the compile
2275 error. Exim's code has been fixed.
2277 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2280 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2282 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2284 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2285 because it turns out to be too big or there is a timeout) there is no
2286 "Completed" line in the log. When some messages of this type were
2287 selected by exigrep, they were listed as "not completed". Others were
2288 picked up by some special patterns. I have improved the selection
2289 criteria to be more general.
2291 PH/38 The host_find_failed option in the manualroute router can now be set
2292 to "ignore", to completely ignore a host whose IP address cannot be
2293 found. If all hosts are ignored, the behaviour is controlled by the new
2294 host_all_ignored option.
2296 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2297 homing or because of multiple MX records with /mx) generated more than
2298 one IP address, and the following item turned out to be the local host,
2299 all the secondary addresses of the first item were incorrectly removed
2300 from the list, along with the local host and any following hosts (which
2301 is what is supposed to happen).
2303 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2304 whoever called Exim into the -H file. In the case of the daemon it was
2305 behaving confusingly. When first started, it used values for whoever
2306 started the daemon, but after a SIGHUP it used the Exim user (because it
2307 calls itself on a restart). I have changed the code so that it now always
2310 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2311 message are rejected with the same error (e.g. no authentication or bad
2312 sender address), and a DATA command is nevertheless sent (as can happen
2313 with PIPELINING or a stupid MUA), the error message that was given to the
2314 RCPT commands is included in the rejection of the DATA command. This is
2315 intended to be helpful for MUAs that show only the final error to their
2318 PH/42 Another patch from the Sieve maintainer.
2320 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2321 Eximstats - Fixed some broken HTML links and added missing column headers
2323 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2324 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2326 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2328 PH/43 Yet another patch from the Sieve maintainer.
2330 PH/44 I found a way to check for a TCP/IP connection going away before sending
2331 the response to the final '.' that terminates a message, but only in the
2332 case where the client has not sent further data following the '.'
2333 (unfortunately, this is allowed). However, in many cases there won't be
2334 any further data because there won't be any more messages to send. A call
2335 to select() can be used: if it shows that the input is "ready", there is
2336 either input waiting, or the socket has been closed. An attempt to read
2337 the next input character can distinguish the two cases. Previously, Exim
2338 would have sent an OK response which the client would never have see.
2339 This could lead to message repetition. This fix should cure that, at
2340 least in a lot of common cases.
2342 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2343 advertised in response to EHLO.
2349 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2350 fixed by 4.65/MH/01 (is this a record?) are fixed:
2352 (i) An empty string was always treated as zero by the numeric comparison
2353 operators. This behaviour has been restored.
2355 (ii) It is documented that the numeric comparison operators always treat
2356 their arguments as decimal numbers. This was broken in that numbers
2357 starting with 0 were being interpreted as octal.
2359 While fixing these problems I realized that there was another issue that
2360 hadn't been noticed. Values of message_size_limit (both the global option
2361 and the transport option) were treated as octal if they started with 0.
2362 The documentation was vague. These values are now always treated as
2363 decimal, and I will make that clear in the documentation.
2369 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2370 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2373 MH/01 Don't check that the operands of numeric comparison operators are
2374 integers when their expansion is in "skipping" mode (fixes bug
2375 introduced by 4.64-PH/07).
2377 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2378 child addresses, Exim now panics and dies. Previously, because the count
2379 is held in a short int, deliveries were likely to be lost. As such a
2380 large number of recipients for a single message is ridiculous
2381 (performance will be very, very poor), I have chosen to impose a limit
2382 rather than extend the field.
2388 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2389 leftover -K file (the existence of which was triggered by #402).
2390 While we were at it, introduced process PID as part of the -K
2391 filename. This should rule out race conditions when creating
2394 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2395 processing considerably. Previous code took too long for large mails,
2396 triggering a timeout which in turn triggers #401.
2398 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2399 in the DK code in transports.c. sendfile() is not really portable,
2400 hence the _LINUX specificness.
2402 TF/01 In the add_headers option to the mail command in an Exim filter,
2403 there was a bug that Exim would claim a syntax error in any
2404 header after the first one which had an odd number of characters
2407 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2408 callout verification, Exim cached a "reject" for the entire domain. This
2409 is correct for most verifications, but it is not correct for a recipient
2410 verification with use_sender or use_postmaster set, because in that case
2411 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2412 case of MAIL FROM:<> rejection from other early rejections (e.g.
2413 rejection of HELO). When verifying a recipient using a non-null MAIL
2414 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2415 Whatever the result of the callout, the value of the domain cache is
2416 left unchanged (for any other kind of callout, getting as far as trying
2417 RCPT means that the domain itself is ok).
2419 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2422 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2423 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2424 session, the daemon ignores ECONNECT errors and logs others; it now
2425 ignores EPIPE as well.
2427 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2428 (quoted-printable decoding).
2430 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2431 later the small subsequent patch to fix an introduced bug.
2433 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2435 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2437 PH/08 An error is now given if message_size_limit is specified negative.
2439 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2440 to be given (somewhat) arbitrary names.
2442 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2445 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2446 miscellaneous code fixes
2448 PH/10 Added the log_reject_target ACL modifier to specify where to log
2451 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2452 hostname. This is wrong, because it relates to the incoming message (and
2453 probably the interface on which it is arriving) and not to the outgoing
2454 callout (which could be using a different interface). This has been
2455 changed to use the value of the helo_data option from the smtp transport
2456 instead - this is what is used when a message is actually being sent. If
2457 there is no remote transport (possible with a router that sets up host
2458 addresses), $smtp_active_hostname is used.
2460 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2461 tweaks were necessary in order to get it to work (see also 21 below):
2462 (a) The code assumed that strncpy() returns a negative number on buffer
2463 overflow, which isn't the case. Replaced with Exim's string_format()
2465 (b) There were several signed/unsigned issues. I just did the minimum
2466 hacking in of casts. There is scope for a larger refactoring.
2467 (c) The code used strcasecmp() which is not a standard C function.
2468 Replaced with Exim's strcmpic() function.
2469 (d) The code set only $1; it now sets $auth1 as well.
2470 (e) A simple test gave the error "authentication client didn't specify
2471 service in request". It would seem that Dovecot has changed its
2472 interface. Fortunately there's a specification; I followed it and
2473 changed what the client sends and it appears to be working now.
2475 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2478 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2479 address A is aliased to B and C, where B exists and C does not. Without
2480 -v the output is "A verified" because verification stops after a
2481 successful redirection if more than one address is generated. However,
2482 with -v the child addresses are also verified. Exim was outputting "A
2483 failed to verify" and then showing the successful verification for C,
2484 with its parentage. It now outputs "B failed to verify", showing B's
2485 parentage before showing the successful verification of C.
2487 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2488 look up a TXT record in a specific list after matching in a combined
2491 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2492 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2493 they consult the DNS. I had assumed they would set it the way they
2494 wanted; and indeed my experiments on Linux seem to show that in some
2495 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2496 To be on the safe side, however, I have now made the interface to
2497 host_find_byname() similar to host_find_bydns(), with an argument
2498 containing the DNS resolver options. The host_find_byname() function now
2499 sets these options at its start, just as host_find_bydns() does. The smtp
2500 transport options dns_qualify_single and dns_search_parents are passed to
2501 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2502 of host_find_byname() use the default settings of RES_DEFNAMES
2503 (qualify_single) but not RES_DNSRCH (search_parents).
2505 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2506 spool_read_header() do less string testing, by means of a preliminary
2507 switch on the second character of optional "-foo" lines. (This is
2508 overdue, caused by the large number of possibilities that now exist.
2509 Originally there were few.) While I was there, I also converted the
2510 str(n)cmp tests so they don't re-test the leading "-" and the first
2511 character, in the hope this might squeeze out yet more improvement.
2513 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2514 flag allowing group syntax was set by the header_syntax check but not
2515 turned off, possible causing trouble later; (2) The flag was not being
2516 set at all for the header_verify test, causing "group"-style headers to
2517 be rejected. I have now set it in this case, and also caused header_
2518 verify to ignore an empty address taken from a group. While doing this, I
2519 came across some other cases where the code for allowing group syntax
2520 while scanning a header line wasn't quite right (mostly, not resetting
2521 the flag correctly in the right place). These bugs could have caused
2522 trouble for malformed header lines. I hope it is now all correct.
2524 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2525 with the "reply" argument non-NULL. The code, however (which originally
2526 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2527 but it didn't always do it. This confused somebody who was copying the
2528 code for some other use. I have removed all the tests.
2530 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2531 feature that was used to support insecure browsers during the U.S. crypto
2532 embargo. It requires special client support, and Exim is probably the
2533 only MTA that supported it -- and would never use it because real RSA is
2534 always available. This code has been removed, because it had the bad
2535 effect of slowing Exim down by computing (never used) parameters for the
2536 RSA_EXPORT functionality.
2538 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2539 authenticator to fail if there's a tab character in the incoming data
2540 (there should never be unless someone is messing about, as it's supposed
2541 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2542 if the connection is using TLS or if the remote IP is the same as the
2543 local IP, and the "valid-client-cert option" if a client certificate has
2546 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2547 authenticators. This can be used for authorization after authentication
2548 succeeds. (In the case of plaintext, it servers for both authentication
2551 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2552 if any retry times were supplied.
2554 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2555 connection, where there is no client IP address to check. In this
2556 situation, the verify now always succeeds.
2558 PH/25 Applied John Jetmore's -Mset patch.
2560 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2562 PH/27 In a string expansion for a processed (not raw) header when multiple
2563 headers of the same name were present, leading whitespace was being
2564 removed from all of them, but trailing whitespace was being removed only
2565 from the last one. Now trailing whitespace is removed from each header
2566 before concatenation. Completely empty headers in a concatenation (as
2567 before) are ignored.
2569 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2570 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2572 PH/29 [Removed. This was a change that I later backed out, and forgot to
2573 correct the ChangeLog entry (that I had efficiently created) before
2574 committing the later change.]
2576 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2577 address errors (4xx response to RCPT) over the same connection as other
2578 messages routed to the same hosts. Such deliveries are always "forced",
2579 so retry times are not inspected. This resulted in far too many retries
2580 for the affected addresses. The effect occurred only when there were more
2581 hosts than the hosts_max_try setting in the smtp transport when it had
2582 the 4xx errors. Those hosts that it had tried were not added to the list
2583 of hosts for which the message was waiting, so if all were tried, there
2584 was no problem. Two fixes have been applied:
2586 (i) If there are any address or message errors in an SMTP delivery, none
2587 of the hosts (tried or untried) are now added to the list of hosts
2588 for which the message is waiting, so the message should not be a
2589 candidate for sending over the same connection that was used for a
2590 successful delivery of some other message. This seems entirely
2591 reasonable: after all the message is NOT "waiting for some host".
2592 This is so "obvious" that I'm not sure why it wasn't done
2593 previously. Hope I haven't missed anything, but it can't do any
2594 harm, as the worst effect is to miss an optimization.
2596 (ii) If, despite (i), such a delivery is accidentally attempted, the
2597 routing retry time is respected, so at least it doesn't keep
2598 hammering the server.
2600 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2601 in ${readsocket because some servers need this prod.
2603 PH/32 Added some extra debug output when updating a wait-xxx database.
2605 PH/33 The hint "could be header name not terminated by colon", which has been
2606 given for certain expansion errors for a long time, was not being given
2607 for the ${if def:h_colon_omitted{... case.
2609 PH/34 The spec says: "With one important exception, whenever a domain list is
2610 being scanned, $domain contains the subject domain." There was at least
2611 one case where this was not true.
2613 PH/35 The error "getsockname() failed: connection reset by peer" was being
2614 written to the panic log as well as the main log, but it isn't really
2615 panic-worthy as it just means the connection died rather early on. I have
2616 removed the panic log writing for the ECONNRESET error when getsockname()
2619 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2620 runs only) independently of the message's sender address. This meant
2621 that, if the 4xx error was in fact related to the sender, a different
2622 message to the same recipient with a different sender could confuse
2623 things. In particular, this can happen when sending to a greylisting
2624 server, but other circumstances could also provoke similar problems.
2625 I have changed the default so that the retry time for these errors is now
2626 based a combination of the sender and recipient addresses. This change
2627 can be overridden by setting address_retry_include_sender=false in the
2630 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2631 remote server are returned as part of bounce messages. This was not
2632 happening for LMTP over a pipe (the lmtp transport), but now it is the
2633 same for both kinds of LMTP.
2635 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2636 in header lines that were in fact CNAMEs. This is no longer the case.
2638 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2639 and queue runs started by the daemon processed all messages. This has
2640 been fixed so that -R and -S can now usefully be given with -q<time>.
2642 PH/40 Import PCRE release 6.7 (fixes some bugs).
2644 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2646 PH/42 Give an error if -q is specified more than once.
2648 PH/43 Renamed the variables $interface_address and $interface_port as
2649 $received_ip_address and $received_port, to make it clear that these
2650 values apply to message reception, and not to the outgoing interface when
2651 a message is delivered. (The old names remain recognized, of course.)
2653 PH/44 There was no timeout on the connect() call when using a Unix domain
2654 socket in the ${readsocket expansion. There now is.
2656 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2657 be meaningful with "accept".
2659 SC/01 Eximstats V1.43
2660 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2662 SC/02 Eximstats V1.44
2663 Use a glob alias rather than an array ref in the generated
2664 parser. This improves both readability and performance.
2666 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2667 Collect SpamAssassin and rejection statistics.
2668 Don't display local sender or destination tables unless
2669 there is data to show.
2670 Added average volumes into the top table text output.
2672 SC/04 Eximstats V1.46
2673 Collect data on the number of addresses (recipients)
2674 as well as the number of messages.
2676 SC/05 Eximstats V1.47
2677 Added 'Message too big' to the list of mail rejection
2678 reasons (thanks to Marco Gaiarin).
2680 SC/06 Eximstats V1.48
2681 Mainlog lines which have GMT offsets and are too short to
2682 have a flag are now skipped.
2684 SC/07 Eximstats V1.49 (Alain Williams)
2685 Added the -emptyok flag.
2687 SC/08 Eximstats V1.50
2688 Fixes for obtaining the IP address from reject messages.
2690 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2691 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2692 whitespace changes from 4.64-PH/27
2694 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2697 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2698 are found, allow negative numbers in numeric criteria)
2700 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2702 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2703 to match changes made in 4.64-PH/43
2705 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2707 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2708 "domain" is actually an IP address, and force a failure. This locks out
2709 those revolvers/nameservers that support "A-for-A" lookups, in
2710 contravention of the specifications.
2712 PH/48 When a host name was looked up from an IP address, and the subsequent
2713 forward lookup of the name timed out, the host name was left in
2714 $sender_host_name, contrary to the specification.
2716 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2717 restricted to single-key lookups, Exim was not diagnosing an error if
2718 * or *@ was used with a query-style lookup.
2720 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2722 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2723 long ago, but noone interested enough thought of it. Let's just say that
2724 the "1.1" means that there are some new functions that weren't there at
2725 some point in the past.
2727 PH/51 Error processing for expansion failure of helo_data from an smtp
2728 transport during callout processing was broken.
2730 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2731 tested/used via the -bh/-bhc/-bs options.
2733 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2734 bug, fixed in subsequent PCRE releases).
2736 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2737 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2739 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2745 SC/01 Use a glob alias rather than an array ref in eximstats generated
2746 parser. This improves both readability and performance.
2748 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2749 Don't display local sender or destination tables in eximstats unless
2750 there is data to show.
2751 Added average volumes into the eximstats top table text output.
2753 SC/03 Collect data on the number of addresses (recipients) as well
2754 as the number of messages in eximstats.
2756 TF/01 Correct an error in the documentation for the redirect router. Exim
2757 does (usually) call initgroups() when daemonizing.
2759 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2760 with consistent privilege compared to when running as a daemon.
2762 TF/03 Note in the spec that $authenticated_id is not set for local
2763 submissions from trusted users.
2765 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2766 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2768 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2769 by adding some example configuration directives to the default
2770 configuration file. A little bit of work is required to uncomment the
2771 directives and define how usernames and passwords are checked, but
2772 there is now a framework to start from.
2774 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2775 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2776 without this. I don't know how relevant this is to other LDAP libraries.
2778 PH/02 Add the verb name to the "unknown ACL verb" error.
2780 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2782 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2784 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2785 directory not expanded when it should be if an expanded home directory
2786 was set for the address (which is overridden by the transport).
2788 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2791 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2792 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2793 because it is too late at that time, and has no effect.
2795 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2796 security issue with \' (bugzilla #107). I could not use the
2797 PQescapeStringConn() function, because it needs a PGconn value as one of
2800 PH/08 When testing addresses using -bt, indicate those final addresses that
2801 are duplicates that would not cause an additional delivery. At least one
2802 person was confused, thinking that -bt output corresponded to deliveries.
2803 (Suppressing duplicates isn't a good idea as you lose the information
2804 about possibly different redirections that led to the duplicates.)
2806 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2807 systems where poll() doesn't work, in particular OS X.
2809 PH/10 Added more information to debugging output for retry time not reached.
2811 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2812 operations in malware.c.
2814 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2817 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2818 syslog_duplication set false, log lines that would normally be written
2819 both the the main log and to the reject log were not written to syslog at
2822 PH/14 In the default configuration, change the use of "message" in ACL warn
2823 statements to "add_header".
2825 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2826 not followed by a command (e.g. "seen endif").
2828 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2829 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2832 PH/17 Added extra conditions to the default value of delay_warning_condition
2836 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2837 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2838 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2841 The Auto-Submitted: and various List- headers are standardised, whereas I
2842 don't think Precedence: ever was.
2844 PH/18 Refactored debugging code in route_finduser() to show more information,
2845 in particular, the error code if getpwnam() issues one.
2847 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2848 This is apparently needed in addition to the PH/07 change above to avoid
2849 any possible encoding problems.
2851 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2852 but not after initializing Perl.
2854 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2855 output them only if debugging. By default they are written stderr,
2856 apparently, which is not desirable.
2858 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2861 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2864 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2866 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2867 authenticated or an ident call has been made. Suppress the default
2868 values for $authenticated_id and $authenticated_sender (but permit -oMai
2869 and -oMas) when testing with -bh.
2871 PH/24 Re-jigged the order of the tests in the default configuration so that the
2872 tests for valid domains and recipients precede the DNS black list and CSA
2873 tests, on the grounds that those ones are more expensive.
2875 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2876 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2877 This bug exists in every version of Exim that I still have, right back to
2880 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2881 However, an attempt to turn on case-sensitivity in a regex key by
2882 including (?-i) didn't work because the subject string was already
2883 lowercased, and the effects were non-intuitive. It turns out that a
2884 one-line patch can be used to allow (?-i) to work as expected.
2890 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2891 other effects) broke the use of negated acl sub-conditions.
2893 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2896 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2897 "Deny" causes Exim to reject the incoming connection with a 554 error.
2898 Unfortunately, if there is a major crisis, such as a disk failure,
2899 tcp-wrappers gives "deny", whereas what one would like would be some
2900 kind of temporary error. A kludge has been added to help with this.
2901 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2902 554 error is used if errno is still zero or contains ENOENT (which occurs
2903 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2906 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2908 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2909 errors. Otherwise a message that provokes a temporary error (when other
2910 messages do not) can cause a whole host to time out.
2912 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2913 addresses were routed directly to files or pipes from a redirect router.
2914 File deliveries just didn't batch; pipe deliveries might have suffered
2917 PH/06 A failure to get a lock for a hints database would erroneously always say
2918 "Failed to get write lock", even when it was really a read lock.
2920 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2921 of 0600. This has been changed to use the value of the lockfile_mode
2922 option (which defaults to 0600).
2924 PH/08 Applied small patch from the Sieve maintainer.
2926 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2927 folder from quota calculations, a direct delivery into this folder messed
2928 up the contents of the maildirsize file. This was because the regex was
2929 used only to exclude .Trash (or whatever) when the size of the mailbox
2930 was calculated. There was no check that a delivery was happening into an
2931 excluded directory. This bug has been fixed by ignoring all quota
2932 processing for deliveries into excluded directories.
2934 PH/10 Added the maildirfolder_create_regex option to appendfile.
2940 PH/01 The code for finding all the local interface addresses on a FreeBSD
2941 system running IPv6 was broken. This may well have applied to all BSD
2942 systems, as well as to others that have similar system calls. The broken
2943 code found IPv4 interfaces correctly, but gave incorrect values for the
2944 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2945 that it would not match correctly against @[] and not recognize the IPv6
2948 PH/02 The ipliteral router was not recognizing addresses of the form user@
2949 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2951 PH/03 Added disable_ipv6.
2953 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2954 decoded form, because it is most often used to construct To: headers
2955 lines in autoreplies, and the decoded form may well be syntactically
2956 invalid. However, $reply_address has leading white space removed, and all
2957 newlines turned into spaces so that the autoreply transport does not
2960 PH/05 If group was specified without a user on a router, and no group or user
2961 was specified on a transport, the group from the router was ignored.
2963 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2964 for visible compile-time settings that can be used to change these
2965 numbers, for those that want even more. Backwards compatibility with old
2966 spool files has been maintained. However, going back to a previous Exim
2967 release will lost any variables that are in spool files.
2969 PH/07 Two small changes when running in the test harness: increase delay when
2970 passing a TCP/IP connection to a new process, in case the original
2971 process has to generate a bounce, and remove special handling of
2972 127.0.0.2 (sic), which is no longer necessary.
2974 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2975 be the same on different OS.
2977 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2980 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2981 whether --show-vars was specified or not
2983 JJ/02 exipick: Added support for new ACL variable spool format introduced
2986 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2987 syntactically invalid From: or Reply-to: line, and a filter used this to
2988 generate an autoreply, and therefore failed to obtain an address for the
2989 autoreply, Exim could try to deliver to a non-existent relative file
2990 name, causing unrelated and misleading errors. What now happens is that
2991 it logs this as a hard delivery error, but does not attempt to create a
2994 PH/11 The exinext utility has a -C option for testing purposes, but although
2995 the given file was scanned by exinext itself; it wasn't being passed on
2996 when Exim was called.
2998 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2999 an end-of-file indication when reading a command response.
3001 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3002 compiled. In many other places in Exim, IPv6 addresses are always
3003 recognized, so I have changed this. It also means that IPv4 domain
3004 literals of the form [IPV4:n.n.n.n] are now always recognized.
3006 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3007 used if the router is not running as root, for example, when verifying at
3008 ACL time, or when using -bh. The debugging output from this situation was
3009 non-existent - all you got was a failure to exec. I have made two
3012 (a) Failures to set uid/gid, the current directory, or a process leader
3013 in a subprocess such as that created by queryprogram now generate
3014 suitable debugging output when -d is set.
3016 (b) The queryprogram router detects when it is not running as root,
3017 outputs suitable debugging information if -d is set, and then runs
3018 the subprocess without attempting to change uid/gid.
3020 PH/15 Minor change to Makefile for building test_host (undocumented testing
3023 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3024 additional section of a DNS packet that returns MX or SRV records.
3025 Instead, it always explicitly searches for A/AAAA records. This avoids
3026 major problems that occur when a DNS server includes only records of one
3027 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3028 fixed another bug: if SRV records were looked up and the corresponding
3029 address records were *not* found in the additional section, the port
3030 values from the SRV records were lost.
3032 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3033 using the correct key (the original address) when searching the retry
3034 rules in order to find which one to use for generating the retry hint.
3036 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3037 adding the default one. Similarly, if it contains a Reply-To: header, the
3038 errors_reply_to option, if set, is not used.
3040 PH/19 When calculating a retry time, Exim used to measure the "time since
3041 failure" by looking at the "first failed" field in the retry record. Now
3042 it does not use this if it is later than than the arrival time of the
3043 message. Instead it uses the arrival time. This makes for better
3044 behaviour in cases where some deliveries succeed, thus re-setting the
3045 "first failed" field. An example is a quota failure for a huge message
3046 when small messages continue to be delivered. Without this change, the
3047 "time since failure" will always be short, possible causing more frequent
3048 delivery attempts for the huge message than are intended.
3049 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3051 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3052 $1, $2, $3) because the numerical variables can be reset during some
3053 expansion items (e.g. "match"), thereby losing the authentication data.
3055 PH/21 Make -bV show the size of off_t variables so that the test suite can
3056 decide whether to run tests for quotas > 2G.
3058 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3059 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3060 filecount value is greater than 2G or if a quota value is greater than 2G
3061 on a system where the size of off_t is not greater than 4, a panic error
3064 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3065 never match. The debug and -bh output now contains an explicit error
3066 message indicating a malformed IPv4 address or mask.
3068 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3069 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3070 PH/23 above applies.
3072 PH/25 Do not write to syslog when running in the test harness. The only
3073 occasion when this arises is a failure to open the main or panic logs
3074 (for which there is an explicit test).
3076 PH/26 Added the /no_tell option to "control=freeze".
3078 PH/27 If a host name lookup failed very early in a connection, for example, if
3079 the IP address matched host_lookup and the reverse lookup yielded a name
3080 that did not have a forward lookup, an error message of the form "no IP
3081 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3082 could be logged. Now it outputs the IP address instead of "NULL".
3084 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3085 allows the sender and the authenticated sender to be set when
3086 submitting a message from within Exim. Since child_open_exim() is
3087 documented for local_scan(), the new function should be too.
3089 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3090 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3091 results in an empty string is now treated as unset.
3093 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3095 PH/31 Added sender_verify_fail logging option.
3097 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3098 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3099 tidied the source and removed it altogether.
3101 PH/33 When a queue run was abandoned because the load average was too high, a
3102 log line was always written; now it is written only if the queue_run log
3103 selector is set. In addition, the log line for abandonment now contains
3104 information about the queue run such as the pid. This is always present
3105 in "start" and "stop" lines but was omitted from the "abandon" line.
3107 PH/34 Omit spaces between a header name and the colon in the error message that
3108 is given when verify = headers_syntax fails (if there are lots of them,
3109 the message gets confusing).
3111 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3112 names, as there are now some PTR records that contain slashes. This check
3113 is only to protect against broken name servers that fall over on strange
3114 characters, so the fact that it applies to all lookups doesn't matter.
3116 PH/36 Now that the new test suite is complete, we can remove some of the
3117 special code in Exim that was needed for the old test suite. For example,
3118 sorting DNS records because real resolvers return them in an arbitrary
3119 order. The new test suite's fake resolver always returns records in the
3122 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3123 bounces) except when queue_only is set, to avoid logging races between
3124 the different processes.
3126 PH/38 Panic-die if .include specifies a non-absolute path.
3128 PH/39 A tweak to the "H" retry rule from its user.
3130 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3131 a label. They prevented compilation on older perls.
3133 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3134 a warning to be raised on newish perls.
3136 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3137 on queue. Changes to match documented behaviour of showing count of
3138 messages matching specified criteria.
3140 PH/40 Changed the default ident timeout from 30s to 5s.
3142 PH/41 Added support for the use of login_cap features, on those BSD systems
3143 that have them, for controlling the resources used by pipe deliveries.
3145 PH/42 The content-scanning code uses fopen() to create files in which to put
3146 message data. Previously it was not paying any attention to the mode of
3147 the files. Exim runs with umask(0) because the rest of the code creates
3148 files with open(), and sets the required mode explicitly. Thus, these
3149 files were ending up world-writeable. This was not a big issue, because,
3150 being within the spool directory, they were not world-accessible. I have
3151 created a function called modefopen, which takes an additional mode
3152 argument. It sets umask(777), creates the file, chmods it to the required
3153 mode, then resets the umask. All the relevant calls to fopen() in the
3154 content scanning code have been changed to use this function.
3156 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3157 to 24 hours. This avoids potential overflow problems when processing G
3158 and H retry rules. I suspect nobody ever tinkers with this value.
3160 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3162 PH/45 When the plaintext authenticator is running as a client, the server's
3163 challenges are checked to ensure they are valid base64 strings. By
3164 default, the authentication attempt is cancelled if an invalid string is
3165 received. Setting client_ignore_invalid_base64 true ignores these errors.
3166 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3167 they are received. Thus, the responses can be made to depend on the
3168 challenges. If an invalid string is ignored, an empty string is placed in
3171 PH/46 Messages that are created by the autoreply transport now contains a
3172 References: header, in accordance with RFCs 2822 and 3834.
3174 PH/47 Added authenticated_sender_force to the smtp transport.
3176 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3178 PH/49 Installed latest patch from the Sieve maintainer.
3180 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3181 was also set, the appendfile transport was unnecessarily scanning a
3182 directory of message files (e.g. for maildir delivery) to find the count
3183 of files (along with the size), even though it did not need this
3184 information. It now does the scan only if it needs to find either the
3185 size of the count of files.
3187 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3189 PH/52 Two bugs concerned with error handling when the smtp transport is
3192 (i) Exim was not creating retry information for temporary errors given
3193 for individual recipients after the DATA command when the smtp transport
3194 was used in LMTP mode. This meant that they could be retried too
3195 frequently, and not timed out correctly.
3197 (ii) Exim was setting the flag that allows error details to be returned
3198 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3199 recipients that were returned after the DATA command.
3201 PH/53 This is related to PH/52, but is more general: for any failing address,
3202 when detailed error information was permitted to be returned to the
3203 sender, but the error was temporary, then after the final timeout, only
3204 "retry timeout exceeded" was returned. Now it returns the full error as
3205 well as "retry timeout exceeded".
3207 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3208 do this, and (what is worse) MTAs that accept it.
3210 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3211 will now be deprecated.
3213 PH/56 New os.c-cygwin from the Cygwin maintainer.
3215 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3216 formats (previously only available in exim formats via -bpr, -bpru,
3217 and -bpra. Now also available in native and exiqgrep formats)
3219 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3220 with very large, slow to parse queues
3222 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3224 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3226 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3227 responses to authentication challenges, though it was showing the
3228 challenges; (ii) I've removed the CR characters from the debug output for
3231 PH/58 Allow for the insertion of a newline as well as a space when a string
3232 is turned into more than one encoded-word during RFC 2047 encoding. The
3233 Sieve code now uses this.
3235 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3236 data_4xx, lost_connection, tls_required.
3238 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3239 message was being sent as an SMTP response.
3241 PH/61 Add -l and -k options to exicyclog.
3243 PH/62 When verifying, if an address was redirected to one new address, so that
3244 verification continued, and the new address failed or deferred after
3245 having set something in $address_data, the value of $address_data was not
3246 passed back to the ACL. This was different to the case when no
3247 redirection occurred. The value is now passed back in both cases.
3249 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3250 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3251 use login_cap.h, so on its own it isn't the distinguishing feature. The
3252 new name refers directly to the setclassresources() function.
3254 PH/65 Added configuration files for NetBSD3.
3256 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3258 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3259 is preferred over IPv4.
3261 PH/68 The bounce_return_message and bounce_return_body options were not being
3262 honoured for bounces generated during the reception of non-SMTP messages.
3263 In particular, this applied to messages rejected by the ACL. This bug has
3264 been fixed. However, if bounce_return_message is true and bounce_return_
3265 body is false, the headers that are returned for a non-SMTP message
3266 include only those that have been read before the error was detected.
3267 (In the case of an ACL rejection, they have all been read.)
3269 PH/69 The HTML version of the specification is now built in a directory called
3270 spec_html instead of spec.html, because the latter looks like a path with
3271 a MIME-type, and this confuses some software.
3273 PH/70 Catch two compiler warnings in sieve.c.
3275 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3276 function verify_get_ident() calls ip_connect() to connect a socket, but
3277 if the "connect()" function timed out, ip_connect() used to close the
3278 socket. However, verify_get_ident() also closes the socket later, and in
3279 between Exim writes to the log, which may get opened at this point. When
3280 the socket was closed in ip_connect(), the log could get the same file
3281 descriptor number as the socket. This naturally causes chaos. The fix is
3282 not to close the socket in ip_connect(); the socket should be closed by
3283 the function that creates it. There was only one place in the code where
3284 this was missing, in the iplookup router, which I don't think anybody now
3285 uses, but I've fixed it anyway.
3287 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3288 well as to direct DNS lookups. Otherwise the handling of names in host
3289 lists is inconsistent and therefore confusing.
3295 PH/01 Two changes to the default runtime configuration:
3297 (1) Move the checks for relay_from_hosts and authenticated clients from
3298 after to before the (commented out) DNS black list checks.
3300 (2) Add control=submission to the relay_from_hosts and authenticated
3301 clients checks, on the grounds that messages accepted by these
3302 statements are most likely to be submissions.
3304 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3306 (1) Generate an error if the third argument for the ${prvs expansion is
3309 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3312 (3) Reset the variables that are obtained from the first argument of
3313 ${prvscheck and used in the second argument before leaving the code,
3314 because their memory is reclaimed, so using them afterwards may do
3317 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3318 one (it's much easier than Tom thought :-).
3320 (5) Because of (4), we can now allow for the use of $prvscheck_result
3321 inside the third argument.
3323 PH/03 For some reason, the default setting of PATH when running a command from
3324 a pipe transport was just "/usr/bin". I have changed it to
3327 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3328 anything to be listed in the output from -bV.
3330 PH/05 When a filter generated an autoreply, the entire To: header line was
3331 quoted in the delivery log line, like this:
3333 => >A.N.Other <ano@some.domain> <original@ddress> ...
3335 This has been changed so that it extracts the operative address. There
3336 may be more than one such address. If so, they are comma-separated, like
3339 => >ano@some.domain,ona@other.domain <original@ddress> ...
3341 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3342 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3343 not being looked up in the rDNS to get a host name, Exim was showing the
3344 IP address twice in Received: lines, even though the IP addresses were
3345 identical. For example:
3347 Received: from [1.2.3.4] (helo=[1.2.3.4])
3349 However, if the real host name was known, it was omitting the HELO data
3350 if it matched the actual IP address. This has been tidied up so that it
3351 doesn't show the same IP address twice.
3353 PH/07 When both +timestamp and +memory debugging was on, the value given by
3354 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3355 called by the debug printing, thereby overwriting the timestamp buffer.
3356 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3358 PH/08 When the original message was included in an autoreply transport, it
3359 always said "this is a copy of the message, including all the headers",
3360 even if body_only or headers_only was set. It now gives an appropriate
3363 PH/09 Applied a patch from the Sieve maintainer which:
3365 o fixes some comments
3366 o adds the (disabled) notify extension core
3367 o adds some debug output for the result of if/elsif tests
3368 o points to the current vacation draft in the documentation
3369 and documents the missing references header update
3373 o fixes a bug in processing the envelope test (when testing
3374 multiple envelope elements, the last element determined the
3377 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3378 Electronic Mail") by including:
3380 Auto-submitted: auto-generated
3382 in the messages that it generates (bounce messages and others, such as
3383 warnings). In the case of bounce messages for non-SMTP messages, there was
3384 also a typo: it was using "Auto_submitted" (underscore instead of
3385 hyphen). Since every message generated by Exim is necessarily in response
3386 to another message, thes have all been changed to:
3388 Auto-Submitted: auto-replied
3390 in accordance with these statements in the RFC:
3392 The auto-replied keyword:
3394 - SHOULD be used on messages sent in direct response to another
3395 message by an automatic process,
3397 - MUST NOT be used on manually-generated messages,
3399 - MAY be used on Delivery Status Notifications (DSNs) and Message
3400 Disposition Notifications (MDNs),
3402 - MUST NOT be used on messages generated by automatic or periodic
3403 processes, except for messages which are automatic responses to
3406 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3407 to the default Received: header definition.
3409 PH/12 Added log selector acl_warn_skipped (default on).
3411 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3412 variables because (a) they are in the wrong storage pool and (b) even if
3413 they were copied, it wouldn't work properly because of the caching.
3415 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3416 checking when decoding. Apparently there are clients that generate
3417 overlong encoded strings. Why am I not surprised?
3419 PH/15 If the first argument of "${if match_address" was not empty, but did not
3420 contain an "@" character, Exim crashed. Now it writes a panic log message
3421 and treats the condition as false.
3423 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3425 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3426 extension "envelope-auth". The code is finished and in agreement with
3427 other implementations, but there is no documentation so far and in fact,
3428 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3429 not changing the active code.
3431 Print executed "if" and "elsif" statements when debugging is used. This
3432 helps a great deal to understand what a filter does.
3434 Document more things not specified clearly in RFC3028. I had all this
3435 sorted out, when out of a sudden new issues came to my mind. Oops."
3437 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3440 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3441 canonical form (as documented). However, after a host name lookup from
3442 the IP address, check_host() was doing a simple string comparison with
3443 addresses acquired from the DNS when checking that the found name did
3444 have the original IP as one of its addresses. Since any found IPv6
3445 addresses are likely to be in abbreviated form, the comparison could
3446 fail. Luckily, there already exists a function for doing the comparison
3447 by converting both addresses to binary, so now that is used instead of
3448 the text comparison.
3450 PH/20 There was another similar case to PH/19, when a complete host name was
3451 given in a host list; looking up its IP address could give an abbreviated
3452 form, whereas the current host's name might or might not be abbreviated.
3453 The same fix has been applied.
3459 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3460 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3463 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3464 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3466 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3468 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3469 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3470 Also, for "Auto-Submitted", treat anything other than "no" as
3471 non-personal, in accordance with RFC 3834. (Previously it treated
3472 anything starting "auto-" as non-personal.)
3474 TF/01 The control=submission/name=... option had a problem with syntax
3475 errors if the name included a slash character. The /name= option
3476 now slurps the rest of the string, so it can include any characters
3477 but it must come last in the list of options (after /sender_retain
3480 PH/05 Some modifications to the interface to the fake nameserver for the new
3488 TK/01 Added the "success_on_redirect" address verification option. See
3489 NewStuff for rationale and an example.
3491 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3493 PH/02 Patch to exigrep to allow it to work on syslog lines.
3495 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3496 fread() to skip over the body file's header line, because in Cygwin the
3497 header line is locked and is inaccessible.
3499 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3500 co-exist for some time) to make it clear that it is the Exim ID that is
3501 referenced, not the Message-ID: header line.
3503 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3504 string_format() function, because snprintf() does not exist on all
3507 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3508 ${stat: expansion item.
3510 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3511 protocol synchronization error", to keep the pedants happy.
3513 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3514 well as for IRIX systems, when gcc is being used. See the host.c source
3517 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3519 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3522 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3523 transport and to the smtp transport in LMTP mode.
3525 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3527 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3528 run a filter in a subprocess. This could lead to confusion in subsequent
3529 lookups in the parent process. There should also be a search_tidyup() at
3530 the end of the subprocess.
3532 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3533 only if the host matched helo_try_verify_hosts, which caused the
3534 verification to occur when the EHLO/HELO command was issued. The ACL just
3535 tested the remembered result. Now, if a previous verification attempt has
3536 not happened, "verify = helo" does it there and then.
3538 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3540 TK/03 Fix log output including CR from clamd.
3542 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3543 exist provoked a memory error which could cause a segfault.
3545 PH/15 Installed PCRE 6.2
3547 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3549 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3550 of the problem. Specifically, suggested +O2 rather than +O1 for the
3553 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3555 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3556 record, the port was not correctly logged when the outgoing_port log
3557 selector was set (it logged the transort's default port).
3559 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3560 fallback_hosts, and "hosts" in the smtp transport.
3562 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3563 host errors such as "Connection refused".
3565 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3566 authentication with radiusclient 0.4.9:
3568 - Error returned from rc_read_config was caught wrongly
3569 - Username/password not passed on to radius server due to wrong length.
3571 The presumption is that some radiusclient API changes for 4.51/PH/17
3572 were not taken care of correctly. The code is still untested by me (my
3573 Linux distribution still has 0.3.2 of radiusclient), but it was
3574 contributed by a Radius user.
3576 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3577 expanding the "port" option of the smtp transport.
3579 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3580 while reading a MIME header. Thanks to Tom Hughes for a patch.
3582 PH/24 Include config.h inside local_scan.h so that configuration settings are
3585 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3586 that in an ACL for RCPT (for example), you can examine exactly what was
3589 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3590 commands, but it was not correctly comparing the address with the actual
3591 client host address. Thus, it would show the EHLO address in Received:
3592 header lines when this was not necessary.
3594 PH/27 Added the % operator to ${eval:}.
3596 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3597 it should be ignoring failures (because with -C, for example, it has lost
3598 privilege). It wasn't ignoring creation failures other than "already
3601 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3604 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3605 because an input line was too long, either on its own, or by virtue of
3606 too many continuations, the temporary file was not being removed, and the
3607 return code was incorrect.
3609 PH/31 Missing "BOOL" in function definition in filtertest.c.
3611 PH/32 Applied Sieve patches from the maintainer.
3613 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3615 PH/33 Added "verify = not_blind".
3617 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3618 Local/Makefile (with some defaults set). These are used in built scripts
3619 such as exicyclog, but they have never been used in the exim_install
3620 script (though there are many overriding facilities there). I have
3621 arranged that the exim_install script now takes note of these two
3624 PH/35 Installed configuration files for Dragonfly.
3626 PH/36 When a locally submitted message by a trusted user did not contain a
3627 From: header, and the sender address was obtained from -f or from an SMTP
3628 MAIL command, and the trusted user did not use -F to supply a sender
3629 name, $originator_name was incorrectly used when constructing a From:
3630 header. Furthermore, $originator_name was used for submission mode
3631 messages from external hosts without From: headers in a similar way,
3632 which is clearly wrong.
3634 PH/37 Added control=suppress_local_fixups.
3636 PH/38 When log_selector = +received_sender was set, and the addition of the
3637 sender made the log line's construction buffer exactly full, or one byte
3638 less than full, an overflow happened when the terminating "\n" was
3641 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3642 when the result of a list match is failure because a DNS lookup failed.
3644 PH/40 RM_COMMAND is now used in the building process.
3646 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3647 the "build-* directories that it finds.
3649 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3650 address in a domain literal was a prefix of an interface address.
3652 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3653 when verifying a sender address, unless rewrite_headers is false.
3655 PH/44 Wrote a long comment about why errors_to addresses are verified as
3656 recipients, not senders.
3658 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3659 the ratelimit ACL was added.
3661 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3663 PH/46 Added extra information about PostgreSQL errors to the error string.
3665 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3666 avoiding the need to install special zones in a real server. This is
3667 backwards compatible; if it can't find the fake resolver, it drops back.
3668 Thus, both old and new test suites can be run.
3670 TF/02 Added util/ratelimit.pl
3672 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3673 clock is set back in time.
3675 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3676 Candler <B.Candler@pobox.com>.
3678 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3679 OK for addresses that are the result of redirections.
3681 PH/48 A number of further additions for the benefit of the new test suite,
3682 including a fake gethostbyname() that interfaces to the fake DNS resolver
3685 TF/06 The fix for widen_domains has also been applied to qualify_single and
3686 search_parents which are the other dnslookup options that can cause
3689 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3692 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3694 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3695 service" if it encountered an MX record with an empty target hostname.
3696 The message is now "an MX or SRV record indicated no SMTP service".
3698 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3699 if the DNS of the sending site is misconfigured. This is quite a
3700 common situation. This change restores the behaviour of treating a
3701 helo verification defer as a failure.
3703 PH/51 If self=fail was set on a router, the bounce message did not include the
3704 actual error message.
3710 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3712 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3713 command itself ended in error, the underlying message about the transport
3714 filter timeout was being overwritten with the pipe command error. Now the
3715 underlying error message should be appended to the second error message.
3717 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3719 PH/02 Reduce the amount of output that "make" produces by default. Full output
3720 can still be requested.
3722 PH/03 The warning log line about a condition test deferring for a "warn" verb
3723 was being output only once per connection, rather than after each
3724 occurrence (because it was using the same function as for successful
3725 "warn" verbs). This seems wrong, so I have changed it.
3727 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3728 it should not have, which might have caused a crash in the right
3729 circumstances, but probably never did.
3731 PH/04 Installed a modified version of Tony Finch's patch to make submission
3732 mode fix the return path as well as the Sender: header line, and to
3733 add a /name= option so that you can make the user's friendly name appear
3736 TF/03 Added the control = fakedefer ACL modifier.
3738 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3739 Mark Lowes for thorough testing.
3741 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3743 TK/03 Merged latest SRS patch from Miles Wilton.
3745 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3746 with the definition in sysexits.h (which is #included earlier).
3747 Fortunately, Exim does not actually use EX_OK. The code used to try to
3748 preserve the sysexits.h value, by assuming that macro definitions were
3749 scanned for macro replacements. I have been disabused of this notion,
3750 so now the code just undefines EX_OK before #including unistd.h.
3752 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3753 in the smtp transport. When a block could not be written in a single
3754 write() function, the timeout was being re-applied to each part-write.
3755 This seems wrong - if the receiver was accepting one byte at a time it
3756 would take for ever. The timeout is now adjusted when this happens. It
3757 doesn't have to be particularly precise.
3759 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3760 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3762 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3763 without a domain if the check to <postmaster@domain> fails.
3765 SC/01 Eximstats: added -xls and the ability to specify output files
3766 (patch written by Frank Heydlauf).
3768 SC/02 Eximstats: use FileHandles for outputting results.
3770 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3772 SC/04 Eximstats: fixed display of large numbers with -nvr option
3774 SC/05 Eximstats: fixed merging of reports with empty tables.
3776 SC/06 Eximstats: added the -include_original_destination flag
3778 SC/07 Eximstats: removed tabs and trailing whitespace.
3780 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3782 TK/06 MBOX spool code: Add real "From " MBOX separator line
3783 so the .eml file is really in mbox format (even though
3784 most programs do not really care). Patch from Alex Miller.
3786 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3787 The latter is generated from $received_to and is only set if the
3788 message has one envelope recipient. SA can use these headers,
3789 obviously out-of-the-box. Patch from Alex Miller.
3791 PH/08 The ${def test on a variable was returning false if the variable's
3792 value was "0", contrary to what the specification has always said!
3793 The result should be true unless the variable is empty.
3795 PH/09 The syntax error of a character other than { following "${if
3796 def:variable_name" (after optional whitespace) was not being diagnosed.
3797 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3798 accidental colon was present, for example, could give incorrect results.
3800 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3801 result is used (not including appendfile, where other changes are about
3804 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3805 This involved changing a lot of size variables from int to off_t. It
3806 should work with maildirs and everything.
3808 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3809 spamd dying while we are connected to it.
3811 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3814 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3815 function for BDB 4.3.
3817 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3819 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3820 the sources for PCRE have been re-arranged and more files are now
3823 PH/15 The code I had for printing potentially long long variables in PH/11
3824 above was not the best (it lost precision). The length of off_t variables
3825 is now inspected at build time, and an appropriate printing format (%ld
3826 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3827 to be "long long int" or "long int". This is needed for the internal
3828 formatting function string_vformat().
3830 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3831 the configuration file to be ":syslog", then the script "guesses" where
3832 the logs files are, rather than using the compiled in default. In our
3833 case the guess is not the same as the compiled default, so the script
3834 suddenly stopped working when I started to use syslog. The patch checks
3835 to see if log_file_path is "". If so, it attempts to read it from exim
3836 with no configuration file to get the compiled in version, before it
3837 falls back to the previous guessing code."
3839 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3840 implementing BATV in an Exim configuration. See NewStuff for the gory
3843 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3844 Makefile that are specific to HP-UX.
3846 PH/18 If the "use_postmaster" option was set for a recipient callout together
3847 with the "random" option, the postmaster address was used as the MAIL
3848 FROM address for the random test, but not for the subsequent recipient
3849 test. It is now used for both.
3851 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3852 patch removes a few documentation additions to RFC 3028, because the
3853 latest draft now contains them. It adds the new en;ascii-case comparator
3854 and a new error check for 8bit text in MIME parts. Comparator and
3855 require names are now matched exactly. I enabled the subaddress
3856 extension, but it is not well tested yet (read: it works for me)."
3858 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3859 rework some of the code of TK/09 above to avoid the hardwired use of
3860 "%lld" and "long long". Replaced the call to snprintf() with a call to
3863 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3864 records point to non-existent hosts", "retry timeout exceeded", and
3865 "retry time not reached for any host after a long failure period".
3867 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3868 experimental DomainKeys support:
3870 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3871 (2) On an error such as an illegally used "control", the wrong name for
3872 the control was given.
3874 These problems did NOT occur unless DomainKeys support was compiled.
3876 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3878 PH/24 Added ${if match_ip condition.
3880 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3881 databases so that it will be absolutely obvious if a crash occurs in the
3882 DB library. This is a regular occurrence (often caused by mis-matched
3885 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3886 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3887 picked up on a user's system that detects such things. There doesn't seem
3888 to be a gcc warning option for this - only an attribute that has to be
3889 put on the function's prototype. It seems that in Fedora Core 4 they have
3890 set this on a number of new functions. No doubt there will be more in due
3893 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3894 specify a transport. However, if an address that was verified by such a
3895 router was the subject of a callout, Exim crashed because it tried to
3896 read the rcpt_include_affixes from the non-existent transport. Now it
3897 just assumes that the setting of that option is false. This bug was
3898 introduced by 4.51/PH/31.
3900 PH/28 Changed -d+all to exclude +memory, because that information is very
3901 rarely of interest, but it makes the output a lot bigger. People tend to
3902 do -d+all out of habit.
3904 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3905 code in os-type was giving problems when libc.so lives in lib64, like on
3908 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3909 aren't the modern standard, and it seems that some systems' include files
3910 don't always have them. Exim was already checking for some of the newer
3911 ones like T_AAAA, and defining it itself. I've added checks for all the
3912 record types that Exim uses.
3914 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3915 not automatically generating a new one, as it is supposed to. This
3916 prevented TLS from working. If the file did exist, but contained invalid
3917 data, a new version was generated, as expected. It was only the case of a
3918 non-existent file that was broken.
3920 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3921 with a change in libdomainkeys > 0.64.
3923 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3924 from DNS. If the selector record carries the flag, it now has
3925 precedence over the domain-wide flag.
3927 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3929 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3930 the use of a port name that isn't defined in /etc/services) occurred, the
3931 message was deferred as in a normal delivery, and thus remained on the
3932 spool, instead of being failed because of the mua_wrapper setting. This
3933 is now fixed, and I tidied up some of the mua_wrapper messages at the
3936 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3937 the messages in a hash of arrays rather than using individual hashes.
3938 This is a bit cleaner and results in dramatic memory savings, albeit
3939 at a slight CPU cost.
3941 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3942 as requested by Marc Sherman.
3944 SC/10 Eximstats: added histograms for user specified patterns as requested
3947 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3949 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3950 fopen() in the content-scanning modules that did not already have it.
3956 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3957 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3959 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3961 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3963 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3964 to test Sieve filters that use "vacation".
3966 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3967 that changes the way the GnuTLS parameters are stored in the cache file.
3968 The new format can be generated externally. For backward compatibility,
3969 if the data in the cache doesn't make sense, Exim assumes it has read an
3970 old-format file, and it generates new data and writes a new file. This
3971 means that you can't go back to an older release without removing the
3974 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3975 work if there are any delivery delays because "one_time" forces the
3976 parent to be marked "delivered", so its unseen clone is never tried
3977 again. For this reason, Exim now forbids the simultaneous setting of
3980 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3981 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3982 entry if you want to know the details. The fix, however, neglected to
3983 consider the case when local delivery batching is involved. The test for
3984 "previously delivered" was not happening when checking to see if an
3985 address could be batched with a previous (undelivered) one; under
3986 certain circumstances this could lead to multiple deliveries to the same
3989 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3990 in its include files, and this causes problems building Exim.
3992 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3993 header_syntax) but Exim was just ignoring anything given after a slash.
3994 In particular, this caused confusion with an attempt to use "verify =
3995 reverse_host_lookup/defer_ok". An error is now given when options are
3996 supplied for verify items that do not have them. (Maybe reverse_host_
3997 lookup should have a defer_ok option, but that's a different point.)
3999 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4000 defined by RFC 821) to 2048, because there were problems with some AUTH
4001 commands, and RFC 1869 says the size should be increased for extended
4002 SMTP commands that take arguments.
4004 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4007 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4008 "unknown" error; now it says that the functionality isn't in the binary.
4010 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4011 an address' error message when a string expansion fails (syntax or
4012 whatever). Otherwise the password may appear in the log. Following change
4013 PH/42 below, there is no longer a chance of it appearing in a bounce
4016 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4018 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4019 of its addresses was ever tried. (Bugzilla bug #2.)
4021 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4022 the result incorrectly in the debug output. (It correctly added a newline
4023 to what was transported.)
4025 TF/01 Added $received_time.
4027 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4028 commented out examples of how to interface to a virus scanner and to
4029 SpamAssassin. Also added commented examples of av_scanner and
4030 spamd_address settings.
4032 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4033 and controls are allowed in which ACLs. There were a couple of minor
4034 errors. Some of the entries in the conditions table (which is a table of
4035 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4036 negation of where the condition IS allowed.
4038 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4040 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4041 header file does not have a version number, so I've had to invent a new
4042 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4043 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4044 radiusclient), but it was contributed by a Radius user.
4046 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4047 files or directories, for OpenSSL.
4049 PH/19 When an Exim process that is running as root has to create an Exim log
4050 file, it does so in a subprocess that runs as exim:exim so as to get the
4051 ownership right at creation (otherwise, other Exim processes might see
4052 the file with the wrong ownership). There was no test for failure of this
4053 fork() call, which would lead to the process getting stuck as it waited
4054 for a non-existent subprocess. Forks do occasionally fail when resources
4055 run out. I reviewed all the other calls to fork(); they all seem to check
4058 PH/20 When checking for unexpected SMTP input at connect time (before writing
4059 the banner), Exim was not dealing correctly with a non-positive return
4060 from the read() function. If the client had disconnected by this time,
4061 the result was a log entry for a synchronization error with an empty
4062 string after "input=" when read() returned zero. If read() returned -1
4063 (an event I could not check), uninitialized data bytes were printed.
4064 There were reports of junk text (parts of files, etc) appearing after
4067 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4069 PH/22 Added support for macro redefinition, and (re)definition in between
4070 driver and ACL definitions.
4072 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4073 forgetting to use the resulting value; it was using the unexpanded value.
4075 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4076 hadn't been configured. The fix is from Juergen Kreileder, who
4077 understands it better than I do:
4079 "Here's what I see happening with three configured cyrus_sasl
4080 authenticators configured (plain, login, cram-md5):
4082 On startup auth_cyrus_sasl_init() gets called for each of these.
4083 This means three calls to sasl_listmech() without a specified mech_list.
4084 => SASL tests which mechs of all available mechs actually work
4085 => three warnings about OTP not working
4086 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4088 With the patch, sasl_listmech() also gets called three times. But now
4089 SASL's mech_list option is set to the server_mech specified in the the
4090 authenticator. Or in other words, the answer from sasl_listmech()
4091 gets limited to just the mech you're testing for (which is different
4093 => the return list contains just 'plain' or 'login', 'cram-md5' or
4094 nothing depending on the value of ob->server_mech.
4096 I've just tested the patch: Authentication still works fine,
4097 unavailable mechs specified in the exim configuration are still
4098 caught, and the auth.log warnings about OTP are gone."
4100 PH/25 When debugging is enabled, the contents of the command line are added
4101 to the debugging output, even when log_selector=+arguments is not
4104 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4105 answer is "GNU", and only if the return is "GNU/something" is the answer
4108 PH/27 $acl_verify_message is now set immediately after the failure of a
4109 verification in an ACL, and so is available in subsequent modifiers. In
4110 particular, the message can be preserved by coding like this:
4112 warn !verify = sender
4113 set acl_m0 = $acl_verify_message
4115 Previously, $acl_verify_message was set only while expanding "message"
4116 and "log_message" when a very denied access.
4118 PH/28 Modified OS/os.c-Linux with
4120 -#ifndef OS_LOAD_AVERAGE
4121 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4123 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4124 nomenclature these days.)
4126 PH/29 Installed patch from the Sieve maintainer that adds the options
4127 sieve_useraddress and sieve_subaddress to the redirect router.
4129 PH/30 In these circumstances:
4130 . Two addresses routed to the same list of hosts;
4131 . First host does not offer TLS;
4132 . First host accepts first address;
4133 . First host gives temporary error to second address;
4134 . Second host offers TLS and a TLS session is established;
4135 . Second host accepts second address.
4136 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4137 and peerdn, if requested) that were in fact used only for the second
4140 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4141 attention to any local part prefix or suffix that was matched by the
4142 router that accepted the address. It now behaves in the same way as it
4143 does for delivery: the affixes are removed from the local part unless
4144 rcpt_include_affixes is set on the transport.
4146 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4147 timeout during the DATA phase of an incoming message.
4149 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4150 applied a patch sanctioned by the Sieve maintainer.
4152 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4153 the uid or gid is negative. A case of a negative gid caused this to be
4154 noticed. The fix allows for either to be negative.
4156 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4157 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4160 PH/36 The change PH/12 above was broken. Fixed it.
4162 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4163 the grounds that routing the same address twice would always produce the
4164 same answer. This might have been true once, but it is certainly no
4165 longer true now. Routing a child address may depend on the previous
4166 routing that produced that child. Some complicated redirection strategies
4167 went wrong when messages had multiple recipients, and made Exim's
4168 behaviour dependent on the order in which the addresses were given.
4170 I have moved the duplicate checking until after the routing is complete.
4171 Exim scans the addresses that are assigned to local and remote
4172 transports, and removes any duplicates. This means that more work will be
4173 done, as duplicates will always all be routed, but duplicates are
4174 presumably rare, so I don't expect this is of any significance.
4176 For deliveries to pipes, files, and autoreplies, the duplicate checking
4177 still happens during the routing process, since they are not going to be
4180 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4181 It corrects a timeout issue with spamd. This is Ian's comment: "The
4182 background is that sometimes spamd either never reads data from a
4183 connection it has accepted, or it never writes response data. The exiscan
4184 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4185 blindly assumes that writes won't block so it may never time out."
4187 PH/39 Allow G after quota size as well as K and M.
4189 PH/40 The value set for $authenticated_id in an authenticator may not contain
4190 binary zeroes or newlines because the value is written to log lines and
4191 to spool files. There was no check on this. Now the value is run through
4192 the string_printing() function so that such characters are converted to
4193 printable escape sequences.
4195 PH/41 $message_linecount is a new variable that contains the total number of
4196 lines in the message. Compare $body_linecount, which is the count for the
4199 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4200 bounce and delay warning messages, except in certain special cases, which
4203 (a) An SMTP error message from a remote host;
4204 (b) A message specified in a :fail: redirection;
4205 (c) A message specified in a "fail" command in a system filter;
4206 (d) A message specified in a FAIL return from the queryprogram router;
4207 (e) A message specified by the cannot_route_message router option.
4209 In these cases only, Exim does include the error details in bounce and
4210 warning messages. There are also a few cases where bland messages such
4211 as "unrouteable address" or "local delivery error" are given.
4213 PH/43 $value is now also set for the "else" part of a ${run expansion.
4215 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4216 being worked on, but at least Exim now implements the latest version to
4219 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4220 process to complete was treated as a delivery failure, a timeout while
4221 writing the message to the pipe was logged, but erroneously treated as a
4222 successful delivery. Such timeouts include transport filter timeouts. For
4223 consistency with the overall process timeout, these timeouts are now
4224 treated as errors, giving rise to delivery failures by default. However,
4225 there is now a new Boolean option for the pipe transport called
4226 timeout_defer, which, if set TRUE, converts the failures into defers for
4227 both kinds of timeout. A transport filter timeout is now identified in
4230 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4231 systems where "make" and "gmake" are different, calling "gmake" at top
4232 level broke things. I've arranged for the value of $(MAKE) to be passed
4233 from the Makefile to this script so that it can call the same version of
4237 A note about Exim versions 4.44 and 4.50
4238 ----------------------------------------
4240 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4241 changes of various kinds. As a consequence, a big documentation update was
4242 needed. This delayed the release for rather longer than seemed good, especially
4243 in the light of a couple of (minor) security issues. Therefore, the changes
4244 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4245 release. So 4.44 and 4.50 are in effect two different branches that both start
4248 I have left the 4.50 change log unchanged; it contains all the changes since
4249 4.43. The change log for 4.44 is below; many of its items are identical to
4250 those for 4.50. This seems to be the most sensible way to preserve the
4251 historical information.
4257 1. Minor wording change to the doc/README.SIEVE file.
4259 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4260 computation of the current number of files was incorrect.
4262 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4263 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4266 4. Give more explanation in the error message when the command for a transport
4267 filter fails to execute.
4269 5. There are several places where Exim runs a non-Exim command in a
4270 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4271 was being done only for the command run by the queryprogram router. It is
4272 now done for all such subprocesses. The other cases are: ${run, transport
4273 filters, and the commands run by the lmtp and pipe transports.
4275 6. Added CONFIGURE_GROUP build-time option.
4277 7. Some older OS have a limit of 256 on the maximum number of file
4278 descriptors. Exim was using setrlimit() to set 1000 as a large value
4279 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4280 systems. I've change it so that if it can't get 1000, it tries for 256.
4282 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4283 was an oversight, and furthermore, ever since the addition of extra
4284 controls (e.g. 4.43/32), the checks on when to allow different forms of
4285 "control" were broken. There should now be diagnostics for all cases when a
4286 control that does not make sense is encountered.
4288 9. Added the /retain_sender option to "control=submission".
4290 10. $recipients is now available in the predata ACL (oversight).
4292 11. Tidy the search cache before the fork to do a delivery from a message
4293 received from the command line. Otherwise the child will trigger a lookup
4294 failure and thereby defer the delivery if it tries to use (for example) a
4295 cached ldap connection that the parent has called unbind on.
4297 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4298 of $address_data from the recipient verification was clobbered by the
4299 sender verification.
4301 13. The value of address_data from a sender verification is now available in
4302 $sender_address_data in subsequent conditions in the ACL statement.
4304 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4306 15. Added a new option "connect=<time>" to callout options, to set a different
4309 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4310 was its contents. (It was OK if the option was not defined at all.)
4312 17. A "Completed" log line is now written for messages that are removed from
4313 the spool by the -Mrm option.
4315 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4316 information about exactly what failed.
4318 19. Added -dd to debug only the daemon process.
4320 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4321 handles timeouts, both on the server side and network timeouts. Renamed the
4322 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4324 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4325 It is now set to "smtps".
4327 22. $host_address is now set to the target address during the checking of
4328 ignore_target_hosts.
4330 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4331 being passed; this would have caused $sender_host_name to have been used if
4332 matching the list had actually called for a host name (not very likely,
4333 since this list is usually IP addresses). A host name is now passed as
4336 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4337 code that specifies a non-restarting handler (typically sigaction() in
4338 modern systems) in an attempt to fix a rare and obscure crash bug.
4340 25. Narrowed the window for a race in the daemon that could cause it to ignore
4341 SIGCHLD signals. This is not a major problem, because they are used only to
4342 wake it up if nothing else does.
4344 26. A malformed maildirsize file could cause Exim to calculate negative values
4345 for the mailbox size or file count. Odd effects could occur as a result.
4346 The maildirsize information is now recalculated if the size or filecount
4349 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4350 support for a long time. Removed HAVE_SYS_VFS_H.
4352 28. Installed the latest version of exipick from John Jetmore.
4354 29. In an address list, if the pattern was not a regular expression, an empty
4355 subject address (from a bounce message) matched only if the pattern was an
4356 empty string. Non-empty patterns were not even tested. This was the wrong
4357 because it is perfectly reasonable to use an empty address as part of a
4358 database query. An empty address is now tested by patterns that are
4359 lookups. However, all the other forms of pattern expect the subject to
4360 contain a local part and a domain, and therefore, for them, an empty
4361 address still always fails if the pattern is not itself empty.
4363 30. Exim went into a mad DNS loop when attempting to do a callout where the
4364 host was specified on an smtp transport, and looking it up yielded more
4365 than one IP address.
4367 31. Re-factored the code for checking spool and log partition space into a
4368 function that finds that data and another that does the check. The former
4369 is then used to implement four new variables: $spool_space, $log_space,
4370 $spool_inodes, and $log_inodes.
4372 32. The RFC2047 encoding function was originally intended for short strings
4373 such as real names; it was not keeping to the 75-character limit for
4374 encoded words that the RFC imposes. It now respects the limit, and
4375 generates multiple encoded words if necessary. To be on the safe side, I
4376 have increased the buffer size for the ${rfc2047: expansion operator from
4379 33. It is now permitted to omit both strings after an "if" condition; if the
4380 condition is true, the result is "true". As before, when the second string
4381 is omitted, a false condition yields an empty string. This makes it less
4382 cumbersome to write custom ACL and router conditions.
4384 34. Failure to deliver a bounce message always caused it to be frozen, even if
4385 there was an errors_to setting on the router. The errors_to setting is now
4388 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4389 canonical form (fully expanded) before being placed in
4390 $sender_host_address.
4392 36. The table in the code that translates DNS record types into text (T_A to
4393 "A" for instance) was missing entries for NS and CNAME. It is just possible
4394 that this could have caused confusion if both these types were looked up
4395 for the same domain, because the text type is used as part of Exim's
4396 per-process caching. But the chance of anyone hitting this buglet seems
4399 37. The dnsdb lookup has been extended in a number of ways.
4401 (1) There is a new type, "zns", which walks up the domain tree until it
4402 finds some nameserver records. It should be used with care.
4404 (2) There is a new type, "mxh", which is like "mx" except that it returns
4405 just the host names, not the priorities.
4407 (3) It is now possible to give a list of domains (or IP addresses) to be
4408 looked up. The behaviour when one of the lookups defers can be
4409 controlled by a keyword.
4411 (4) It is now possible to specify the separator character for use when
4412 multiple records are returned.
4414 38. The dnslists ACL condition has been extended: it is now possible to supply
4415 a list of IP addresses and/or domains to be looked up in a particular DNS
4418 39. Added log_selector=+queue_time_overall.
4420 40. When running the queue in the test harness, wait just a tad after forking a
4421 delivery process, to get repeatability of debugging output.
4423 41. Include certificate and key file names in error message when GnuTLS fails
4424 to set them up, because the GnuTLS error message doesn't include the name
4425 of the failing file when there is a problem reading it.
4427 42. Allow both -bf and -bF in the same test run.
4429 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4431 44. The "Exiscan patch" is now merged into the mainline Exim source.
4433 45. Sometimes the final signoff response after QUIT could fail to get
4434 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4435 < 0 before doing a fflush(). This bug looks as though it goes back to the
4436 introduction of TLS in release 3.20, but "sometimes" must have been rare
4437 because the tests only now provoked it.
4439 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4440 (this can affect the format of dates).
4442 47. exim_tidydb, when checking for the continued existence of a message for
4443 which it has found a message-specific retry record, was not finding
4444 messages that were in split spool directories. Consequently, it was
4445 deleting retry records that should have stayed in existence.
4447 48. Steve fixed some bugs in eximstats.
4449 49. The SPA authentication driver was not abandoning authentication and moving
4450 on to the next authenticator when an expansion was forced to fail,
4451 contradicting the general specification for all authenticators. Instead it
4452 was generating a temporary error. It now behaves as specified.
4454 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4455 (the order specifies the preference for clients). The order is now AES256,
4456 AES128, 3DES, ARCFOUR128.
4458 51. Small patch to Sieve code - explicitly set From: when generating an
4461 52. Exim crashed if a remote delivery caused a very long error message to be
4462 recorded - for instance if somebody sent an entire SpamAssassin report back
4463 as a large number of 550 error lines. This bug was coincidentally fixed by
4464 increasing the size of one of Exim's internal buffers (big_buffer) that
4465 happened as part of the Exiscan merge. However, to be on the safe side, I
4466 have made the code more robust (and fixed the comments that describe what
4469 53. Now that there can be additional text after "Completed" in log lines (if
4470 the queue_time_overall log selector is set), a one-byte patch to exigrep
4471 was needed to allow it to recognize "Completed" as not the last thing in
4474 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4475 patch that reportedly fixes this has been added. I am not expert enough to
4476 create a test for it. This is what the patch creator wrote:
4478 "I found a little strange behaviour of ldap code when working with
4479 Windows 2003 AD Domain, where users was placed in more than one
4480 Organization Units. When I tried to give exim partial DN, the exit code
4481 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4482 But simultaneously result of request was absolutely normal ldap result,
4483 so I produce this patch..."
4485 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4486 REFERENCE, so I have modified the code to exclude the patch when that macro
4489 55. Some experimental protocols are using DNS PTR records for new purposes. The
4490 keys for these records are domain names, not reversed IP addresses. The
4491 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4492 leaves it alone. Component reversal etc. now happens only for IP addresses.
4495 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4497 57. Double the size of the debug message buffer (to 2048) so that more of very
4498 long debug lines gets shown.
4500 58. The exicyclog utility now does better if the number of log files to keep
4501 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4503 59. Two changes related to the smtp_active_hostname option:
4505 (1) $smtp_active_hostname is now available as a variable.
4506 (2) The default for smtp_banner uses $smtp_active_hostname instead
4507 of $primary_hostname.
4509 60. The host_aton() function is supposed to be passed a string that is known
4510 to be a valid IP address. However, in the case of IPv6 addresses, it was
4511 not checking this. This is a hostage to fortune. Exim now panics and dies
4512 if the condition is not met. A case was found where this could be provoked
4513 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4514 components; fortuitously, this particular loophole had already been fixed
4515 by change 4.50/55 above.
4517 If there are any other similar loopholes, the new check in host_aton()
4518 itself should stop them being exploited. The report I received stated that
4519 data on the command line could provoke the exploit when Exim was running as
4520 exim, but did not say which command line option was involved. All I could
4521 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4522 running as the user.
4525 61. There was a buffer overflow vulnerability in the SPA authentication code
4526 (which came originally from the Samba project). I have added a test to the
4527 spa_base64_to_bits() function which I hope fixes it.
4530 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4531 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4533 63. The daemon start-up calls getloadavg() while still root for those OS that
4534 need the first call to be done as root, but it missed one case: when
4535 deliver_queue_load_max is set with deliver_drop_privilege. This is
4536 necessary for the benefit of the queue runner, because there is no re-exec
4537 when deliver_drop_privilege is set.
4539 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4540 This has been fixed.
4542 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4543 was in use, was not putting the data itself into the right store pool;
4544 consequently, it could be overwritten for a subsequent message in the same
4545 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4548 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4550 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4551 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4552 the function were treating the return as a boolean value, which happened to
4553 work because 0=false and not-0=true, but is not correct code.
4555 68. The host_aton() function was not handling scoped IPv6 addresses (those
4556 with, for example, "%eth0" on the end) correctly.
4558 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4559 negated items (that is, ~something) in unsigned ints. Some compilers
4560 apparently mutter when there is no cast.
4562 70. If an address verification called from an ACL failed, and did not produce a
4563 user-specific message (i.e. there was only a "system" message), nothing was
4564 put in $acl_verify_message. In this situation, it now puts the system
4567 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4568 session; change 4.31/43 added the unwanted input to the log line - except
4569 that it did not do this in the start of session case. It now does.
4571 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4572 This is wrong and can cause the other end to generate a synchronization
4573 error if it is another Exim or anything else that does the synchronization
4574 check. A QUIT command is no longer sent after a timeout.
4576 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4577 during host lookups.
4579 74. The defer_ok option of callout verification was not working if it was used
4580 when verifying addresses in header lines, that is, for this case:
4582 verify = header_sender/callout=defer_ok
4584 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4585 those file descriptors could be used for SMTP connections. If anything
4586 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4587 could be sent to the SMTP client, causing chaos. The daemon now opens
4588 stdin, stdout, and stderr to /dev/null when it puts itself into the
4591 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4592 log by default. The user can override this with suitable Perl magic.
4594 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4595 the log message when discard triggers, was not working for the DATA ACL or
4596 for the non-SMTP ACL.
4598 78. Error message wording change in sieve.c.
4600 79. If smtp_accept_max_per_host was set, the number of connections could be
4601 restricted to fewer than expected, because the daemon was trying to set up
4602 a new connection before checking whether the processes handling previous
4603 connections had finished. The check for completed processes is now done
4604 earlier. On busy systems, this bug wouldn't be noticed because something
4605 else would have woken the daemon, and it would have reaped the completed
4608 80. If a message was submitted locally by a user whose login name contained one
4609 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4610 It caused a spool format error. I have fixed the spool reading code. A
4611 related problem was that the "from" clause in the Received: line became
4612 illegal because of the space(s). It is now covered by ${quote_local_part.
4614 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4617 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4619 83. Patch from Sieve maintainer to fix unterminated string problem in
4620 "vacation" handling.
4622 84. Some minor changes to the Linux configuration files to help with other
4623 OS variants using glibc.
4625 85. One more patch for Sieve to update vacation handling to latest spec.
4628 ----------------------------------------------------
4629 See the note above about the 4.44 and 4.50 releases.
4630 ----------------------------------------------------
4636 1. Change 4.43/35 introduced a bug that caused file counts to be
4637 incorrectly computed when quota_filecount was set in an appendfile
4640 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4641 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4644 3. Give more explanation in the error message when the command for a transport
4645 filter fails to execute.
4647 4. There are several places where Exim runs a non-Exim command in a
4648 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4649 was being done only for the command run by the queryprogram router. It is
4650 now done for all such subprocesses. The other cases are: ${run, transport
4651 filters, and the commands run by the lmtp and pipe transports.
4653 5. Some older OS have a limit of 256 on the maximum number of file
4654 descriptors. Exim was using setrlimit() to set 1000 as a large value
4655 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4656 systems. I've change it so that if it can't get 1000, it tries for 256.
4658 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4659 was an oversight, and furthermore, ever since the addition of extra
4660 controls (e.g. 4.43/32), the checks on when to allow different forms of
4661 "control" were broken. There should now be diagnostics for all cases when a
4662 control that does not make sense is encountered.
4664 7. $recipients is now available in the predata ACL (oversight).
4666 8. Tidy the search cache before the fork to do a delivery from a message
4667 received from the command line. Otherwise the child will trigger a lookup
4668 failure and thereby defer the delivery if it tries to use (for example) a
4669 cached ldap connection that the parent has called unbind on.
4671 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4672 of $address_data from the recipient verification was clobbered by the
4673 sender verification.
4675 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4676 was its contents. (It was OK if the option was not defined at all.)
4678 11. A "Completed" log line is now written for messages that are removed from
4679 the spool by the -Mrm option.
4681 12. $host_address is now set to the target address during the checking of
4682 ignore_target_hosts.
4684 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4685 being passed; this would have caused $sender_host_name to have been used if
4686 matching the list had actually called for a host name (not very likely,
4687 since this list is usually IP addresses). A host name is now passed as
4690 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4691 code that specifies a non-restarting handler (typically sigaction() in
4692 modern systems) in an attempt to fix a rare and obscure crash bug.
4694 15. Narrowed the window for a race in the daemon that could cause it to ignore
4695 SIGCHLD signals. This is not a major problem, because they are used only to
4696 wake it up if nothing else does.
4698 16. A malformed maildirsize file could cause Exim to calculate negative values
4699 for the mailbox size or file count. Odd effects could occur as a result.
4700 The maildirsize information is now recalculated if the size or filecount
4703 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4704 support for a long time. Removed HAVE_SYS_VFS_H.
4706 18. Updated exipick to current release from John Jetmore.
4708 19. Allow an empty sender to be matched against a lookup in an address list.
4709 Previously the only cases considered were a regular expression, or an
4712 20. Exim went into a mad DNS lookup loop when doing a callout where the
4713 host was specified on the transport, if the DNS lookup yielded more than
4716 21. The RFC2047 encoding function was originally intended for short strings
4717 such as real names; it was not keeping to the 75-character limit for
4718 encoded words that the RFC imposes. It now respects the limit, and
4719 generates multiple encoded words if necessary. To be on the safe side, I
4720 have increased the buffer size for the ${rfc2047: expansion operator from
4723 22. Failure to deliver a bounce message always caused it to be frozen, even if
4724 there was an errors_to setting on the router. The errors_to setting is now
4727 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4728 canonical form (fully expanded) before being placed in
4729 $sender_host_address.
4731 24. Updated eximstats to version 1.33
4733 25. Include certificate and key file names in error message when GnuTLS fails
4734 to set them up, because the GnuTLS error message doesn't include the name
4735 of the failing file when there is a problem reading it.
4737 26. Expand error message when OpenSSL has problems setting up cert/key files.
4740 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4741 (this can affect the format of dates).
4743 28. exim_tidydb, when checking for the continued existence of a message for
4744 which it has found a message-specific retry record, was not finding
4745 messages that were in split spool directories. Consequently, it was
4746 deleting retry records that should have stayed in existence.
4748 29. eximstats updated to version 1.35
4749 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4750 1.35 - bugfix such that pie charts by volume are generated correctly
4752 30. The SPA authentication driver was not abandoning authentication and moving
4753 on to the next authenticator when an expansion was forced to fail,
4754 contradicting the general specification for all authenticators. Instead it
4755 was generating a temporary error. It now behaves as specified.
4757 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4758 (the order specifies the preference for clients). The order is now AES256,
4759 AES128, 3DES, ARCFOUR128.
4761 31. Small patch to Sieve code - explicitly set From: when generating an
4764 32. Exim crashed if a remote delivery caused a very long error message to be
4765 recorded - for instance if somebody sent an entire SpamAssassin report back
4766 as a large number of 550 error lines. This bug was coincidentally fixed by
4767 increasing the size of one of Exim's internal buffers (big_buffer) that
4768 happened as part of the Exiscan merge. However, to be on the safe side, I
4769 have made the code more robust (and fixed the comments that describe what
4772 33. Some experimental protocols are using DNS PTR records for new purposes. The
4773 keys for these records are domain names, not reversed IP addresses. The
4774 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4775 leaves it alone. Component reversal etc. now happens only for IP addresses.
4778 34. The host_aton() function is supposed to be passed a string that is known
4779 to be a valid IP address. However, in the case of IPv6 addresses, it was
4780 not checking this. This is a hostage to fortune. Exim now panics and dies
4781 if the condition is not met. A case was found where this could be provoked
4782 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4783 components; fortuitously, this particular loophole had already been fixed
4784 by change 4.50/55 or 4.44/33 above.
4786 If there are any other similar loopholes, the new check in host_aton()
4787 itself should stop them being exploited. The report I received stated that
4788 data on the command line could provoke the exploit when Exim was running as
4789 exim, but did not say which command line option was involved. All I could
4790 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4791 running as the user.
4794 35. There was a buffer overflow vulnerability in the SPA authentication code
4795 (which came originally from the Samba project). I have added a test to the
4796 spa_base64_to_bits() function which I hope fixes it.
4799 36. The daemon start-up calls getloadavg() while still root for those OS that
4800 need the first call to be done as root, but it missed one case: when
4801 deliver_queue_load_max is set with deliver_drop_privilege. This is
4802 necessary for the benefit of the queue runner, because there is no re-exec
4803 when deliver_drop_privilege is set.
4805 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4806 was in use, was not putting the data itself into the right store pool;
4807 consequently, it could be overwritten for a subsequent message in the same
4808 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4811 38. Sometimes the final signoff response after QUIT could fail to get
4812 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4813 < 0 before doing a fflush(). This bug looks as though it goes back to the
4814 introduction of TLS in release 3.20, but "sometimes" must have been rare
4815 because the tests only now provoked it.
4821 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4822 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4823 Now it returns an integer. A number of calls were still expecting a T/F
4824 return. Fortuitously, in all cases, the tests worked in OK situations,
4825 which is the norm. However, things would have gone wrong on any write
4826 failures on the smtp file descriptor. This function is used when sending
4827 messages over SMTP and also when doing verify callouts.
4829 2. When Exim is called to do synchronous delivery of a locally submitted
4830 message (the -odf or -odi options), it no longer closes stderr before doing
4833 3. Implemented the mua_wrapper option.
4835 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4837 5. Implemented the functions header_remove(), header_testname(),
4838 header_add_at_position(), and receive_remove_recipient(), and exported them
4841 6. If an ACL "warn" statement specified the addition of headers, Exim already
4842 inserted X-ACL-Warn: at the start if there was no header name. However, it
4843 was not making this test for the second and subsequent header lines if
4844 there were newlines in the string. This meant that an invalid header could
4845 be inserted if Exim was badly configured.
4847 7. Allow an ACL "warn" statement to add header lines at the start or after all
4848 the Received: headers, as well as at the end.
4850 8. Added the rcpt_4xx retry error code.
4852 9. Added postmaster_mailfrom=xxx to callout verification option.
4854 10. Added mailfrom=xxxx to the callout verification option, for verify=
4857 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4858 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4860 12. Inserted some casts to stop certain compilers warning when using pointer
4861 differences as field lengths or precisions in printf-type calls (mostly
4862 affecting debugging statements).
4864 13. Added optional readline() support for -be (dynamically loaded).
4866 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4867 same clock tick as a message's arrival, so that its received time was the
4868 same as the "first fail" time on the retry record, and that message
4869 remained on the queue past the ultimate address timeout, every queue runner
4870 would try a delivery (because it was past the ultimate address timeout) but
4871 after another failure, the ultimate address timeout, which should have then
4872 bounced the address, did not kick in. This was a "< instead of <=" error;
4873 in most cases the first failure would have been in the next clock tick
4874 after the received time, and all would be well.
4876 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4877 being recognized when the domain list was tested by the match_domain
4878 condition in an expansion string.
4880 16. Added the ${str2b64: operator.
4882 17. Exim was always calling setrlimit() to set a large limit for the number of
4883 processes, without checking whether the existing limit was already
4884 adequate. (It did check for the limit on file descriptors.) Furthermore,
4885 errors from getrlimit() and setrlimit() were being ignored. Now they are
4886 logged to the main and panic logs, but Exim does carry on, to try to do its
4887 job under whatever limits there are.
4889 18. Imported PCRE 5.0.
4891 19. Trivial typo in log message " temporarily refused connection" (the leading
4894 20. If the log selector return_path_on_delivery was set and an address was
4895 redirected to /dev/null, the delivery process crashed because it assumed
4896 that a return path would always be set for a "successful" delivery. In this
4897 case, the whole delivery is bypassed as an optimization, and therefore no
4900 21. Internal re-arrangement: the function for sending a challenge and reading
4901 a response while authentication was assuming a zero-terminated challenge
4902 string. It's now changed to take a pointer and a length, to allow for
4903 binary data in such strings.
4905 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4907 23. Exim was not respecting finduser_retries when seeking the login of the
4908 uid under which it was called; it was always trying 10 times. (The default
4909 setting of finduser_retries is zero.) Also, it was sleeping after the final
4910 failure, which is pointless.
4912 24. Implemented tls_on_connect_ports.
4914 25. Implemented acl_smtp_predata.
4916 26. If the domain in control=submission is set empty, Exim assumes that the
4917 authenticated id is a complete email address when it generates From: or
4918 Sender: header lines.
4920 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4921 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4922 chown and chgrp in /bin and hostname in /usr/bin.
4924 28. Exim was keeping the "process log" file open after each use, just as it
4925 does for the main log. This opens the possibility of it remaining open for
4926 long periods when the USR1 signal hits a daemon. Occasional processlog
4927 errors were reported, that could have been caused by this. Anyway, it seems
4928 much more sensible not to leave this file open at all, so that is what now
4931 29. The long-running daemon process does not normally write to the log once it
4932 has entered its main loop, and it closes the log before doing so. This is
4933 so that log files can straightforwardly be renamed and moved. However,
4934 there are a couple of unusual error situations where the daemon does write
4935 log entries, and I had neglected to close the log afterwards.
4937 30. The text of an SMTP error response that was received during a remote
4938 delivery was being truncated at 512 bytes. This is too short for some of
4939 the long messages that one sometimes sees. I've increased the limit to
4942 31. It is now possible to make retry rules that apply only when a message has a
4943 specific sender, in particular, an empty sender.
4945 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4946 it possible to be selective about when SMTP synchronization is enforced.
4948 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4950 32. Implemented hosts_connection_nolog.
4952 33. Added an ACL for QUIT.
4954 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4957 35. Added mailbox_size and mailbox_filecount to appendfile.
4959 36. Added control = no_multiline_responses to ACLs.
4961 37. There was a bug in the logic of the code that waits for the clock to tick
4962 in the case where the clock went backwards by a substantial amount such
4963 that the microsecond fraction of "now" was more than the microsecond
4964 fraction of "then" (but the whole seconds number was less).
4966 38. Added support for the libradius Radius client library this is found on
4967 FreeBSD (previously only the radiusclient library was supported).
4973 1. When certain lookups returned multiple values in the form name=value, the
4974 quoting of the values was not always being done properly. Specifically:
4975 (a) If the value started with a double quote, but contained no whitespace,
4977 (b) If the value contained whitespace other than a space character (i.e.
4978 tabs or newlines or carriage returns) it was not quoted.
4979 This fix has been applied to the mysql and pgsql lookups by writing a
4980 separate quoting function and calling it from the lookup code. The fix
4981 should probably also be applied to nisplus, ibase and oracle lookups, but
4982 since I cannot test any of those, I have not disturbed their existing code.
4984 2. A hit in the callout cache for a specific address caused a log line with no
4985 reason for rejecting RCPT. Now it says "Previous (cached) callout
4986 verification failure".
4988 3. There was an off-by-one bug in the queryprogram router. An over-long
4989 return line was truncated at 256 instead of 255 characters, thereby
4990 overflowing its buffer with the terminating zero. As well as fixing this, I
4991 have increased the buffer size to 1024 (and made a note to document this).
4993 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4994 when Exim is waiting for an SMTP response from a remote server, Exim
4995 restarts its select() call on the socket, thereby resetting its timeout.
4996 This is not a problem when such interrupts are rare. Somebody set up a cron
4997 job to run exiwhat every 2 minutes, which is less than the normal select()
4998 timeout (5 or 10 minutes). This meant that the select() timeout never
4999 kicked in because it was always reset. I have fixed this by comparing the
5000 time when an interrupt arrives with the time at the start of the first call
5001 to select(). If more time than the timeout has elapsed, the interrupt is
5002 treated as a timeout.
5004 5. Some internal re-factoring in preparation for the addition of Sieve
5005 extensions (by MH). In particular, the "personal" test is moved to a
5006 separate function, and given an option for scanning Cc: and Bcc: (which is
5007 not set for Exim filters).
5009 6. When Exim created an email address using the login of the caller as the
5010 local part (e.g. when creating a From: or Sender: header line), it was not
5011 quoting the local part when it contained special characters such as @.
5013 7. Installed new OpenBSD configuration files.
5015 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5016 try to make them clearer.
5018 9. Callout options, other than the timeout value, were being ignored when
5019 verifying sender addresses in header lines. For example, when using
5021 verify = header_sender/callout=no_cache
5023 the cache was (incorrectly) being used.
5025 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5026 only the Cygwin environment.
5028 11. When return_path_on_delivery was set as a log selector, if different remote
5029 addresses in the same message used different return paths and parallel
5030 remote delivery occurred, the wrong values would sometimes be logged.
5031 (Whenever a remote delivery process finished, the return path value from
5032 the most recently started remote delivery process was logged.)
5034 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5035 lines when AUTH and/or TLS are in use. This is the "received protocol"
5036 field. Exim used to use "asmtp" for authenticated SMTP, without any
5037 indication (in the protocol name) for TLS use. Now it follows the RFC and
5038 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5039 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5040 names appear in log lines as well as in Received: header lines.
5042 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5043 extensions, and comparison tests, and to fix some bugs.
5045 14. Changes to the "personal" filter test:
5047 (1) The test was buggy in that it was just doing the equivalent of
5048 "contains" tests on header lines. For example, if a user's address was
5049 anne@some.where, the "personal" test would incorrectly be true for
5051 To: susanne@some.where
5053 This test is now done by extracting each address from the header in turn,
5054 and checking the entire address. Other tests that are part of "personal"
5055 are now done using regular expressions (for example, to check local parts
5056 of addresses in From: header lines).
5058 (2) The list of non-personal local parts in From: addresses has been
5059 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5060 taken from the Sieve specification recommendations.
5062 (3) If the message contains any header line starting with "List-" it is
5063 treated as non-personal.
5065 (4) The test for "circular" in the Subject: header line has been removed
5066 because it now seems ill-conceived.
5068 15. Minor typos in src/EDITME comments corrected.
5070 16. Installed latest exipick from John Jetmore.
5072 17. If headers_add on a router specified a text string that was too long for
5073 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5074 of string_sprintf() is now avoided.
5076 18. $message_body_size was not set (it was always zero) when running the DATA
5077 ACL and the local_scan() function.
5079 19. For the "mail" command in an Exim filter, no default was being set for
5080 the once_repeat time, causing a random time value to be used if "once" was
5081 specified. (If the value happened to be <= 0, no repeat happened.) The
5082 default is now 0s, meaning "never repeat". The "vacation" command was OK
5083 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5084 (I found it when inspecting the code).
5086 20. There is now an overall timeout for performing a callout verification. It
5087 defaults to 4 times the callout timeout, which applies to individual SMTP
5088 commands during the callout. The overall timeout applies when there is more
5089 than one host that can be tried. The timeout is checked before trying the
5090 next host. This prevents very long delays if there are a large number of
5091 hosts and all are timing out (e.g. when the network connections are timing
5092 out). The value of the overall timeout can be changed by specifying an
5093 additional sub-option for "callout", called "maxwait". For example:
5095 verify = sender/callout=5s,maxwait=20s
5097 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5098 to the end before writing, but this should make it even safer).
5100 22. Exim was forgetting that it had advertised PIPELINING for the second and
5101 subsequent messages on an SMTP connection. It was also not resetting its
5102 memory on STARTTLS and an internal HELO.
5104 23. When Exim logs an SMTP synchronization error within a session, it now
5105 records whether PIPELINING has been advertised or not.
5107 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5108 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5109 rather than long int.
5111 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5113 26. Added the never_mail option to autoreply.
5119 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5120 crash if the getsockname() call failed; this can happen if a connection is
5121 closed very soon after it is established. The problem was simply in the
5122 order in which certain operations were done, causing Exim to try to write
5123 to the SMTP stream before it had set up the file descriptor. The bug has
5124 been fixed by making things happen in the correct order.
5130 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5131 before the connection was closed, thus losing the rejection response.
5133 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5134 some early Solaris releases, but causes trouble in current releases where
5135 socklen_t is defined.
5137 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5140 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5143 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5144 that releasing the top of it at the end releases what was used for sub-
5145 expansions (unless the block got too big). However, discard this block if
5146 the first thing is a variable or header, so that we can use its block when
5147 it is dynamic (useful for very large $message_headers, for example).
5149 6. Lookups now cache *every* query, not just the most recent. A new, separate
5150 store pool is used for this. It can be recovered when all lookup caches are
5151 flushed. Lookups now release memory at the end of their result strings.
5152 This has involved some general refactoring of the lookup sources.
5154 7. Some code has been added to the store_xxx() functions to reduce the amount
5155 of flapping under certain conditions.
5157 8. log_incoming_interface used to affect only the <= reception log lines. Now
5158 it causes the local interface and port to be added to several more SMTP log
5159 lines, for example "SMTP connection from", and rejection lines.
5161 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5163 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5165 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5166 could be overwritten at the end of the current message (or the start of a
5167 new message if it was set in a HELO ACL). The value is now preserved for
5168 the duration of the SMTP connection.
5170 12. If a transport had a headers_rewrite setting, and a matching header line
5171 contained an unqualified address, that address was qualified, even if it
5172 did not match any rewriting rules. The underlying bug was that the values
5173 of the flags that permit the existence of unqualified sender and recipient
5174 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5175 non-local messages, and by -bnq for local messages) were not being
5176 preserved with the message after it was received.
5178 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5179 "next input=" as part of the text comprising the host identity instead of
5180 the correct text. The code was using the same buffer for two different
5181 strings. However, depending on which order the printing function evaluated
5182 its arguments, the bug did not always show up. Under Linux, for example, my
5183 test suite worked just fine.
5185 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5186 doesn't work with some older versions of Perl. It has been changed to "my",
5187 which in any case is probably the better facility to use.
5189 15. A really picky compiler found some instances of statements for creating
5190 error messages that either had too many or two few arguments for the format
5193 16. The size of the buffer for calls to the DNS resolver has been increased
5194 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5195 for addresses that have a lot of PTR records. This alleviates a problem; it
5196 does not fully solve it.
5198 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5199 buffer now truncates the list and logs the incident, which is the same
5200 action as happens when Exim is looking up a host name and its aliases.
5201 Previously in this situation something unpredictable would happen;
5202 sometimes it was "internal error: store_reset failed".
5204 18. If a server dropped the connection unexpectedly when an Exim client was
5205 using GnuTLS and trying to read a response, the client delivery process
5206 crashed while trying to generate an error log message.
5208 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5209 string, for example:
5211 warn message = H1: something\nH2: something
5213 the text was added as a single header line from Exim's point of view
5214 though it ended up OK in the delivered message. However, searching for the
5215 second and subsequent header lines using $h_h2: did not work. This has been
5216 fixed. Similarly, if a system filter added multiple headers in this way,
5217 the routers could not see them.
5219 20. Expanded the error message when iplsearch is called with an invalid key to
5220 suggest using net-iplsearch in a host list.
5222 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5223 ACLs are no longer actually imposed (and a message to that effect is
5226 22. If a "gecos" field in a passwd entry contained escaped characters, in
5227 particular, if it contained a \" sequence, Exim got it wrong when building
5228 a From: or a Sender: header from that name. A second bug also caused
5229 incorrect handling when an unquoted " was present following a character
5230 that needed quoting.
5232 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5233 was not being matched caselessly.
5235 24. Arranged for all hyphens in the exim.8 source to be escaped with
5238 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5239 properly. Recipient callout cache records were still being keyed to include
5240 the sender, even when use_sender was set false. This led to far more
5241 callouts that were necessary. The sender is no longer included in the key
5242 when use_sender is false.
5244 26. Added "control = submission" modifier to ACLs.
5246 27. Added the ${base62d: operator to decode base 62 numbers.
5248 28. dnsdb lookups can now access SRV records.
5250 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5251 the configuration file.
5253 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5254 (-v) mode. This makes the output for a verbose queue run more intelligible.
5256 31. Added a use_postmaster feature to recipient callouts.
5258 32. Added the $body_zerocount variable, containing the number of binary zero
5259 bytes in the message body.
5261 33. The time of last modification of the "new" subdirectory is now used as the
5262 "mailbox time last read" when there is a quota error for a maildir
5265 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5267 35. Added +ignore_unknown as a special item in host lists.
5269 36. Code for decoding IPv6 addresses in host lists is now included, even if
5270 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5271 address was recognized as an IP address, but was then not correctly decoded
5272 into binary, causing unexpected and incorrect effects when compared with
5279 1. Very minor rewording of debugging text in manualroute to say "list of
5280 hosts" instead of "hostlist".
5282 2. If verify=header_syntax was set, and a header line with an unqualified
5283 address (no domain) and a large number of spaces between the end of the
5284 name and the colon was received, the reception process suffered a buffer
5285 overflow, and (when I tested it) crashed. This was caused by some obsolete
5286 code that should have been removed. The fix is to remove it!
5288 3. When running in the test harness, delay a bit after writing a bounce
5289 message to get a bit more predictability in the log output.
5291 4. Added a call to search_tidyup() just before forking a reception process. In
5292 theory, someone could use a lookup in the expansion of smtp_accept_max_
5293 per_host which, without the tidyup, could leave open a database connection.
5295 5. Added the variables $recipient_data and $sender_data which get set from a
5296 lookup success in an ACL "recipients" or "senders" condition, or a router
5297 "senders" option, similar to $domain_data and $local_part_data.
5299 6. Moved the writing of debug_print from before to after the "senders" test
5302 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5303 problems for message scanning, either using a data ACL, or using
5304 local_scan() because the Received: header was not generated till after they
5305 were called (in order to set the time as the time of reception completion).
5306 I have revised the way this works. The header is now generated after the
5307 body is received, but before the ACL or local_scan() are called. After they
5308 are run, the timestamp in the header is updated.
5314 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5315 before starting a queue runner without re-exec. This happened only when
5316 deliver_drop_privilege was set or when the Exim user was set to root. The
5317 effect of the bug was that timeouts during subsequent deliveries caused
5318 crashes instead of being properly handled. The handler is now left at its
5319 default (and expected) setting.
5321 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5322 message, again when deliver_drop_privilege is set or Exim is run as root.
5323 The bug described in (1) was not present in this case, but the tidying up
5324 of the other signals was missing. I have made the two cases consistent.
5326 3. The ignore_target_hosts setting on a manualroute router was being ignored
5327 for hosts that were looked up using the /MX notation.
5329 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5332 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5333 operated on the sender address. After changing the $sender_address to <>
5334 for the sender address verify, Exim was re-instated it as the original
5335 (before rewriting) address, but remembering that it had rewritten it, so it
5336 wasn't rewriting it again. This bug also had the effect of breaking the
5337 sender address verification caching when the sender address was rewritten.
5339 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5340 This has been changed so that if the ip literal address matches
5341 ignore_target_hosts, the router declines.
5343 7. Added expansion conditions match_domain, match_address, and match_local_
5344 part (NOT match_host).
5346 8. The placeholder for the Received: header didn't have a length field set.
5348 9. Added code to Exim itself and to exim_lock to test for a specific race
5349 condition that could lead to file corruption when using MBX delivery. The
5350 issue is with the lockfile that is created in /tmp. If this file is removed
5351 after a process has opened it but before that process has acquired a lock,
5352 there is the potential for a second process to recreate the file and also
5353 acquire a lock. This could lead to two Exim processes writing to the file
5354 at the same time. The added code performs the same test as UW imapd; it
5355 checks after acquiring the lock that its file descriptor still refers to
5356 the same named file.
5358 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5359 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5362 11. Added the smtp_active_hostname option. If used, this will typically be made
5363 to depend on the incoming interface address. Because $interface_address is
5364 not set up until the daemon has forked a reception process, error responses
5365 that can happen earlier (such as "too many connections") no longer contain
5368 12. If an expansion in a condition on a "warn" statement fails because a lookup
5369 defers, the "warn" statement is abandoned, and the next ACL statement is
5370 processed. Previously this caused the whole ACL to be aborted.
5372 13. Added the iplsearch lookup type.
5374 14. Added ident_timeout as a log selector.
5376 15. Added tls_certificate_verified as a log selector.
5378 16. Added a global option tls_require_ciphers (compare the smtp transport
5379 option of the same name). This controls incoming TLS connections.
5381 17. I finally figured out how to make tls_require_ciphers do a similar thing
5382 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5383 before starting the TLS session.
5385 18. Tabs are now shown as \t in -bP output.
5387 19. If the log selector return_path_on_delivery was set, Exim crashed when
5388 bouncing a message because it had too many Received: header lines.
5390 20. If two routers both had headers_remove settings, and the first one included
5391 a superfluous trailing colon, the final name in the first list and the
5392 first name in the second list were incorrectly joined into one item (with a
5393 colon in the middle).
5399 1. Added -C and -D options to the exinext utility, mainly to make it easier
5400 to include in the automated testing, but these could be helpful when
5401 multiple configurations are in use.
5403 2. The exinext utility was not formatting the output nicely when there was
5404 an alternate port involved in the retry record key, nor when there was a
5405 message id as well (for retries that were specific to a specific message
5406 and a specific host). It was also confused by IPv6 addresses, because of
5407 the additional colons they contain. I have fixed the IPv4 problem, and
5408 patched it up to do a reasonable job for IPv6.
5410 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5411 delivery, the log line now contains "pipelined" if PIPELINING was used.
5413 4. An SMTP transport process used to panic and die if the bind() call to set
5414 an explicit outgoing interface failed. This has been changed; it is now
5415 treated in the same way as a connect() failure.
5417 5. A reference to $sender_host_name in the part of a conditional expansion
5418 that was being skipped was still causing a DNS lookup. This no longer
5421 6. The def: expansion condition was not recognizing references to header lines
5422 that used bh_ and bheader_.
5424 7. Added the _cache feature to named lists.
5426 8. The code for checking quota_filecount in the appendfile transport was
5427 allowing one more file than it should have been.
5429 9. For compatibility with Sendmail, the command line option
5437 and sets the incoming protocol and host name (for trusted callers). The
5438 host name and its colon can be omitted when only the protocol is to be set.
5439 Note the Exim already has two private options, -pd and -ps, that refer to
5440 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5441 "s", but I don't think that's a major issue.
5443 10. A number of refactoring changes to the code, none of which should affect
5446 (a) The number of logging options was getting close to filling up the
5447 32-bit word that was used as a bit map. I have split them into two classes:
5448 those that are passed in the argument to log_write(), and those that are
5449 only ever tested independently outside of that function. These are now in
5450 separate 32-bit words, so there is plenty of room for expansion again.
5451 There is no change in the user interface or the logging behaviour.
5453 (b) When building, for example, log lines, the code previously used a
5454 macro that called string_cat() twice, in order to add two strings. This is
5455 not really sufficiently general. Furthermore, there was one instance where
5456 it was actually wrong because one of the argument was used twice, and in
5457 one call a function was used. (As it happened, calling the function twice
5458 did not affect the overall behaviour.) The macro has been replaced by a
5459 function that can join an arbitrary number of extra strings onto a growing
5462 (c) The code for expansion conditions now uses a table and a binary chop
5463 instead of a serial search (which was left over from when there were very
5464 few conditions). Also, it now recognizes conditions like "pam" even when
5465 the relevant support is not compiled in: a suitably worded error message is
5466 given if an attempt is made to use such a condition.
5468 11. Added ${time_interval:xxxxx}.
5470 12. A bug was causing one of the ddress fields not to be passed back correctly
5471 from remote delivery subprocesses. The field in question was not being
5472 subsequently used, so this caused to problems in practice.
5474 13. Added new log selectors queue_time and deliver_time.
5476 14. Might have fixed a bug in maildirsizefile handling that threw up
5477 "unexpected character" debug warnings, and recalculated the data
5478 unnecessarily. In any case, I expanded the warning message to give more
5481 15. Added the message "Restricted characters in address" to the statements in
5482 the default ACL that block characters like @ and % in local parts.
5484 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5485 Three changes have been made:
5487 (a) There was a serious bug; a negative response to MAIL caused the whole
5488 recipient domain to be cached as invalid, thereby blocking all messages
5489 to all local parts at the same domain, from all senders. This bug has
5490 been fixed. The domain is no longer cached after a negative response to
5491 MAIL if the sender used is not empty.
5493 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5496 (c) A new callout option, "use_sender" has been added for people who want
5497 the modified behaviour.
5503 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5506 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5507 indeed breaks things for older releases.
5509 3. Added additional logging to the case where there is a problem reading data
5510 from a filter that is running in a subprocess using a pipe, in order to
5511 try to track down a specific problem.
5513 4. Testing facility fudge: when running in the test harness and attempting
5514 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5515 getting "No route to host". Convert this to a timeout.
5517 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5520 6. Some OS don't have socklen_t but use size_t instead. This affects the
5521 fifth argument of getsockopt() amongst other things. This is now
5522 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5523 can be set for individual OS. I have set it for SunOS5, OSF1, and
5524 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5525 some earlier ones do not.
5527 7. Change 4.30/15 was not doing the test caselessly.
5529 8. The standard form for an IPv6 address literal was being rejected by address
5530 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5531 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5532 this, as well as the form without the "IPv6" on the front (but only when
5533 address literals are enabled, of course).
5535 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5537 10. Exim crashed if a message with an empty sender address specified by -f
5538 encountered a router with an errors_to setting. This could be provoked only
5539 by a command such as
5543 where an empty string was supplied; "<>" did not hit this bug.
5545 11. Installed PCRE release 4.5.
5547 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5548 remained set. It is now erased.
5550 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5551 times from message ids (which are base 36 rather than the normal 62).
5553 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5554 were being counted as actual protocol errors, and logged if the log
5555 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5556 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5557 and DATA following a set of rejected RCPTs do not count as protocol errors.
5558 In other words, Exim assumes they were pipelined, though this may not
5559 actually be the case. Of course, in all cases the client gets an
5560 appropriate error code.
5562 15. If a lookup fails in an ACL condition, a message about the failure may
5563 be available; it is used if testing the ACL cannot continue, because most
5564 such messages specify what the cause of the deferral is. However, some
5565 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5566 that caused an old message to be retained and used if a later statement
5567 caused a defer, replacing the real cause of the deferral.
5569 16. If an IP address had so many PTR records that the DNS lookup buffer
5570 was not large enough to hold them, Exim could crash while trying to process
5571 the truncated data. It now detects and logs this case.
5573 17. Further to 4.21/58, another change has been made: if (and only if) the
5574 first line of a message (the first header line) ends with CRLF, a bare LF
5575 in a subsequent header line has a space inserted after it, so as not to
5576 terminate the header.
5578 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5579 unnecessarily, used atoi() instead of strtol(), and didn't check the
5580 termination when getting file sizes from file names by regex.
5582 19. Completely re-implemented the support for maildirsize files, in the light
5583 of a number of problems with the previous contributed implementation
5584 (4.30/29). In particular:
5586 . If the quota is zero, the maildirsize file is maintained, but no quota is
5589 . If the maildir directory does not exist, it is created before any attempt
5590 to write a maildirsize file.
5592 . The quota value in the file is just a cache; if the quota is changed in
5593 the transport, the new value overrides.
5595 . A regular expression is available for excluding directories from the
5598 20. The autoreply transport checks the characters in options that define the
5599 message's headers; it allows continued headers, but it was checking with
5600 isspace() after an embedded newline instead of explicitly looking for a
5603 21. If all the "regular" hosts to which an address was routed had passed their
5604 expiry times, and had not reached their retry times, the address was
5605 bounced, even if fallback hosts were defined. Now Exim should go on to try
5608 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5609 equivalent code in the SMTP transport. Some hosts send humungous responses
5610 to HELO/EHLO, more than 1024 it seems.
5612 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5613 gives compiler warnings in some environments. I've now done it "properly",
5616 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5617 (because of problems with the built-in one) was declared to return uschar *
5618 instead of char *, causing compiler failure.
5620 25. Fixed a file descriptor leak when processing alias/forward files.
5622 26. Fixed a minor format string issue in dbfn.c.
5624 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5626 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5627 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5630 29. When a sender address is verified, it is cached, to save repeating the test
5631 when there is more than one recipient in a message. However, when the
5632 verification involves a callout, it is possible for different callout
5633 options to be set for different recipients. It is too complicated to keep
5634 track of this in the cache, so now Exim always runs a verification when a
5635 callout is required, relying on the callout cache for the optimization.
5636 The overhead is duplication of the address routing, but this should not be
5639 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5640 to be verified with callout=postmaster, and the main callout worked but the
5641 postmaster check failed, the verification correctly failed. However, if a
5642 subsequent RCPT command asked for sender verification *without* the
5643 postmaster check, incorrect caching caused this verification also to fail,
5646 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5647 it was not caching the DNS options (qualify_single, search_parents) that
5648 were used when the lookup failed. A subsequent lookup with different
5649 options therefore always gave the same answer, though there were cases
5650 where it should not have. (Example: a "domains = !$mx_any" option on a
5651 dnslookup router: the "domains" option is always processed without any
5652 widening, but the router might have qualify_single set.) Now Exim uses the
5653 cached value only when the same options are set.
5655 32. Added John Jetmore's "exipick" utility to the distribution.
5657 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5658 than a timeout (e.g. a certificate is required, and is not provided), an
5659 Exim server now closes the connection immediately. Previously it waited for
5660 the client to close - but if the client is SSL, it seems that they each
5661 wait for each other, leading to a delay before one of them times out.
5663 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5664 maintained 0.8.x compatibility because I don't think many are using it, and
5665 it is clearly obsolete.
5667 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5670 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5671 client certificate was expired. A simple patch fixes this, though I don't
5672 understand the full logic of why the verify callback is called multiple
5675 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5676 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5677 which causes problems with some clients (such as the Certicom SSL Plus
5678 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5679 disables the coutermeasure allowing Eudora to connect."
5681 38. Exim was not checking that a write() to a log file succeeded. This could
5682 lead to Bad Things if a log got too big, in particular if it hit a file
5683 size limit. Exim now panics and dies if it cannot write to a log file, just
5684 as it does if it cannot open a log file.
5686 39. Modified OS/Makefile-Linux so that it now contains
5688 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5690 The two -D definitions ensure that Exim is compiled with large file
5691 support, which makes it possible to handle log files that are bigger than
5694 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5695 instance) a domain was checked against a named list that involved a lookup,
5696 causing $domain_data to be set, then another domain was checked against the
5697 same list, then the first domain was re-checked, the value of $domain_data
5698 after the final check could be wrong. In particular, if the second check
5699 failed, it could be set empty. This bug probably also applied to
5702 41. The strip_trailing_dot option was not being applied to the address given
5703 with the -f command-line option.
5705 42. The code for reading a message's header from the spool was incrementing
5706 $received_count, but never initializing it. This meant that the value was
5707 incorrect (doubled) while delivering a message in the same process in which
5708 it was received. In the most common configuration of Exim, this never
5709 happens - a fresh exec is done - but it can happen when
5710 deliver_drop_privilege is set.
5712 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5713 it now includes up to 150 characters of the unexpected data in the log
5716 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5717 and building data strings. The size of both of these buffers was 10 000
5718 bytes - far larger than anybody would *ever* want, thought I. Needless to
5719 say, somebody hit the limit. I have increased the maximum line length to
5720 20 000 and the maximum data length of concatenated lines to 100 000. I have
5721 also fixed two bugs, because there was no checking on these buffers. Tsk,
5722 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5723 buffer is too small.
5725 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5726 lsearch lookups. Now it does.
5728 46. When parsing a route_list item in a manualroute router, a fixed-length
5729 buffer was used for the list of hosts. I made this 1024 bytes long,
5730 thinking that nobody would ever have a list of hosts that long. Wrong.
5731 Somebody had a whole pile of complicated expansion conditions, and the
5732 string was silently truncated, leading to an expansion error. It turns out
5733 that it is easier to change to an unlimited length (owing to other changes
5734 that have happened since this code was originally written) than to build
5735 structure for giving a limitation error. The length of the item that
5736 expands into the list of hosts is now unlimited.
5738 47. The lsearch lookup could not handle data where the length of text line was
5739 more than 4095 characters. Such lines were truncated, leading to shortened
5740 data being returned. It should now handle lines of any length.
5742 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5743 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5746 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5747 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5748 no longer happens in comment lines. A list of replacements is now placed
5749 at the head of all of the source files, except those whose only change is
5750 to replace PERL_COMMAND in the very first #! line.
5752 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5753 messages on the queue, with a bottom-up merge sort, using code contributed
5754 by Michael Haardt. This should make operations like -bp somewhat faster on
5755 large queues. It won't affect queue runners, except when queue_run_in_order
5758 51. Installed eximstats 1.31 in the distribution.
5760 52. Added support for SRV lookups to the dnslookup router.
5762 53. If an ACL referred to $message_body or $message_body_end, the value was not
5763 reset for any messages that followed in the same SMTP session.
5765 54. The store-handling optimization for building very long strings was not
5766 differentiating between the different store pools. I don't think this
5767 actually made any difference in practice, but I've tidied it.
5769 55. While running the routers to verify a sender address, $sender_address
5770 was still set to the sender address. This is wrong, because when routing to
5771 send a bounce to the sender, it would be empty. Therefore, I have changed
5772 it so that, while verifying a sender address, $sender_address is set to <>.
5773 (There is no change to what happens when verifying a recipient address.)
5775 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5776 target A or AAAA records (if not already returned) without resetting the
5777 qualify_single or search_parents options of the DNS resolver. These are
5778 inappropriate in this case because the targets of MX and SRV records must
5779 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5780 target that, when qualified, matched something in the local domain. These
5781 two options are now turned off when doing these lookups.
5783 57. It seems that at least some releases of Reiserfs (which does not have the
5784 concept of a fixed number of inodes) returns zero and not -1 for the
5785 number of available inodes. This interacted badly with check_spool_inodes,
5786 which assumed that -1 was the "no such thing" setting. What I have done is
5787 to check that the total number of inodes is greater than zero before doing
5788 the test of how many are available.
5790 58. When a "warn" ACL statement has a log_message modifier, the message is
5791 remembered, and not repeated. This is to avoid a lot of repetition when a
5792 message has many recipients that cause the same warning to be written.
5793 However, Exim was preserving the list of already written lines for an
5794 entire SMTP session, which doesn't seem right. The memory is now reset if a
5795 new message is started.
5797 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5798 debugging output unless log_rewrite was also set.
5800 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5801 of (int)(handle) when we know that handle contains (void *)(-1).
5803 61. The Exim daemon panic-logs an error return when it closes the incoming
5804 connection. However "connection reset by peer" seems to be common, and
5805 isn't really an error worthy of noting specially, so that particular error
5808 62. When Exim is trying to find all the local interfaces, it used to panic and
5809 die if the ioctl to get the interface flags failed. However, it seems that
5810 on at least one OS (Solaris 9) it is possible to have an interface that is
5811 included in the list of interfaces, but for which you get a failure error
5812 for this call. This happens when the interface is not "plumbed" into a
5813 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5814 failure of the "get flags" call assumes that the interface is down.
5816 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5817 makes life easier for people who are doing arithmetic on fields extracted
5818 from dates, where you often get leading zeros that should not be
5819 interpreted as octal.
5821 64. Added qualify_domain to the redirect router, to override the global
5824 65. If a pathologically long header line contained very many addresses (the
5825 report of this problem mentioned 10 000) and each of them was rewritten,
5826 Exim could use up a very large amount of memory. (It kept on making new
5827 copies of the header line as it rewrote, and never released the old ones.)
5828 At the expense of a bit more processing, the header rewriting function has
5829 been changed so that it no longer eats memory in this way.
5831 66. The generation of the Received: header has been moved from the time that a
5832 message starts to be received, to the time that it finishes. The timestamp
5833 in the Received: header should now be very close to that of the <= log
5834 line. There are two side-effects of this change:
5836 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5837 logged header lines no longer include the local Received: line, because
5838 it has not yet been created. The same applies to a copy of the message
5839 that is returned to a non-SMTP sender when a message is rejected.
5841 (b) When a filter file is tested using -bf, no additional Received: header
5842 is added to the test message. After some thought, I decided that this
5845 This change does not affect the value of $received_for. It is still set
5846 after address rewriting, but before local_scan() is called.
5848 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5850 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5851 gave an unhelpful panic error message, and a defer error. I have managed to
5852 change this behaviour so that it now rejects any supplied certificate,
5853 which seems right, as the list of acceptable certificates is empty.
5855 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5856 gave an unhelpful defer error. I have not managed to make this reject any
5857 supplied certificates, but the error message it gives is "no certificate
5858 supplied", which is not helpful.
5860 70. exigrep's output now also includes lines that are not associated with any
5861 message, but which match the given pattern. Implemented by a patch from
5862 Martin Sluka, which also tidied up the Perl a bit.
5864 71. Recipient callout verification, like sender verification, was using <> in
5865 the MAIL FROM command. This isn't really the right thing, since the actual
5866 sender may affect whether the remote host accepts the recipient or not. I
5867 have changed it to use the actual sender in the callout; this means that
5868 the cache record is now keyed on a recipient/sender pair, not just the
5869 recipient address. There doesn't seem to be a real danger of callout loops,
5870 since a callout by the remote host to check the sender would use <>.
5871 [SEE ABOVE: changed after hitting problems.]
5873 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5874 temporary errors. However, in the case of such a code being given after
5875 the end of a data transmission (i.e. after ".") Exim was failing to write
5876 a retry record for the message. (Yes, there was some broken host that was
5877 actually sending 8xx at this point.)
5879 73. An unknown lookup type in a host list could cause Exim to panic-die when
5880 the list was checked. (An example that provoked this was putting <; in the
5881 middle of a list instead of at the start.) If this happened during a DATA
5882 ACL check, a -D file could be left lying around. This kind of configuration
5883 error no longer causes Exim to die; instead it causes a defer error. The
5884 incident is still logged to the main and panic logs.
5886 74. Buglet left over from Exim 3 conversion. The message "too many messages
5887 in one connection" was written to the rejectlog but not the mainlog, except
5888 when address rewriting (yes!) was being logged.
5890 75. Added write_rejectlog option.
5892 76. When a system filter was run not as root (that is, when system_filter_user
5893 was set), the values of the $n variables were not being returned to the
5894 main process; thus, they were not subsequently available in the $sn
5897 77. Added +return_path_on_delivery log selector.
5899 78. A connection timeout was being treated differently from recipients deferred
5900 when testing hosts_max_try with a message that was older than the host's
5901 retry timeout. (The host should not be counted, thus allowing all hosts to
5902 be tried at least once before bouncing.) This may have been the cause of an
5903 occasionally reported bug whereby a message would remain on the queue
5904 longer than the retry timeout, but would be bounced if a delivery was
5905 forced. I say "may" because I never totally pinned down the problem;
5906 setting up timeout/retry tests is difficult. See also the next item.
5908 79. The ultimate address timeout was not being applied to errors that involved
5909 a combination of host plus message (for example, a timeout on a MAIL
5910 command). When an address resolved to a number of possible hosts, and they
5911 were not all tried for each delivery (e.g. because of hosts_max_try), a
5912 message could remain on the queue longer than the retry timeout.
5914 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5917 81. Fixed an obscure SMTP outgoing bug which required at least the following
5918 conditions: (a) there was another message waiting for the same server;
5919 (b) the server returned 5xx to all RCPT commands in the first message so
5920 that the message was not completed; (c) the server dropped the connection
5921 or gave a negative response to the RSET that Exim sends to abort the
5922 transaction. The observed case was a dropped connection after DATA that had
5923 been sent in pipelining mode. That is, the server had advertised PIPELINING
5924 but was not implementing it correctly. The effect of the bug was incorrect
5925 behaviour, such as trying another host, and this could lead to a crash.
5931 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5932 and daemon.c were passed as pointers to ints; they should have been
5933 pointers to socklen_t variables (which are typically unsigned ints).
5935 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5938 3. Fixed a really odd bug that affected only the testing scheme; patching a
5939 certain fixed string in the binary changed the value of another string that
5940 happened to be identical to the end of the original first string.
5942 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5943 name", it returns that address as the IP address. On some operating
5944 systems (e.g. Solaris), it also passes back the IP address string as the
5945 "host name". However, on others (e.g. Linux), it passes back an empty
5946 string. Exim wasn't checking for this, and was changing the host name to an
5947 empty string, assuming it had been canonicalized.
5949 5. Although rare, it is permitted to have more than one PTR record for a given
5950 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5951 all the names associated with an address, because they do in Solaris.
5952 However, it seems that they do not in Linux for data that comes from the
5953 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5954 I found this out when I moved to a new Linux workstation and tried to run
5955 the Exim test suite.
5957 To get round this problem I have changed the code so that it now does its
5958 own call to the DNS to look up PTR records when searching for a host name.
5959 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5960 addresses that are only in /etc/hosts are still found.
5962 This behaviour is, however, controlled by an option called host_lookup_
5963 order, which defaults to "bydns:byaddr". If people want to use the other
5964 order, or indeed, just use one or the other means of lookup, they can
5965 specify it in this variable.
5967 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5968 some operating systems, this comes back from gethostbyaddr() as an empty
5969 string, and this is what Exim used to test for. However, it seems that in
5970 other systems, "." is yielded. Exim now tests for this case too.
5972 7. The values of check_spool_space and check_log_space are now held internally
5973 as a number of kilobytes instead of an absolute number of bytes. If a
5974 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5975 kilobyte. This means that much larger values can be stored.
5977 8. Exim monitor: an attempt to get the action menu when not actually pointing
5978 at a message produces an empty menu entitled "No message selected". This
5979 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5980 no entries in it ("Shell widget menu has zero width and/or height"). So I
5981 have added a single, blank menu entry in this case.
5983 9. Added ${quote_local_part.
5985 10. MIME decoding is now applied to the contents of Subject: header lines when
5988 11. Now that a reference to $sender_host_address automatically causes a reverse
5989 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5990 host lookup before query-style lookups in lists that might use this
5991 variable. This has therefore been abolished, and the "net-" prefix is no
5992 longer necessary for query-style lookups.
5994 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5995 have been a typo for LFLAGS, so it has been changed.
5997 13. The install script calls Exim with "-C /dev/null" in order to find the
5998 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5999 to be output. However, since Exim outputs its version number before the
6000 error, it didn't break the script. It just looked ugly. I fixed this by
6001 always allowing "-C /dev/null" if the caller is root.
6003 14. Ignore overlarge ACL variable number when reading spool file - insurance
6004 against a later release with more variables having written the file.
6006 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6007 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6008 this, as well as the form without the "IPv6" on the front.
6010 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6011 OS/Makefile-Darwin file.
6013 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6014 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6016 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6017 variable. However, some OS use ints for inodes. I've added cast to long int
6018 to get rid of the compiler warning.
6020 19. I had forgotten to lock out "/../" in configuration file names when
6021 ALT_CONFIG_PREFIX was set.
6023 20. Routers used for verification do not need to specify transports. However,
6024 if such a router generated a host list, and callout was configured, Exim
6025 crashed, because it could not find a port number from the (non-existent)
6026 transport. It now assumes port 25 in this circumstance.
6028 21. Added the -t option to exigrep.
6030 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6031 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6032 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6033 them is set, LOOKUP_LSEARCH is forced.
6035 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6037 24. Added sender and host information to the "rejected by local_scan()" log
6038 line; previously there was no indication of these.
6040 25. Added .include_if_exists.
6042 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6043 Linux. It turns out that not all file systems support this. Apparently some
6044 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6045 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6046 not supported on the file descriptor, is now ignored when Exim is trying to
6047 sync a directory. This applies only to Linux.
6049 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6051 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6052 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6055 29. Added support for maildirsize files from supplied patch (modified a bit).
6057 30. The use of :fail: followed by an empty string could lead Exim to respond to
6058 sender verification failures with (e.g.):
6060 550 Verification failed for <xxx>
6061 550 Sender verify failed
6063 where the first response line was missing the '-' that indicates it is not
6064 the final line of the response.
6066 31. The loop for finding the name of the user that called Exim had a hardwired
6067 limit of 10; it now uses the value of finduser_retries, which is used for
6068 all other user lookups.
6070 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6073 33. Exim was neglecting to zero errno before one call of strtol() when
6074 expanding a string and expecting an integer value. On some systems this
6075 resulted in spurious "integer overflow" errors. Also, it was casting the
6076 result into an int without checking.
6078 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6079 did not work. The code looks as if it has *never* worked, though it appears
6080 to have been documented since at least release 1.62. I have made it work.
6082 35. The "timeout_DNS" error in retry rules, also documented since at least
6083 1.62, also never worked. As it isn't clear exactly what this means, and
6084 clearly it isn't a major issue, I have abolished the feature by treating it
6085 as "timeout", and writing a warning to the main and panic logs.
6087 36. The display of retry rules for -brt wasn't always showing the error code
6090 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6091 timeout_connect_A, timeout_connect_MX.
6093 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6094 to the empty sender.
6096 39. The daemon was not analysing the content of -oX till after it had closed
6097 stderr and disconnected from the controlling terminal. This meant that any
6098 syntax errors were only noted on the panic log, and the return code from
6099 the command was 0. By re-arranging the code a little, I've made the
6100 decoding happen first, so such errors now appear on stderr, and the return
6101 code is 1. However, the actual setting up of the sockets still happens in
6102 the disconnected process, so errors there are still only recorded on the
6105 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6106 connections (as happens on some IP stacks) was logged at start up time as
6107 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6108 it from "IPv6 and IPv4", which means that two separate sockets are being
6111 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6112 says whether AF_INET or AF_INET6 was passed as an argument.
6114 42. Exiwhat output was messed up when time zones were included in log
6117 43. Exiwhat now gives more information about the daemon's listening ports,
6118 and whether -tls-on-connect was used.
6120 44. The "port" option of the smtp transport is now expanded.
6122 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6123 silently ignored. Now an error message is written to the main and panic
6126 46. There's a new ACL modifier called "logwrite" which writes to a log file
6127 as soon as it is encountered.
6129 47. Added $local_user_uid and $local_user_gid at routing time.
6131 48. Exim crashed when trying to verify a sender address that was being
6134 49. Exim was recognizing only a space character after ".include". It now also
6135 recognizes a tab character.
6137 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6138 extracting the relevant information from the specification. The man page no
6139 longer contains scrambled data for the -d option, and I've added a section
6140 at the front about calling Exim under different names.
6142 51. Added "extra_headers" argument to the "mail" command in filter files.
6144 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6147 53. Installed eximstats 1.29.
6149 54. Added transport_filter_timeout as a generic transport option.
6151 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6152 Cc: header lines. This was required by RFC 822, but it not required by RFC
6155 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6156 incoming messages that did not have them. Now it does so only if the
6157 message originates locally, that is, if there is no associated remote host
6158 address. When Resent- header lines are present, this applies to the Resent-
6159 lines rather than the non-Resent- lines.
6161 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6162 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6164 58. Messages for configuration errors now include the name of the main
6165 configuration file - useful now that there may be more than one file in a
6166 list (.included file names were always shown).
6168 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6169 for those rare installations that do not start the daemon as root or run it
6170 setuid root. I've cut out the call to initgroups() if the daemon is not
6173 60. The Exim user and group can now be bound into the binary as text strings
6174 that are looked up at the start of Exim's processing.
6176 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6178 62. Added $mailstore_basename variable.
6180 63. Installed patch to sieve.c from Michael Haardt.
6182 64. When Exim failed to open the panic log after failing to open the main log,
6183 the original message it was trying to log was written to stderr and debug
6184 output, but if they were not available (the usual case in production), it
6185 was lost. Now it is written to syslog before the two lines that record the
6186 failures to open the logs.
6188 65. Users' Exim filters run in subprocesses under the user's uid. It is
6189 possible for a "deliver" command or an alias in a "personal" command to
6190 provoke an address rewrite. If logging of address rewriting is configured,
6191 this fails because the process is not running as root or exim. There may be
6192 a better way of dealing with this, but for the moment (because 4.30 needs
6193 to be released), I have disabled address rewrite logging when running a
6194 filter in a non-root, non-exim process.
6200 1. The buildconfig auxiliary program wasn't quoting the value set for
6201 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6202 not defined. This bug was masked in 4.22 by the effect that was fixed in
6205 2. Some messages that were rejected after a message id was allocated were
6206 shown as "incomplete" by exigrep. It no longer does this for messages that
6207 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6209 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6210 have allow_domain_literals set, the ID did not get logged in the <= line.
6211 Domain literals are now always recognized in Message-ID: header lines.
6213 4. The first argument for a ${extract expansion item is the key name or field
6214 number. Leading and trailing spaces in this item were not being ignored,
6215 causing some misleading effects.
6217 5. When deliver_drop_privilege was set, single queue runner processes started
6218 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6219 same command in the process it spins off) were not dropping privilege.
6221 6. When the daemon running as "exim" started a queue runner, it always
6222 re-executed Exim in the spun-off process. This is a waste of effort when
6223 deliver_drop_privilege is set. The new process now just calls the
6224 queue-runner function directly.
6230 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6233 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6234 ignored. Though the use of -oP was forcing the writing of a pid file, it
6235 was always written to the default place.
6237 3. If the message "no IP address found for host xxxx" is generated during
6238 incoming verification, it is now followed by identification of the incoming
6239 connection (so you can more easily find what provoked it).
6241 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6243 5. Added some features to "harden" Exim a bit more against certain attacks:
6245 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6246 be put in Local/Makefile. This is like the never_users runtime option,
6247 but it cannot be overridden. The default setting is "root".
6249 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6250 prefix string with which any file named in a -C command line option
6253 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6254 is retained for -C and -D only if the caller of Exim is root. Without
6255 it, the exim user may also use -C and -D and retain privilege.
6257 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6258 command line option is disabled.
6260 6. Macro names set by the -D option must start with an upper case letter, just
6261 like macro names defined in the configuration file.
6263 7. Added "dereference=" facility to LDAP.
6265 8. Two instances of the typo "uknown" in the source files are fixed.
6267 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6268 the Configure-Makefile script screwed up while processing it.
6270 10. Incorporated PCRE 4.4.
6272 11. The SMTP synchronization check was not operating right at the start of an
6273 SMTP session. For example, it could not catch a HELO sent before the client
6274 waited for the greeting. There is now a check for outstanding input at the
6275 point when the greeting is written. Because of the duplex, asynchronous
6276 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6277 way, but not yet received, when the check is performed.
6279 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6280 on TCP/IP sockets, because this apparently causes some broken clients to
6283 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6284 unchanged) from the Cygwin maintainer.
6286 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6287 was supported instead of testing for mbx. Effectively a typo.
6289 15. The spa authenticator server code was not checking that the input it
6290 received was valid base64.
6292 16. The debug output line for the "set" modifier in ACLs was not showing the
6293 name of the variable that was being set.
6295 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6297 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6298 lookup on its own. Something else was needed to trigger the lookup. For
6299 example, a match in host_lookup or the need for a host name in a host list.
6300 Now, if $sender_host_name is referenced and the host name has not yet been
6301 looked up, a lookup is performed. If the lookup fails, the variable remains
6302 empty, and $host_lookup_failed is set to "1".
6304 19. Added "eqi" as a case-independent comparison operator.
6306 20. The saslauthd authentication condition could segfault if neither service
6307 nor realm was specified.
6309 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6310 error message that was logged was misleading, and incoming SMTP
6311 connections were dropped. The message is now more accurate, and temporary
6312 errors are given to SMTP connections.
6314 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6315 (except RSET) with a 421 error, until QUIT is received. However, it was
6316 failing to send a response to QUIT.
6318 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6319 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6320 ACL is now run after the helo_try_verify_hosts code.
6322 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6323 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6324 case-independent, so other case variants are also recognized). Apparently
6325 some systems use these upper case variants.
6327 25. If more than two messages were waiting for the same host, and a transport
6328 filter was specified for the transport, Exim sent two messages over the
6329 same TCP/IP connection, and then failed with "socket operation on non-
6330 socket" when it tried to send the third.
6332 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6334 27. The extern definition of crypt16() in expand.c was not being excluded when
6335 the OS had its own crypt16() function.
6337 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6338 as a preferred synonym for return_size_limit, both as an option and as an
6341 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6343 30. Changed the default configuration ACL to relax the local part checking rule
6344 for addresses that are not in any local domains. For these addresses,
6345 slashes and pipe symbols are allowed within local parts, but the sequence
6346 /../ is explicitly forbidden.
6348 31. SPA server authentication was not clearing the challenge buffer before
6351 32. log_message in a "warn" ACL statement was writing to the reject log as
6352 well as to the main log, which contradicts the documentation and doesn't
6353 seem right (because no rejection is happening). So I have stopped it.
6355 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6356 However, I am unable to do any testing of this.
6358 34. Fixed an infelicity in the appendfile transport. When checking directories
6359 for a mailbox, to see if any needed to be created, it was accidentally
6360 using path names with one or more superfluous leading slashes; tracing
6361 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6363 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6364 contents are added to the log line that is written for every discarded
6365 recipient. (Previously a log_message setting was ignored.)
6367 36. The ${quote: operator now quotes the string if it is empty.
6369 37. The install script runs exim in order to find its version number. If for
6370 some reason other than non-existence or emptiness, which it checks, it
6371 could not run './exim', it was installing it with an empty version number,
6372 i.e. as "exim-". This error state is now caught, and the installation is
6375 38. An argument was missing from the function that creates an error message
6376 when Exim fails to connect to the socket for saslauthd authentication.
6377 This could cause Exim to crash, or give a corrupted message.
6379 39. Added isip, isip4, and isip6 to ${if conditions.
6381 40. The ACL variables $acl_xx are now saved with the message, and can be
6382 accessed later in routers, transports, and filters.
6384 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6385 strings in the file are not string-expanded.
6387 42. If a MAIL command specified a SIZE value that was too large to fit into an
6388 int variable, the check against message_size_limit failed. Such values are
6389 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6390 day this will have to be increased, but I don't think I want to be around
6391 when emails are that large.
6398 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6399 iconv() is not standard in FreeBSD.
6401 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6402 IPv6 enabled. The observed symptom was a segmentation fault on return from
6403 the function os_common_find_running_interfaces() in src/os.c.
6405 3. In the check_special_case() function in daemon.c I had used "errno" as an
6406 argument name, which causes warnings on some systems. This was basically a
6407 typo, since it was named "eno" in the comments!
6409 4. The code that waits for the clock to tick (at a resolution of some fraction
6410 of a second) so as to ensure message-id uniqueness was always waiting for
6411 at least one whole tick, when it could have waited for less. [This is
6412 almost certainly not relevant at current processor speeds, where it is
6413 unlikely to ever wait at all. But we try to future-proof.]
6415 5. The function that sleeps for a time interval that includes fractions of a
6416 second contained a race. It did not block SIGALRM between setting the
6417 timer, and suspending (a couple of lines later). If the interval was short
6418 and the sigsuspend() was delayed until after it had expired, the suspension
6419 never ended. On busy systems this could lead to processes getting stuck for
6422 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6423 process, before it forks any delivery processes. The open lookup caching
6424 mechanism meant that the open file or database connection was passed into
6425 the delivery process. The problem was that delivery processes always tidy
6426 up cached lookup data. This could cause a problem for the next delivery
6427 process started by the queue runner, because the external queue runner
6428 process does not know about the closure. So the next delivery process
6429 still has data in the lookup cache. In the case of a file lookup, there was
6430 no problem because closing a file descriptor in a subprocess doesn't affect
6431 the parent. However, if the lookup was caching a connection to a database,
6432 the connection was closed, and the second delivery process was likely to
6433 see errors such as "PGSQL: query failed: server closed the connection
6434 unexpectedly". The problem has been fixed by closing all cached lookups
6435 in a queue runner before running a delivery process.
6437 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6438 seem to have the "const" qualifier which it has on other OS. I've
6441 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6442 *of the same type* (client or server) with the same public name that an
6443 error should be diagnosed.
6445 9. When Exim looked up a host name for an IP address, but failed to find the
6446 original IP address when looking up the host name (a safety check), it
6447 output the message "<ip address> does not match any IP for NULL", which was
6448 confusing, to say the least. The bug was that the host name should have
6449 appeared instead of "NULL".
6451 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6452 user that is built into the binary, and the -C or -D options is used, root
6453 privilege is dropped before the configuration file is read. In addition,
6454 logging is switched to stderr instead of the normal log files. If the
6455 configuration then re-defines the Exim user, the unprivileged environment
6456 is probably not what is expected, so Exim logs a panic warning message (but
6459 However, if deliver_drop_privilege is set, the unprivileged state may well
6460 be exactly what is intended, so the warning has been cut out in that case,
6461 and Exim is allowed to try to write to its normal log files.
6467 1. smtp_return_error_details was not giving details for temporary sender
6468 or receiver verification errors.
6470 2. Diagnose a configuration error if two authenticators have the same public
6473 3. Exim used not to create the message log file for a message until the first
6474 delivery attempt. This could be confusing when incoming messages were held
6475 for policy or load reasons. The message log file is now created at the time
6476 the message is received, and an initial "Received" line is written to it.
6478 4. The automatically generated man page for command line options had a minor
6479 bug that caused no ill effects; however, a more serious problem was that
6480 the procedure for building the man page automatically didn't always
6481 operate. Consequently, release 4.20 contains an out-of-date version. This
6482 shouldn't happen again.
6484 5. When building Exim with embedded Perl support, the script that builds the
6485 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6486 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6488 6. The freeze_tell option was not being used for messages that were frozen on
6489 arrival, either by an ACL or by local_scan().
6491 7. Added the smtp_incomplete_transaction log selector.
6493 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6494 was accepting AUTH without a new EHLO.
6496 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6497 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6500 10. Logging of TCP/IP connections (when configured) now happens in the main
6501 daemon process instead of the child process, so that the TCP/IP connection
6502 count is more accurate (but it can never be perfect).
6504 11. The use of "drop" in a nested ACL was not being handled correctly in the
6505 outer ACL. Now, if condition failure induced by the nested "drop" causes
6506 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6507 or "require"), the connection is dropped.
6509 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6510 that yield "discard" can now be used with an "accept" or a "discard" verb,
6511 but an error is generated for any others (because I can't see a useful way
6512 to define what should happen).
6514 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6515 are now processed in the same way as lines in the Exim configuration file.
6516 In particular, continuation lines are supported.
6518 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6520 15. Added -ti meaning -t -i.
6522 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6523 domains, and warn by logging if others are found.
6525 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6526 returned from GIFCONF when Exim is trying to find the list of interfaces on
6527 a host. The code in os.c has been modified to copy each ifreq to an aligned
6528 structure in all cases.
6530 Also, in some cases, the returned ifreq's were being copied to a 'struct
6531 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6532 means the last couple of bytes of an IPv6 address could be chopped if the
6533 ifreq contained only a normal sockaddr (14 bytes storage).
6535 18. Named domain lists were not supported in the hosts_treat_as_local option.
6536 An entry such as +xxxx was not recognized, and was treated as a literal
6539 19. Ensure that header lines added by a DATA ACL are included in the reject log
6540 if the ACL subsequently rejects the message.
6542 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6543 MD5 (which is deprecated).
6545 21. When testing a filter file using -bf, Exim was writing a message when it
6546 took the sender from a "From " line in the message, but it was not doing so
6547 when it took $return_path from a Return-Path: header line. It now does.
6549 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6550 with a valid header line field name (a series of printing characters
6551 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6553 23. Changed "disc" in the source to "disk" to conform to the documentation and
6554 the book and for uniformity.
6556 24. Ignore Sendmail's -Ooption=value command line item.
6558 25. When execve() failed while trying to run a command in a pipe transport,
6559 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6560 could be confused with a return value of 69 from the command itself. This
6561 has been changed to 127, the value the shell returns if it is asked to run
6562 a non-existent command. The wording for the related log line suggests a
6563 non-existent command as the problem.
6565 26. If received_header_text expands to an empty string, do not add a Received:
6566 header line to the message. (Well, it adds a token one on the spool, but
6567 marks it "old" so that it doesn't get used or transmitted.)
6569 27. Installed eximstats 1.28 (addition of -nt option).
6571 28. There was no check for failure on the call to getsockname() in the daemon
6572 code. This can fail if there is a shortage of resources on the system, with
6573 ENOMEM, for example. A temporary error is now given on failure.
6575 29. Contrary to the C standard, it seems that in some environments, the
6576 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6577 program. Exim now does this explicitly; it affects the formatting of
6578 timestamps using strftime().
6580 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6581 complaints. I've now initialized all the variables, to avoid this.
6583 32. Header lines added by a system filter were not being "seen" during
6584 transport-time rewrites.
6586 33. The info_callback() function passed to OpenSSL is set up with type void
6587 (*)(SSL *, int, int), as described somewhere. However, when calling the
6588 function (actually a macro) that sets it up, the type void(*)() is
6589 expected. I've put in a cast to prevent warnings from picky compilers.
6591 34. If a DNS black list lookup found a CNAME record, but there were no A
6592 records associated with the domain it pointed at, Exim crashed.
6594 35. If a DNS black list lookup returned more than one A record, Exim ignored
6595 all but the first. It now scans all returned addresses if a particular IP
6596 value is being sought. In this situation, the contents of the
6597 $dnslist_value variable are a list of all the addresses, separated by a
6600 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6601 to accept a host name and all its aliases if the forward lookup for any of
6602 them yielded the IP address of the incoming connection. Now it accepts only
6603 those names whose forward lookup yields the correct IP address. Any other
6604 names are discarded. This closes a loophole whereby a rogue DNS
6605 administrator could create reverse DNS records to break through a
6606 wildcarded host restriction in an ACL.
6608 37. If a user filter or a system filter that ran in a subprocess used any of
6609 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6610 the wrong values were passed to the pipe command ($thisaddress had the
6611 value of $0, $0 had the value of $1, etc). This bug was introduced by
6612 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6614 38. Improved the line breaking for long SMTP error messages from ACLs.
6615 Previously, if there was no break point between 40 and 75 characters, Exim
6616 left the rest of the message alone. Two changes have been made: (a) I've
6617 reduced the minimum length to 35 characters; (b) if it can't find a break
6618 point between 35 and 75 characters, it looks ahead and uses the first one
6619 that it finds. This may give the occasional overlong line, but at least the
6620 remaining text gets split now.
6622 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6623 file descriptors might be low, and that setting 1000 would always raise it.
6624 It turns out that in some environments, the limit is already over 1000 and
6625 that lowering it causes trouble. So now Exim takes care not to decrease it.
6627 40. When delivering a message, the value of $return_path is set to $sender_
6628 address at the start of routing (routers may change the value). By an
6629 oversight, this default was not being set up when an address was tested by
6630 -bt or -bv, which affected the outcome if any router or filter referred to
6633 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6634 log only when "message" or "log_message" are set. However, if one of the
6635 conditions was an address verification, or a call to a nested ACL, the
6636 messages generated by the underlying test were being passed through. This
6637 no longer happens. The underlying message is available in $acl_verify_
6638 message for both "message" and "log_message" expansions, so it can be
6639 passed through if needed.
6641 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6642 new expansion $bh_ to give the encoded byte string without charset
6643 translation. Translation happens only if iconv() is available; HAVE_ICONV
6644 indicates this at build time. HEADERS_CHARSET gives the charset to
6645 translate to; headers_charset can change it in the configuration, and
6646 "headers charset" can change it in an individual filter file.
6648 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6649 that creates RFC 2047 encoded "words" labels them as that charset instead
6650 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6651 expansion operator; (ii) when Exim creates a From: line for a local
6652 message; (iii) when a header line is rewritten to include a "phrase" part.
6654 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6655 buggy, causing it to skip the first lines of messages whose message ID
6656 ended in 'D'. This would not have bitten before Exim release 4.14, because
6657 message IDs were unlikely to end in 'D' before then. The effect was to have
6658 incorrect size information for certain domains.
6660 45. #include "config.h" was missing at the start of the crypt16.c module. This
6661 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6664 46. If there was a timeout during a "random" callout check, Exim treated it as
6665 a failure of the random address, and carried on sending RSET and the real
6666 address. If the delay was just some slowness somewhere, the response to the
6667 original RCPT would be taken as a response to RSET and so on, causing
6668 mayhem of various kinds.
6670 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6671 when I implemented it. It didn't allow for the fact that some option values
6672 may legitimately be negative (e.g. size_addition), and it didn't even do
6673 the right test for positive values.
6675 48. Domain names in DNS records are case-independent. Exim always looks them up
6676 in lower case. Some resolvers return domain names in exactly the case they
6677 appear in the zone file, that is, they may contain uppercase letters. Not
6678 all resolvers do this - some return always lower case. Exim was treating a
6679 change of case by a resolver as a change of domain, similar to a widening
6680 of a domain abbreviation. This triggered its re-routing code and so it was
6681 trying to route what was effectively the same domain again. This normally
6682 caused routing to fail (because the router wouldn't handle the domain
6683 twice). Now Exim checks for this case specially, and just changes the
6684 casing of the domain that it ultimately uses when it transmits the message
6687 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6690 50. If a filter generated a file delivery with a non-absolute name (possible if
6691 no home directory exists for the router), the forbid_file option was not
6694 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6695 the existing equality matching.
6697 52. Exim was using ints instead of ino_t variables in some places where it was
6698 dealing with inode numbers.
6700 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6701 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6702 called TMPDIR, and if it finds it is different, it changes its value.
6704 54. The smtp_printf() function is now made available to local_scan() so
6705 additional output lines can be written before returning. There is also an
6706 smtp_fflush() function to enable the detection of a dropped connection.
6707 The variables smtp_input and smtp_batched_input are exported to
6710 55. Changed the default runtime configuration: the message "Unknown user"
6711 has been removed from the ACL, and instead placed on the localuser router,
6712 using the cannot_route_message feature. This means that any verification
6713 failures that generate their own messages won't get overridden. Similarly,
6714 the "Unrouteable address" message that was in the ACL for unverifiable
6715 relay addresses has also been removed.
6717 56. Added hosts_avoid_esmtp to the smtp transport.
6719 57. The exicyclog script was not checking for the esoteric option
6720 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6721 will work only if exicyclog is run under the appropriate euid.
6723 58. Following a discussion on the list, the rules by which Exim recognises line
6724 endings on incoming messages have been changed. The -dropcr and drop_cr
6725 options are now no-ops, retained only for backwards compatibility. The
6726 following line terminators are recognized: LF CRLF CR. However, special
6727 processing applies to CR:
6729 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6730 nor a local message in the state where . is a terminator.
6732 (ii) If a bare CR is encountered in a header line, an extra space is added
6733 after the line terminator so as not to end the header. The reasoning
6734 behind this is that bare CRs in header lines are most likely either
6735 to be mistakes, or people trying to play silly games.
6737 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6738 was being incorrectly given as 18 bytes larger than it should have been.
6739 This is a VOB (very old bug).
6741 60. This may never have affected anything current, but just in case it has:
6742 When the local host is found other than at the start of a list of hosts,
6743 the local host, those with the same MX, and any that follow, are discarded.
6744 When the list in question was part of a longer list of hosts, the following
6745 hosts (not currently being processed) were also being discarded. This no
6746 longer happens. I'm not sure if this situation could ever has previously
6749 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6752 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6753 header. This is something that is recommended in a new Internet Draft, and
6754 is something that is documented as being done by Sendmail. There are two
6755 possible values. For messages generated by the autoreply transport, Exim
6758 Auto-Submitted: auto-replied
6760 whereas for all other generated messages (e.g. bounces) it adds
6762 Auto-Submitted: auto-generated
6764 63. The "personal" condition in filters now includes a test for the
6765 Auto-Submitted: header. If it contains the string "auto-" the message it
6766 not considered personal.
6768 64. Added rcpt_include_affixes as a generic transport option.
6770 65. Added queue_only_override (default true).
6772 66. Added the syslog_duplication option.
6774 67. If what should have been the first header line of a message consisted of
6775 a space followed by a colon, Exim was mis-interpreting it as a header line.
6776 It isn't of course - it is syntactically invalid and should therefore be
6777 treated as the start of the message body. The misbehaviour could have
6778 caused a number of strange effects, including loss of data in subsequent
6779 header lines, and spool format errors.
6781 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6782 client host had authenticated. This control can now be exercised by an ACL
6783 for more flexibility.
6785 69. By default, callouts do not happen when testing with -bh. There is now a
6786 variant, -bhc, which does actually run the callout code, including
6787 consulting and updating the callout cache.
6789 70. Added support for saslauthd authentication, courtesy of Alexander
6792 71. If statvfs() failed on the spool or log directories while checking their
6793 size for availability, Exim confusingly gave the error "space shortage".
6794 Furthermore, in debugging mode it crashed with a floating point exception.
6795 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6796 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6797 serious problem, Exim now writes to the main and panic logs when this
6798 happens, with details of the failure. It then refuses to accept the
6799 incoming message, giving the message "spool directory problem" or "log
6800 directory problem" with a 421 code for SMTP messages.
6802 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6803 0, 1, and 2 exist, because some OS complain for execs without them (see
6804 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6805 descriptors. However, the code omitted to check that the open succeeded,
6806 causing mysterious errors if for some reason the permissions on /dev/null
6807 got screwed. Now Exim writes a message to the main and panic logs, and
6808 bombs out if it can't open /dev/null.
6810 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6811 interact so that it is all more flexible. It is supposed to remain
6812 backwards compatible. Also added extra_local_interfaces.
6814 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6815 to bomb out with an assertion failure - to the client this appears as a
6816 connection drop. This problem occurs in the part of the code that was taken
6817 from the Samba project. Fortunately, the assertion is in a very simple
6818 function, so I have fixed this by reproducing the function inline in the
6819 one place where it is called, and arranging for authentication to fail
6820 instead of killing the process with assert().
6822 75. The SPA client code was not working when the server requested OEM rather
6823 than Unicode encoding.
6825 76. Added code to make require_files with a specific uid setting more usable in
6826 the case where statting the file as root fails - usually a non-root-mounted
6827 NFS file system. When this happens and the failure is EACCES, Exim now
6828 forks a subprocess and does the per-uid checking as the relevant uid.
6830 77. Added process_log_path.
6832 78. If log_file_path was not explicitly set, a setting of check_log_space or
6833 check_log_inodes was ignored.
6835 79. If a space check for the spool or log partitions fails, the incident is now
6836 logged. Of course, in the latter case the data may get lost...
6838 80. Added the %p formatting code to string_format() so that it can be used to
6839 print addresses in debug_print(). Adjusted all the address printing in the
6840 debugging in store.c to use %p rather than %d.
6842 81. There was a concern that a line of code in smtp_in.c could overflow a
6843 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6844 initially expressed, the concern was not well-founded, because trailing
6845 spaces are removed early. However, if the trailing spaces were followed by
6846 a NULL, they did not get removed, so the overflow was possible. Two fixes
6849 (a) I re-wrote the offending code in a cleaner fashion.
6850 (b) If an incoming SMTP command contains a NULL character, it is rejected
6853 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6854 runs initgroups(), so that if the Exim user is in any additional groups,
6855 they will be used during message reception.
6861 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob