1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
40 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
41 Disable the session-cache too, which might reduce our load. Since we
42 currrectly use a new context for every connection, both as server and
43 client, there is no benefit for these.
44 GnuTLS appears to not support tickets server-side by default (we don't
45 call gnutls_session_ticket_enable_server()) but client side is enabled
46 by default on recent versions (3.1.3 +) unless the PFS priority string
49 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
50 <https://reproducible-builds.org/specs/source-date-epoch/>.
52 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
53 the check for any unsuccessful recipients did not notice the limit, and
54 erroneously found still-pending ones.
56 JH/08 Pipeline CHUNKING command and data together, on kernels that support
57 MSG_MORE. Only in-clear (not on TLS connections).
59 JH/09 Avoid using a temporary file during transport using dkim. Unless a
60 transport-filter is involved we can buffer the headers in memory for
61 creating the signature, and read the spool data file once for the
62 signature and again for transmission.
64 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
65 4.77 as the kernel support then wasn't solid, having issues in 64bit
66 mode. Now, it's been long enough. Add support for FreeBSD also.
68 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
69 case where the routing stage had gathered several addresses to send to
70 a host before calling the transport for the first, we previously failed
71 to close down TLS in the old transport process before passing the TCP
72 connection to the new process. The new one sent a STARTTLS command
73 which naturally failed, giving a failed delivery and bloating the retry
74 database. Investigation and fix prototype from Wolfgang Breyha.
76 JH/12 Fix check on SMTP command input synchronisation. Previously there were
77 false-negatives in the check that the sender had not preempted a response
78 or prompt from Exim (running as a server), due to that code's lack of
79 awareness of the SMTP input buffering.
81 PP/04 Add commandline_checks_require_admin option.
82 Exim drops privileges sanely, various checks such as -be aren't a
83 security problem, as long as you trust local users with access to their
84 own account. When invoked by services which pass untrusted data to
85 Exim, this might be an issue. Set this option in main configuration
86 AND make fixes to the calling application, such as using `--` to stop
89 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
90 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
91 responses to those, into a single TLS record each way (this usually means
92 a single packet). As a side issue, smtp_enforce_sync now works on TLS
95 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
96 affects you only if you're dancing at the edge of the param size limits.
97 If you are, and this message makes sense to you, then: raise the
98 configured limit or use OpenSSL 1.1. Nothing we can do for older
101 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
102 cmdline element to get the documented default one. Previously it was
105 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
106 are unsuable for non-message-specific reasons - eg. connection timeout,
107 banner-time rejection.
109 JH/16 Fix logging of delivery remote port, when specified by router, under
112 PP/06 Repair manualroute's ability to take options before hostnames.
119 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
120 than -2003 did; needs libidn2 in addition to libidn.
122 JH/02 The path option on a pipe transport is now expanded before use.
124 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
125 Patch provided by "Björn", documentation fix added too.
127 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
128 missing a wire-to-host endian conversion.
130 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
131 close after a BDAT command line could be taken as a following command,
132 giving a synch failure. Fix by only checking for synch immediately
133 before acknowledging the chunk.
135 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
136 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
137 macro. Patches provided by Josh Soref.
139 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
140 Previously we did not; the RFC seems ambiguous and VRFY is not listed
141 by IANA as a service extension. However, John Klensin suggests that we
144 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
145 the dkim code may be unix-mode line endings rather than smtp wire-format
146 CRLF, so prepend a CR to any bare LF.
148 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
149 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
151 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
152 the first were themselves being wrongly included in the feed into dkim
153 processing; with most chunk sizes in use this resulted in an incorrect
154 body hash calculated value.
156 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
157 DKIM signature block, for verification. Although advised against by
158 standards it is specifically not ruled illegal.
160 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
162 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
163 missing a body hash (the bh= tag).
165 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
166 It seems that HAProxy sends the Proxy Protocol information in clear and
167 only then does a TLS startup, so do the same.
169 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
170 TCP connections (such as for Spamd) unless the daemon successfully set
171 Fast Open mode on its listening sockets. This fixes breakage seen on
172 too-old kernels or those not configured for Fast Open, at the cost of
173 requiring both directions being enabled for TFO, and TFO never being used
174 by non-daemon-related Exim processes.
176 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
177 endings, at least on the first header line. Try to canonify any that get
178 past that check, despite the cost.
180 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
181 now limited to an arbitrary five deep, while parsing addresses with the
182 strip_excess_angle_brackets option enabled.
184 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
185 instead leave the unprompted TLS handshake in socket buffer for the
186 TLS library to consume.
188 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
190 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
192 JH/16 Drop variables when they go out of scope. Memory management drops a whole
193 region in one operation, for speed, and this leaves assigned pointers
194 dangling. Add checks run only under the testsuite which checks all
195 variables at a store-reset and panics on a dangling pointer; add code
196 explicitly nulling out all the variables discovered. Fixes one known
197 bug: a transport crash, where a dangling pointer for $sending_ip_address
198 originally assigned in a verify callout, is re-used.
200 PP/06 Drop '.' from @INC in various Perl scripts.
202 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
204 PP/08 Reduce a number of compilation warnings under clang; building with
205 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
206 should be warning-free.
208 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
210 HS/01 Fix portability problems introduced by PP/08 for platforms where
211 realloc(NULL) is not equivalent to malloc() [SunOS et al].
213 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
214 chunk. This allows us to accept broken chunked messages. We need a more
215 general solution here.
217 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
218 already-broken messages in the queue.
220 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
222 JH/19 Fix reference counting bug in routing-generated-address tracking.
228 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
229 supports it and a size is available (ie. the sending peer gave us one).
231 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
232 years of being deprecated). The replacements are the ACLs
233 acl_smtp_mime and acl_not_smtp_mime.
235 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
236 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
237 or even an in-clear connection were permitted. Now, if the host lookup
238 was dnssec and dane was requested then the host is only used if the
239 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
240 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
241 if one fails this test.
242 This means that a poorly-configured remote DNS will make it incommunicado;
243 but it protects against a DNS-interception attack on it.
245 JH/04 Bug 1810: make continued-use of an open smtp transport connection
246 non-noisy when a race steals the message being considered.
248 JH/05 If main configuration option tls_certificate is unset, generate a
249 self-signed certificate for inbound TLS connections.
251 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
252 in rewrites and routers.
254 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
255 and logged a warning sing 4.83; now they are a configuration file error.
257 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
258 (lacking @domain). Apply the same qualification processing as RCPT.
260 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
262 JH/10 Support ${sha256:} applied to a string (as well as the previous
265 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
266 a cutthrough deliver is pending, as we always want to make a connection.
267 This also avoids re-routing the message when later placing the cutthrough
268 connection after a verify cache hit.
269 Do not update it with the verify result either.
271 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
272 when routing results in more than one destination address.
274 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
275 signing (which inhibits the cutthrough capability). Previously only
276 the presence of an option was tested; now an expansion evaluating as
277 empty is permissible (obviously it should depend only on data available
278 when the cutthrough connection is made).
280 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
281 the relevant preceding SMTP command did not note the pipelining mode.
283 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
284 Previously they were not counted.
286 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
287 as one having no matching records. Previously we deferred the message
288 that needed the lookup.
290 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
291 distinguished as "(=".
293 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
294 for missing MX records. Previously it only worked for missing A records.
296 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
298 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
299 after the data-go-ahead and data-ack. Patch from Jason Betts.
301 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
302 even for a "none" policy. Patch from Tony Meyer.
304 JH/22 Fix continued use of a connection for further deliveries. If a port was
305 specified by a router, it must also match for the delivery to be
308 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
309 When one of the recipients of a message was unsuitable for the connection
310 (has no matching addresses), we lost track of needing to mark it
311 deferred. As a result mail would be lost.
313 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
315 JH/25 Decoding ACL controls is now done using a binary search; the source code
316 takes up less space and should be simpler to maintain. Merge the ACL
317 condition decode tables also, with similar effect.
319 JH/26 Fix problem with one_time used on a redirect router which returned the
320 parent address unchanged. A retry would see the parent address marked as
321 delivered, so not attempt the (identical) child. As a result mail would
324 JH/27 Fix a possible security hole, wherein a process operating with the Exim
325 UID can gain a root shell. Credit to http://www.halfdog.net/ for
326 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
329 JH/28 Enable {spool,log} filesystem space and inode checks as default.
330 Main config options check_{log,spool}_{inodes,space} are now
331 100 inodes, 10MB unless set otherwise in the configuration.
333 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
334 Previously it only applied to the main-section connection policy
337 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
339 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
340 by me. Added RFC7919 DH primes as an alternative.
342 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
343 are not in the system include path.
345 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
346 GnuTLS, when a session startup failed (eg because the client disconnected)
347 Exim did stdio operations after fclose. This was exposed by a recent
348 change which nulled out the file handle after the fclose.
350 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
351 signed directly by the cert-signing cert, rather than an intermediate
352 OCSP-signing cert. This is the model used by LetsEncrypt.
354 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
356 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
357 an incoming connection.
359 HS/02 Bug 1802: Do not half-close the connection after sending a request
362 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
363 fallback to "prime256v1".
365 JH/34 SECURITY: Use proper copy of DATA command in error message.
366 Could leak key material. Remotely exploitable. CVE-2016-9963.
372 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
373 and 3.4.4 - once the server is enabled to respond to an OCSP request
374 it does even when not requested, resulting in a stapling non-aware
375 client dropping the TLS connection.
377 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
378 support variable-length bit vectors. No functional change.
380 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
381 The I= interface field on outgoing lines is now after the H= remote
382 host field, same as incoming lines. There is a separate
383 outgoing_interface log selector which allows you to disable the
386 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
387 If not running log_selector +smtp_connection the mainlog would be held
388 open indefinitely after a "too many connections" event, including to a
389 deleted file after a log rotate. Leave the per net connection logging
390 leaving it open for efficiency as that will be quickly detected by the
391 check on the next write.
393 HS/01 Bug 1671: Fix post transport crash.
394 Processing the wait-<transport> messages could crash the delivery
395 process if the message IDs didn't exist for some reason. When
396 using 'split_spool_directory=yes' the construction of the spool
397 file name failed already, exposing the same netto behaviour.
399 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
400 mime_regex ACL conditions.
402 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
403 to DSN fail messages (bounces): remote IP, remote greeting, remote response
404 to HELO, local diagnostic string.
406 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
407 log line to debug. Even when configured with a tls authenticator many
408 client connections are expected to not authenticate in this way, so
409 an authenticate fail is not an error.
411 HS/02 Add the Exim version string to the process info. This way exiwhat
412 gives some more detail about the running daemon.
414 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
415 matter for fast-change records such as DNSBLs.
417 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
418 retry record, even if constant. There may be multiple transports with
419 different interface settings and the retry behaviour needs to be kept
422 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
424 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
426 JH/10 Bug 840: fix log_defer_output option of pipe transport
428 JH/11 Bug 830: use same host for all RCPTS of a message, even under
429 hosts_randomize. This matters a lot when combined with mua_wrapper.
431 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
432 ${quote_pgsql:<string>} operator.
434 JH/13 Bug 1708: avoid misaligned access in cached lookup.
436 JH/14 Change header file name for freeradius-client. Relevant if compiling
437 with Radius support; from the Gentoo tree and checked under Fedora.
439 JH/15 Bug 1712: Introduce $prdr_requested flag variable
441 JH/16 Bug 1714: Permit an empty string as expansion result for transport
442 option transport_filter, meaning no filtering.
444 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
446 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
447 defaults to "*" (all hosts). The variable is now available when not built
448 with TLS, default unset, mainly to enable keeping the testsuite sane.
449 If a server certificate is not supplied (via tls_certificate) an error is
450 logged, and clients will find TLS connections fail on startup. Presumably
451 they will retry in-clear.
452 Packagers of Exim are strongly encouraged to create a server certificate
453 at installation time.
455 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
456 with the $config_file variable.
458 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
459 in transport context, after the attempt, and per-recipient. The latter type
460 is per host attempted. The event data is the error message, and the errno
461 information encodes the lookup type (A vs. MX) used for the (first) host,
462 and the trailing two digits of the smtp 4xx response.
464 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
465 to write to mainlog (or rejectlog, paniclog) in the window between file
466 creation and permissions/ownership being changed. Particularly affects
467 installations where exicyclog is run as root, rather than exim user;
468 result is that the running daemon panics and dies.
470 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
472 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
473 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
474 "pri" and "weight". Note that the previous implicit priority given by the
475 list order is no longer honoured.
477 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
480 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
481 by defining SUPPORT_SOCKS.
483 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
484 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
485 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
486 variables are renamed to proxy_{local,external}_{address,port}.
488 JH/27 Move Internationalisation support from Experimental to mainline, enabled
489 for a build by defining SUPPORT_I18N
491 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
492 of the query string, and make ${quote_redis:} do that quoting.
494 JH/29 Move Events support from Experimental to mainline, enabled by default
495 and removable for a build by defining DISABLE_EVENT.
497 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
499 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
500 cached by the daemon.
502 JH/32 Move Redis support from Experimental to mainline, enabled for a build
503 by defining LOOKUP_REDIS. The libhiredis library is required.
505 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
506 keys are given for lookup.
508 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
509 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
510 only supported when built with TLS support. The PolarSSL SHA routines
511 are still used when the TLS library is too old for convenient support.
513 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
514 openssl_options), for security. OpenSSL forces this from version 1.1.0
515 server-side so match that on older versions.
517 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
518 allocation for $value could be released as the expansion processing
519 concluded, but leaving the global pointer active for it.
521 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
522 and to use the domains and local_parts ACL conditions.
524 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
525 incorrectly not doubled on cutthrough transmission, hence seen as a
526 body-termination at the receiving system - resulting in truncated mails.
527 Commonly the sender saw a TCP-level error, and retransmitted the message
528 via the normal store-and-forward channel. This could result in duplicates
529 received - but deduplicating mailstores were liable to retain only the
530 initial truncated version.
532 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
534 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
536 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
537 we're in there, support oversigning also; bug 1309.
539 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
541 HS/04 Add support for keep_environment and add_environment options.
543 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
544 either intentional arithmetic overflow during PRNG, or testing config-
547 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
548 delivery resulted in actual delivery. Cancel cutthrough before DATA
551 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
552 rejects a recipient: pass the reject to the originator.
554 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
555 Many were false-positives and ignorable, but it's worth fixing the
558 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
559 for the new environment-manipulation done at startup. Move the routines
560 from being local to tls.c to being global via the os.c file.
562 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
563 an extract embedded as result-arg for a map, the first arg for extract
564 is unavailable so we cannot tell if this is a numbered or keyed
565 extraction. Accept either.
571 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
574 JH/02 The smtp transport option "multi_domain" is now expanded.
576 JH/03 The smtp transport now requests PRDR by default, if the server offers
579 JH/04 Certificate name checking on server certificates, when exim is a client,
580 is now done by default. The transport option tls_verify_cert_hostnames
581 can be used to disable this per-host. The build option
582 EXPERIMENTAL_CERTNAMES is withdrawn.
584 JH/05 The value of the tls_verify_certificates smtp transport and main options
585 default to the word "system" to access the system default CA bundle.
586 For GnuTLS, only version 3.0.20 or later.
588 JH/06 Verification of the server certificate for a TLS connection is now tried
589 (but not required) by default. The verification status is now logged by
590 default, for both outbound TLS and client-certificate supplying inbound
593 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
596 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
597 Status Notification (bounce) messages are now MIME format per RFC 3464.
598 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
599 under the control of the dsn_advertise_hosts option, and routers may
600 have a dsn_lasthop option.
602 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
603 default, modifiable by a malware= option. The list separator for
604 the options can now be changed in the usual way. Bug 68.
606 JH/10 The smtp_receive_timeout main option is now expanded before use.
608 JH/11 The incoming_interface log option now also enables logging of the
609 local interface on delivery outgoing connections.
611 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
612 if the interface and destination host and port all match.
614 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
617 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
618 Patch from Andrew Lewis.
620 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
621 now supports optional time-restrictions, weighting, and priority
622 modifiers per server. Patch originally by <rommer@active.by>.
624 JH/16 The spamd_address main option now supports a mixed list of local
625 and remote servers. Remote servers can be IPv6 addresses, and
626 specify a port-range.
628 JH/17 Bug 68: The spamd_address main option now supports an optional
629 timeout value per server.
631 JH/18 Bug 1581: Router and transport options headers_add/remove can
632 now have the list separator specified.
634 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
637 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
640 JH/21 Support for the A6 type of dns record is withdrawn.
642 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
643 rather than the verbs used.
645 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
646 from 255 to 1024 chars.
648 JH/24 Verification callouts now attempt to use TLS by default.
650 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
651 are generic router options now. The defaults didn't change.
653 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
654 Original patch from Alexander Shikoff, worked over by JH.
656 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
657 files if ZCAT_COMMAND is not executable.
659 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
661 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
663 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
664 Normally benign, it bites when the pair was led to by a CNAME;
665 modern usage is to not canonicalize the domain to a CNAME target
666 (and we were inconsistent anyway for A-only vs AAAA+A).
668 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
670 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
671 when evaluating $sender_host_dnssec.
673 JH/31 Check the HELO verification lookup for DNSSEC, adding new
674 $sender_helo_dnssec variable.
676 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
678 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
680 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
682 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
683 documented as working, but never had. Support all but $spam_report.
685 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
686 added for tls authenticator.
688 HS/03 Add perl_taintmode main config option
694 TL/01 When running the test suite, the README says that variables such as
695 no_msglog_check are global and can be placed anywhere in a specific
696 test's script, however it was observed that placement needed to be near
697 the beginning for it to behave that way. Changed the runtest perl
698 script to read through the entire script once to detect and set these
699 variables, reset to the beginning of the script, and then run through
700 the script parsing/test process like normal.
702 TL/02 The BSD's have an arc4random API. One of the functions to induce
703 adding randomness was arc4random_stir(), but it has been removed in
704 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
705 function when detected.
707 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
708 cause callback expansion.
710 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
711 syntax errors in an expansion can be treated as a string instead of
712 logging or causing an error, due to the internal use of bool_lax
713 instead of bool when processing it.
715 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
716 server certificates when making smtp deliveries.
718 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
720 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
722 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
724 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
725 Merged patch from Sebastian Wiedenroth.
727 JH/05 Fix results-pipe from transport process. Several recipients, combined
728 with certificate use, exposed issues where response data items split
729 over buffer boundaries were not parsed properly. This eventually
730 resulted in duplicates being sent. This issue only became common enough
731 to notice due to the introduction of connection certificate information,
732 the item size being so much larger. Found and fixed by Wolfgang Breyha.
734 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
735 size buffer was used, resulting in syntax errors when an expansion
738 JH/07 Add support for directories of certificates when compiled with a GnuTLS
739 version 3.3.6 or later.
741 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
742 is EXPERIMENTAL_EVENT, the main-configuration and transport options
743 both become "event_action", the variables become $event_name, $event_data
744 and $event_defer_errno. There is a new variable $verify_mode, usable in
745 routers, transports and related events. The tls:cert event is now also
746 raised for inbound connections, if the main configuration event_action
749 TL/06 In test suite, disable OCSP for old versions of openssl which contained
750 early OCSP support, but no stapling (appears to be less than 1.0.0).
752 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
753 server certificate names available under the smtp transport option
754 "tls_verify_cert_hostname" now do not permit multi-component wildcard
757 JH/10 Time-related extraction expansions from certificates now use the main
758 option "timezone" setting for output formatting, and are consistent
759 between OpenSSL and GnuTLS compilations. Bug 1541.
761 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
762 encoded parameter in the incoming message. Bug 1558.
764 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
765 include certificate info, eximon was claiming there were spoolfile
768 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
770 JH/14 Log delivery-related information more consistently, using the sequence
771 "H=<name> [<ip>]" wherever possible.
773 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
774 are problematic for Debian distribution, omit them from the release
777 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
779 JH/16 Fix string representation of time values on 64bit time_t architectures.
782 JH/17 Fix a null-indirection in certextract expansions when a nondefault
783 output list separator was used.
788 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
789 checkers that were complaining about end of non-void function with no
792 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
793 This was a regression introduced in 4.83 by another bugfix.
795 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
797 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
798 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
804 TF/01 Correctly close the server side of TLS when forking for delivery.
806 When a message was received over SMTP with TLS, Exim failed to clear up
807 the incoming connection properly after forking off the child process to
808 deliver the message. In some situations the subsequent outgoing
809 delivery connection happened to have the same fd number as the incoming
810 connection previously had. Exim would try to use TLS and fail, logging
811 a "Bad file descriptor" error.
813 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
814 utilities have not been installed.
816 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
817 temporary space as the ACL may create new global variables.
819 TL/01 LDAP support uses per connection or global context settings, depending
820 upon the detected version of the libraries at build time.
822 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
823 to extract and use the src ip:port in logging and expansions as if it
824 were a direct connection from the outside internet. PPv2 support was
825 updated based on HAProxy spec change in May 2014.
827 JH/02 Add ${listextract {number}{list}{success}{fail}}.
829 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
830 Properly escape header and check for NULL return.
832 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
835 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
837 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
838 characters in header names, implemented as a verify condition.
839 Contributed by Michael Fischer v. Mollard.
841 TL/05 Rename SPF condition results err_perm and err_temp to standardized
842 results permerror and temperror. Previous values are deprecated but
843 still accepted. In a future release, err_perm and err_temp will be
844 completely removed, which will be a backward incompatibility if the
845 ACL tests for either of these two old results. Patch contributed by
846 user bes-internal on the mailing list.
848 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
850 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
851 selectors, in both main and reject logs.
853 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
856 JH/07 Add malware type "sock" for talking to simple daemon.
858 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
860 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
861 routers/transports under cutthrough routing.
863 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
864 numbers. Touch up "bool" conditional to keep the same definition.
866 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
868 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
870 JH/12 Expand items in router/transport headers_add or headers_remove lists
871 individually rather than the list as a whole. Bug 1452.
873 Required for reasonable handling of multiple headers_ options when
874 they may be empty; requires that headers_remove items with embedded
875 colons must have them doubled (or the list-separator changed).
877 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
878 view the policy declared in the DMARC record. Currently, $dmarc_status
879 is a combined value of both the record presence and the result of the
882 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
884 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
885 dnslookup router and the smtp transport (applying to the forward
888 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
889 of ldap servers used for a specific lookup. Patch provided by Heiko
892 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
893 New variable $lookup_dnssec_authenticated for observability.
895 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
896 Patch submitted by Lars Timman.
898 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
900 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
901 Requires trusted mode and valid format message id, aborts otherwise.
902 Patch contributed by Heiko Schlichting.
904 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
905 certextract with support for various fields. Bug 1358.
907 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
908 is requested by default, modifiable by smtp transport option
911 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
912 operate on certificate variables to give certificate fingerprints
913 Also new ${sha256:cert_variable}.
915 JH/23 The PRDR feature is moved from being Experimental into the mainline.
917 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
918 Christian Aistleitner.
920 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
922 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
923 file. Patch from Wolfgang Breyha.
925 JH/25 Expand the coverage of the delivery $host and $host_address to
926 client authenticators run in verify callout. Bug 1476.
928 JH/26 Port service names are now accepted for tls_on_connect_ports, to
929 align with daemon_smtp_ports. Bug 72.
931 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
932 support and error reporting did not work properly.
934 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
935 and is readable. Patch from Andrew Colin Kissa.
937 TL/14 Enhance documentation of ${run expansion and how it parses the
938 commandline after expansion, particularly in the case when an
939 unquoted variable expansion results in an empty value.
941 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
943 PP/02 Fix internal collision of T_APL on systems which support RFC3123
944 by renaming away from it. Addresses GH issue 15, reported by
947 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
949 TL/15 SECURITY: prevent double expansion in math comparison functions
950 (can expand unsanitized data). Not remotely exploitable.
957 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
959 PP/02 Make -n do something, by making it not do something.
960 When combined with -bP, the name of an option is not output.
962 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
965 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
966 $sender_host_name and config options to manage this, and basic check
969 PP/05 DSCP support for outbound connections and control modifier for inbound.
971 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
972 (Only plugin which currently uses this is kerberos4, which nobody should
973 be using, but we should make it available and other future plugins might
974 conceivably use it, even though it would break NAT; stuff *should* be
975 using channel bindings instead).
977 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
978 name; added for Sendmail compatibility; requires admin caller.
979 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
980 just ignore it); requires trusted caller.
981 Also parse but ignore: -Ac -Am -X<logfile>
984 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
986 TL/02 Add +smtp_confirmation as a default logging option.
988 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
989 Patch by Magnus Holmgren from 2007-02-20.
991 TL/04 Bugzilla 1281 - Spec typo.
992 Bugzilla 1283 - Spec typo.
993 Bugzilla 1290 - Spec grammar fixes.
995 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
997 TL/06 Add Experimental DMARC support using libopendmarc libraries.
999 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1000 mailing list by by Dmitry Isaikin.
1002 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1004 JH/02 Support "G" suffix to numbers in ${if comparisons.
1006 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1008 NM/01 Bugzilla 1197 - Spec typo
1009 Bugzilla 1196 - Spec examples corrections
1011 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1013 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1014 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1017 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1018 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1020 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1021 "acl {{name}{arg}...}", and optional args on acl condition
1024 JH/05 Permit multiple router/transport headers_add/remove lines.
1026 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1028 JH/07 Avoid using a waiting database for a single-message-only transport.
1029 Performance patch from Paul Fisher. Bugzilla 1262.
1031 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1034 JH/09 Add $headers_added variable, with content from use of ACL modifier
1035 add_header (but not yet added to the message). Bugzilla 199.
1037 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1038 Pulled from Bugzilla 817 by Wolfgang Breyha.
1040 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1042 (nb: this is the same fix as in Exim 4.80.1)
1044 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1047 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1050 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1052 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1053 advertise SMTP AUTH mechanism to us, instead of a generic
1054 protocol violation error. Also, make Exim more robust to bad
1055 data from the Dovecot auth socket.
1057 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1059 When a queue runner is handling a message, Exim first routes the
1060 recipient addresses, during which it prunes them based on the retry
1061 hints database. After that it attempts to deliver the message to
1062 any remaining recipients. It then updates the hints database using
1065 So if a recipient address works intermittently, it can get repeatedly
1066 deferred at routing time. The retry hints record remains fresh so the
1067 address never reaches the final cutoff time.
1069 This is a fairly common occurrence when a user is bumping up against
1070 their storage quota. Exim had some logic in its local delivery code
1071 to deal with this. However it did not apply to per-recipient defers
1072 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1074 This change adds a proper retry rule check during routing so that the
1075 final cutoff time is checked against the message's age. We only do
1076 this check if there is an address retry record and there is not a
1077 domain retry record; this implies that previous attempts to handle
1078 the address had the retry_use_local_parts option turned on. We use
1079 this as an approximation for the destination being like a local
1080 delivery, as in LMTP.
1082 I suspect this new check makes the old local delivery cutoff check
1083 redundant, but I have not verified this so I left the code in place.
1085 TF/02 Correct gecos expansion when From: is a prefix of the username.
1087 Test 0254 submits a message to Exim with the header
1091 When I ran the test suite under the user fanf2, Exim expanded
1092 the header to contain my full name, whereas it should have added
1093 a Resent-Sender: header. It erroneously treats any prefix of the
1094 username as equal to the username.
1096 This change corrects that bug.
1098 GF/01 DCC debug and logging tidyup
1099 Error conditions log to paniclog rather than rejectlog.
1100 Debug lines prefixed by "DCC: " to remove any ambiguity.
1102 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1104 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1105 Bug spotted by Jeremy Harris; was flawed since initial commit.
1106 Would have resulted in OCSP responses post-SNI triggering an Exim
1107 NULL dereference and crash.
1109 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1111 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1112 Bug detection, analysis and fix by Samuel Thibault.
1113 Bugzilla 1331, Debian bug #698092.
1115 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1117 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1118 Server implementation by Todd Lyons, client by JH.
1119 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1120 config variable "prdr_enable" controls whether the server
1121 advertises the facility. If the client requests PRDR a new
1122 acl_data_smtp_prdr ACL is called once for each recipient, after
1123 the body content is received and before the acl_smtp_data ACL.
1124 The client is controlled by both of: a hosts_try_prdr option
1125 on the smtp transport, and the server advertisement.
1126 Default client logging of deliveries and rejections involving
1127 PRDR are flagged with the string "PRDR".
1129 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1130 fclose(). Diagnosis by Todd Lyons.
1132 PP/17 Update configure.default to handle IPv6 localhost better.
1133 Patch by Alain Williams (plus minor tweaks).
1136 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1137 This is now consistent with GnuTLS, and is now documented: the
1138 previous undocumented portable approach to treating the option as
1139 unset was to force an expansion failure. That still works, and
1140 an empty string is now equivalent.
1142 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1143 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1144 not performing validation itself.
1146 PP/20 Added force_command boolean option to pipe transport.
1147 Patch from Nick Koston, of cPanel Inc.
1149 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1152 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1154 PP/21 Fix eximon continuous updating with timestamped log-files.
1155 Broken in a format-string cleanup in 4.80, missed when I repaired the
1156 other false fix of the same issue.
1157 Report and fix from Heiko Schlichting.
1160 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1161 Report from Prashanth Katuri.
1163 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1164 It's SecureTransport, so affects any MacOS clients which use the
1165 system-integrated TLS libraries, including email clients.
1167 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1168 using a MIME ACL for non-SMTP local injection.
1169 Report and assistance in diagnosis by Warren Baker.
1171 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1173 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1175 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1176 last id that failed so it may be referenced in subsequent ACL's.
1178 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1181 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1182 ldap library initialization, allowing self-signed CA's to be
1183 used. Also properly sets require_cert option later in code by
1184 using NULL (global ldap config) instead of ldap handle (per
1185 session). Bug diagnosis and testing by alxgomz.
1187 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1188 the src/util/ subdirectory.
1190 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1191 renamed to Transport Post Delivery Action by Jeremy Harris, as
1194 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1195 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1196 redis_servers = needs to be configured which will be used by the redis
1197 lookup. Patch from Warren Baker, of The Packet Hub.
1199 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1201 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1202 hostname or reverse DNS when processing a host list. Used suggestions
1203 from multiple comments on this bug.
1205 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1207 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1208 Missed a few lines, added it to make the runtest require no keyboard
1211 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1212 contains upper case chars. Make router use caseful_local_part.
1214 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1215 support when GnuTLS has been built with p11-kit.
1221 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1223 This, or similar/improved, will also be change PP/11 of 4.82.
1229 PP/01 Handle short writes when writing local log-files.
1230 In practice, only affects FreeBSD (8 onwards).
1231 Bugzilla 1053, with thanks to Dmitry Isaikin.
1233 NM/01 Bugzilla 949 - Documentation tweak
1235 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1238 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1240 PP/02 Implemented gsasl authenticator.
1242 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1244 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1245 `pkg-config foo` for cflags/libs.
1247 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1248 with rest of GSASL and with heimdal_gssapi.
1250 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1251 `pkg-config foo` for cflags/libs for the TLS implementation.
1253 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1254 properties get this fed in as external SSF. A number of robustness
1255 and debugging improvements to the cyrus_sasl authenticator.
1257 PP/08 cyrus_sasl server now expands the server_realm option.
1259 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1260 Patch by Jeremy Harris.
1262 PP/10 Added dbmjz lookup type.
1264 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1266 PP/12 MAIL args handles TAB as well as SP, for better interop with
1267 non-compliant senders.
1268 Analysis and variant patch by Todd Lyons.
1270 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1271 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1272 Patch from Dirk Mueller <dmueller@suse.com>
1274 PP/13 tls_peerdn now print-escaped for spool files.
1275 Observed some $tls_peerdn in wild which contained \n, which resulted
1276 in spool file corruption.
1278 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1279 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1280 or write after TLS renegotiation, which otherwise led to messages
1283 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1284 as a tracking header (ie: a signed header comes before the signature).
1285 Patch from Wolfgang Breyha.
1287 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1288 comma-sep list; embedded commas doubled.
1290 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1292 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1294 Report and patch from Dmitry Banschikov.
1296 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1297 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1298 failures. We appear to now support TLS1.1+ with Exim.
1300 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1301 lets Exim select keys and certificates based upon TLS SNI from client.
1302 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1303 before an outbound SMTP session. New log_selector, +tls_sni.
1305 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1306 NULL dereference. Report and patch from Alun Jones.
1308 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1309 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1310 Not seeing resolver debug output on NetBSD, but suspect this is a
1311 resolver implementation change.
1313 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1314 Left warnings. Added "eximon gdb" invocation mode.
1316 PP/21 Defaulting "accept_8bitmime" to true, not false.
1318 PP/22 Added -bw for inetd wait mode support.
1320 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1321 locate the relevant includes and libraries. Made this the default.
1323 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1324 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1326 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1327 This may cause build issues on older platforms.
1329 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1330 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1331 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1332 Added SNI support via GnuTLS too.
1333 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1335 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1337 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1338 Applied second patch from Janne, implementing suggestion to default
1339 multiple-strings-in-record handling to match SPF spec.
1341 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1343 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1344 read-only, out of scope).
1345 Patch from Wolfgang Breyha, report from Stuart Northfield.
1347 PP/29 Fix three issues highlighted by clang analyser static analysis.
1348 Only crash-plausible issue would require the Cambridge-specific
1349 iplookup router and a misconfiguration.
1350 Report from Marcin Mirosław.
1352 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1354 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1355 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1356 As part of this, removing so much warning spew let me fix some minor
1357 real issues in debug logging.
1359 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1360 assignment on my part. Fixed.
1362 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1363 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1364 Janne Snabb (who went above and beyond: thank you).
1366 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1367 string otherwise requires a connection and a bunch more work and it's
1368 relatively easy to get wrong. Should also expose TLS library linkage
1371 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1372 64-bit ${eval} (JH/03).
1374 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1375 GNU libc to support some of the 64-bit stuff, should not lead to
1376 conflicts. Defined before os.h is pulled in, so if a given platform
1377 needs to override this, it can.
1379 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1380 protection layer was required, which is not implemented.
1381 Bugzilla 1254, patch from Wolfgang Breyha.
1383 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1384 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1385 tls_dhparam take prime identifiers. Also unbreak combination of
1386 OpenSSL+DH_params+TLSSNI.
1388 PP/39 Disable SSLv2 by default in OpenSSL support.
1394 PP/01 Solaris build fix for Oracle's LDAP libraries.
1395 Bugzilla 1109, patch from Stephen Usher.
1397 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1399 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1402 TF/02 Fix a couple more cases where we did not log the error message
1403 when unlink() failed. See also change 4.74-TF/03.
1405 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1406 lock up or crash if it happened to be inside a call to libc when it
1407 got a SIGUSR1 from exiwhat.
1409 The SIGUSR1 handler appends the current process status to the process
1410 log which is later printed by exiwhat. It used to use the general
1411 purpose logging code to do this, but several functions it calls are
1412 not safe for signals.
1414 The new output code in the SIGUSR1 handler is specific to the process
1415 log, and simple enough that it's easy to inspect for signal safety.
1416 Removing some special cases also simplifies the general logging code.
1417 Removing the spurious timestamps from the process log simplifies
1420 TF/04 Improved ratelimit ACL condition.
1422 The /noupdate option has been deprecated in favour of /readonly which
1423 has clearer semantics. The /leaky, /strict, and /readonly update modes
1424 are mutually exclusive. The update mode is no longer included in the
1425 database key; it just determines when the database is updated. (This
1426 means that when you upgrade Exim will forget old rate measurements.)
1428 Exim now checks that the per_* options are used with an update mode that
1429 makes sense for the current ACL. For example, when Exim is processing a
1430 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1431 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1432 must specify per_mail/readonly. If you omit the update mode it defaults to
1433 /leaky where that makes sense (as before) or /readonly where required.
1435 The /noupdate option is now undocumented but still supported for
1436 backwards compatibility. It is equivalent to /readonly except that in
1437 ACLs where /readonly is required you may specify /leaky/noupdate or
1438 /strict/noupdate which are treated the same as /readonly.
1440 A useful new feature is the /count= option. This is a generalization
1441 of the per_byte option, so that you can measure the throughput of other
1442 aggregate values. For example, the per_byte option is now equivalent
1443 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1445 The per_rcpt option has been generalized using the /count= mechanism
1446 (though it's more complicated than the per_byte equivalence). When it is
1447 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1448 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1449 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1450 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1451 Note that using per_rcpt with a non-readonly update mode in more than
1452 one ACL will cause the recipients to be double-counted. (The per_mail
1453 and per_byte options don't have this problem.)
1455 The handling of very low rates has changed slightly. If the computed rate
1456 is less than the event's count (usually one) then this event is the first
1457 after a long gap. In this case the rate is set to the same as this event's
1458 count, so that the first message of a spam run is counted properly.
1460 The major new feature is a mechanism for counting the rate of unique
1461 events. The new per_addr option counts the number of different
1462 recipients that someone has sent messages to in the last time period. It
1463 behaves like per_rcpt if all the recipient addresses are different, but
1464 duplicate recipient addresses do not increase the measured rate. Like
1465 the /count= option this is a general mechanism, so the per_addr option
1466 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1467 example, measure the rate that a client uses different sender addresses
1468 with the options per_mail/unique=$sender_address. There are further
1469 details in the main documentation.
1471 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1473 TF/06 Removed a few PCRE remnants.
1475 TF/07 Automatically extract Exim's version number from tags in the git
1476 repository when doing development or release builds.
1478 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1479 Bugzilla 879. Patch from Paul Fisher.
1481 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1482 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1485 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1487 PP/05 Variable $av_failed, true if the AV scanner deferred.
1488 Bugzilla 1078. Patch from John Horne.
1490 PP/06 Stop make process more reliably on build failure.
1491 Bugzilla 1087. Patch from Heiko Schlittermann.
1493 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1494 Bugzilla 1089. Patch from Heiko Schlittermann.
1496 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1497 Bugzilla 1131. Patch from Holger Weiß.
1499 PP/09 Handle IPv6 addresses with SPF.
1500 Bugzilla 860. Patch from Wolfgang Breyha.
1502 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1504 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1507 PP/11 match_* no longer expand right-hand-side by default.
1508 New compile-time build option, EXPAND_LISTMATCH_RHS.
1509 New expansion conditions, "inlist", "inlisti".
1511 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1513 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1515 PP/14 fix log_write() format string regression from TF/03.
1516 Bugzilla 1152. Patch from Dmitry Isaikin.
1522 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1524 PP/02 Harmonised TLS library version reporting; only show if debugging.
1525 Layout now matches that introduced for other libraries in 4.74 PP/03.
1527 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1529 PP/04 New "dns_use_edns0" global option.
1531 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1534 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1535 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1537 TK/01 Updated PolarSSL code to 0.14.2.
1538 Bugzilla 1097. Patch from Andreas Metzler.
1540 PP/07 Catch divide-by-zero in ${eval:...}.
1541 Fixes bugzilla 1102.
1543 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1546 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1547 format-string attack -- SECURITY: remote arbitrary code execution.
1549 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1550 time unintentionally subject to list matching rules, letting the header
1551 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1552 arbitrary string expansion). This allowed for information disclosure.
1554 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1555 INT_MIN/-1 -- value coerced to INT_MAX.
1561 NM/01 Workaround for PCRE version dependency in version reporting
1564 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1565 This fixes portability to compilers other than gcc, notably
1566 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1568 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1569 makefiles for portability to HP-UX and POSIX correctness.
1571 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1572 Also via indented variable definition in the Makefile.
1573 (Debugging by Oliver Heesakkers).
1575 PP/02 Restore caching of spamd results with expanded spamd_address.
1576 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1578 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1579 Improves build reliability. Fix from: Frank Elsner
1581 NM/02 Fix wide character breakage in the rfc2047 coding
1582 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1584 NM/03 Allow underscore in dnslist lookups
1585 Fixes bug 1026. Patch from Graeme Fowler
1587 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1588 Code patches from Adam Ciarcinski of NetBSD.
1590 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1593 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1594 is logged, to avoid truncation. Patch from John Horne.
1596 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1597 Patch from Jakob Hirsch.
1599 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1600 SQL string expansion failure details.
1601 Patch from Andrey Oktyabrski.
1603 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1604 Patch from Simon Arlott.
1606 PP/09 New lookups functionality failed to compile on old gcc which rejects
1607 extern declarations in function scope.
1608 Patch from Oliver Fleischmann
1610 PP/10 Use sig_atomic_t for flags set from signal handlers.
1611 Check getgroups() return and improve debugging.
1612 Fixed developed for diagnosis in bug 927 (which turned out to be
1615 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1616 Patch from Mark Zealey.
1618 PP/12 Bugzilla 1056: Improved spamd server selection.
1619 Patch from Mark Zealey.
1621 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1622 Based on patch from Heiko Schlittermann.
1624 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1625 Patch from Uwe Doering, sign-off by Michael Haardt.
1627 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1628 variable declaration deep within a block. Bug and patch from
1631 PP/15 lookups-Makefile IRIX compatibility coercion.
1633 PP/16 Make DISABLE_DKIM build knob functional.
1635 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1636 Patch by Simon Arlott
1638 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1639 variable argument macros. Our copy now differs from upstream.
1645 TF/01 Failure to get a lock on a hints database can have serious
1646 consequences so log it to the panic log.
1648 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1649 controlled using the smtp_confirmation log selector.
1651 TF/03 Include the error message when we fail to unlink a spool file.
1653 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1654 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1655 for maintaining out-of-tree patches for some time.
1657 PP/01 Bugzilla 139: Documentation and portability issues.
1658 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1659 Handle per-OS dynamic-module compilation flags.
1661 PP/02 Let /dev/null have normal permissions.
1662 The 4.73 fixes were a little too stringent and complained about the
1663 permissions on /dev/null. Exempt it from some checks.
1664 Reported by Andreas M. Kirchwitz.
1666 PP/03 Report version information for many libraries, including
1667 Exim version information for dynamically loaded libraries. Created
1668 version.h, now support a version extension string for distributors
1669 who patch heavily. Dynamic module ABI change.
1671 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1672 privilege escalation vulnerability whereby the Exim run-time user
1673 can cause root to append content of the attacker's choosing to
1676 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1679 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1680 If dropping privileges for untrusted macros, we disabled normal logging
1681 on the basis that it would fail; for the Exim run-time user, this is not
1682 the case, and it resulted in successful deliveries going unlogged.
1683 Fixed. Reported by Andreas Metzler.
1689 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1690 only prepend for the Resent-* case. Fixes regression introduced in
1691 Exim 4.70 by NM/22 for Bugzilla 607.
1693 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1694 increasing numbers of administrators be bitten by this.
1696 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1698 PP/03 Bugzilla 994: added openssl_options main configuration option.
1700 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1702 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1704 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1706 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1707 without a peer certificate, leading to a segfault because of an
1708 assumption that peers always have certificates. Be a little more
1709 paranoid. Problem reported by Martin Tscholak.
1711 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1712 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1713 NB: ClamAV planning to remove STREAM in "middle of 2010".
1714 CL also introduces -bmalware, various -d+acl logging additions and
1715 more caution in buffer sizes.
1717 PP/09 Implemented reverse_ip expansion operator.
1719 PP/10 Bugzilla 937: provide a "debug" ACL control.
1721 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1723 PP/12 Bugzilla 973: Implement --version.
1725 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1727 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1729 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1731 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1732 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1733 ignore trailing whitespace.
1735 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1737 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1740 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1741 Notification from Dr Andrew Aitchison.
1743 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1744 ExtendedDetectionInfo response format.
1745 Notification from John Horne.
1747 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1750 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1751 XSL and documented dependency on system catalogs, with examples of how
1754 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1757 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1758 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1759 configuration file which is writeable by the Exim user or group.
1761 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1762 of configuration files to cover files specified with the -C option if
1763 they are going to be used with root privileges, not just the default
1766 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1767 option (effectively making it always true).
1769 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1770 files to be used while preserving root privileges.
1772 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1773 that rogue child processes cannot use them.
1775 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1776 run-time user, instead of root.
1778 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1779 Exim run-time user without dropping privileges.
1781 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1782 result string, instead of calling string_vformat() twice with the same
1785 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1786 for other users. Others should always drop root privileges if they use
1787 -C on the command line, even for a whitelisted configure file.
1789 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1791 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1797 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1798 $data_path, and $header_path variables; fixed documentation bugs and
1801 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1802 exipick to access non-standard spools, including the "frozen" queue
1805 NM/01 Bugzilla 965: Support mysql stored procedures.
1806 Patch from Alain Williams
1808 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1810 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1811 Patch from Andreas Metzler
1813 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1814 Patch from Kirill Miazine
1816 NM/05 Bugzilla 671: Added umask to procmail example.
1818 JJ/03 installed exipick 20100323.0, fixing doc bug
1820 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1821 directory. Notification and patch from Dan Rosenberg.
1823 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1825 TK/02 Improve log output when DKIM signing operation fails.
1827 MH/01 Treat the transport option dkim_domain as a colon separated
1828 list, not as a single string, and sign the message with each element,
1829 omitting multiple occurences of the same signer.
1831 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1832 Bugzilla 985, 986. Patch by Simon Arlott
1834 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1835 Patch by Simon Arlott
1837 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1838 MBX locking. Notification from Dan Rosenberg.
1844 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1846 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1848 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1850 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1852 NM/04 Bugzilla 915: Flag broken perl installation during build.
1858 TK/01 Added patch by Johannes Berg that expands the main option
1859 "spamd_address" if it starts with a dollar sign.
1861 TK/02 Write list of recipients to X-Envelope-Sender header when building
1862 the mbox-format spool file for content scanning (suggested by Jakob
1865 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1866 (http://www.dcc-servers.net/) support via dccifd. Activated by
1867 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1869 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1870 by Mark Daniel Reidel <mr@df.eu>.
1872 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1873 When building exim an external PCRE library is now needed -
1874 PCRE is a system library on the majority of modern systems.
1875 See entry on PCRE_LIBS in EDITME file.
1877 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1878 conversation. Added nologin parameter to request.
1879 Patch contributed by Kirill Miazine.
1881 TF/01 Do not log submission mode rewrites if they do not change the address.
1883 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1885 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1886 log files in place. Contributed by Roberto Lima.
1888 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1890 TF/03 Bugzilla 615: When checking the local_parts router precondition
1891 after a local_part_suffix or local_part_prefix option, Exim now
1892 does not use the address's named list lookup cache, since this
1893 contains cached lookups for the whole local part.
1895 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1896 Robert Millan. Documentation is in experimental-spec.txt.
1898 TF/04 Bugzilla 668: Fix parallel build (make -j).
1900 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1902 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1903 Patch provided by Jan Srzednicki.
1905 TF/05 Leading white space used to be stripped from $spam_report which
1906 wrecked the formatting. Now it is preserved.
1908 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1909 that they are available at delivery time.
1911 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1913 TF/08 TLS error reporting now respects the incoming_interface and
1914 incoming_port log selectors.
1916 TF/09 Produce a more useful error message if an SMTP transport's hosts
1917 setting expands to an empty string.
1919 NM/06 Bugzilla 744: EXPN did not work under TLS.
1920 Patch provided by Phil Pennock.
1922 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1923 Patch provided by Richard Godbee.
1925 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1926 acl_smtp_notquit, added index entry.
1928 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1929 Patch provided by Eugene Bujak.
1931 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1932 accept(). Patch provided by Maxim Dounin.
1934 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1935 Patch provided by Phil Pennock.
1937 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1939 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1940 Patch provided by Brad "anomie" Jorsch.
1942 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1943 Patch provided by Dean Brooks.
1945 TK/05 Add native DKIM support (does not depend on external libraries).
1947 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1948 Patch provided by Graeme Fowler.
1950 NM/16 Bugzilla 851: Documentation example syntax fix.
1952 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1954 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1957 NM/19 Bugzilla 745: TLS version reporting.
1958 Patch provided by Phil Pennock.
1960 NM/20 Bugzilla 167: bool: condition support.
1961 Patch provided by Phil Pennock.
1963 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1964 clients. Patch provided by Phil Pennock.
1966 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1967 Patch provided by Brad "anomie" Jorsch.
1969 NM/23 Bugzilla 687: Fix misparses in eximstats.
1970 Patch provided by Heiko Schlittermann.
1972 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1973 Patch provided by Heiko Schlittermann.
1975 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1976 plus update to original patch.
1978 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1980 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1981 Patch provided by David Brownlee.
1983 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1985 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1987 NM/30 Bugzilla 888: TLS documentation bugfixes.
1989 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1991 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1992 Unlike the original bugzilla I have changed all shell scripts in src tree.
1994 NM/33 Bugzilla 898: Transport filter timeout fix.
1995 Patch by Todd Rinaldo.
1997 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1998 Patch by Serge Demonchaux.
2000 NM/35 Bugzilla 39: Base64 decode bug fixes.
2001 Patch by Jakob Hirsch.
2003 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2005 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2007 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2009 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2015 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2016 ALT-N's libdkim that I have put here:
2017 http://duncanthrax.net/exim-experimental/
2019 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2020 were called 'true' and it seems that C99 defines that as a reserved
2021 keyword to be used with 'bool' variable types. That means you could
2022 not include C99-style headers which use bools without triggering
2023 build errors in sieve.c.
2025 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2026 as mailq or other aliases. Changed the --help handling significantly
2027 to do whats expected. exim_usage() emits usage/help information.
2029 SC/01 Added the -bylocaldomain option to eximstats.
2031 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2033 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2035 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2041 PH/01 Another patch from the Sieve maintainer.
2043 PH/02 When an IPv6 address is converted to a string for single-key lookup
2044 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2045 dots are used instead of colons so that keys in lsearch files need not
2046 contain colons. This was done some time before quoting was made available
2047 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2048 (notated using the quote facility) so as to distinguish them from IPv4
2049 keys. This meant that lookups for IP addresses in host lists did not work
2050 for iplsearch lookups.
2052 This has been fixed by arranging for IPv6 addresses to be expressed with
2053 colons if the lookup type is iplsearch. This is not incompatible, because
2054 previously such lookups could never work.
2056 The situation is now rather anomalous, since one *can* have colons in
2057 ordinary lsearch keys. However, making the change in all cases is
2058 incompatible and would probably break a number of configurations.
2060 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2063 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2064 conversion specification without a maximum field width, thereby enabling
2065 a rogue spamd server to cause a buffer overflow. While nobody in their
2066 right mind would setup Exim to query an untrusted spamd server, an
2067 attacker that gains access to a server running spamd could potentially
2068 exploit this vulnerability to run arbitrary code as the Exim user.
2070 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2071 $primary_hostname instead of what libspf2 thinks the hosts name is.
2073 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2074 a directory entry by the name of the lookup key. Previously, if a
2075 symlink pointed to a non-existing file or a file in a directory that
2076 Exim lacked permissions to read, a lookup for a key matching that
2077 symlink would fail. Now it is enough that a matching directory entry
2078 exists, symlink or not. (Bugzilla 503.)
2080 PH/03 The body_linecount and body_zerocount variables are now exported in the
2083 PH/04 Added the $dnslist_matched variable.
2085 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2086 This means they are set thereafter only if the connection becomes
2089 PH/06 Added the client_condition to authenticators so that some can be skipped
2090 by clients under certain conditions.
2092 PH/07 The error message for a badly-placed control=no_multiline_responses left
2093 "_responses" off the end of the name.
2095 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2097 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2098 (without spaces) instead of just copying the configuration text.
2100 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2102 PH/11 Added $max_received_linelength.
2104 PH/12 Added +ignore_defer and +include_defer to host lists.
2106 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2107 way in which PCRE > 7.0 is built.
2109 PH/14 Implemented queue_only_load_latch.
2111 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2112 MAIL command. The effect was to mangle the value on 64-bit systems.
2114 PH/16 Another patch from the Sieve maintainer.
2116 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2118 PH/18 If a system quota error occurred while trying to create the file for
2119 a maildir delivery, the message "Mailbox is full" was not appended to the
2120 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2121 only to a quota excession during the actual writing of the file.
2123 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2124 characters?) which causes problems in log lines. The DN values are now
2125 passed through string_printing() before being added to log lines.
2127 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2128 and InterBase are left for another time.)
2130 PH/21 Added message_body_newlines option.
2132 PH/22 Guard against possible overflow in moan_check_errorcopy().
2134 PH/23 POSIX allows open() to be a macro; guard against that.
2136 PH/24 If the recipient of an error message contained an @ in the local part
2137 (suitably quoted, of course), incorrect values were put in $domain and
2138 $local_part during the evaluation of errors_copy.
2144 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2145 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2148 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2149 issue a MAIL command.
2151 PH/02 In an ACL statement such as
2153 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2155 if a client was not listed at all, or was listed with a value other than
2156 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2157 the condition was not true (as it should be), so access was not denied.
2158 The bug was that the ! inversion was incorrectly passed on to the second
2159 item. This has been fixed.
2161 PH/03 Added additional dnslists conditions == and =& which are different from
2162 = and & when the dns lookup returns more than one IP address.
2164 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2165 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2167 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2168 FSYNC, which compiles an option called disable_fsync that allows for
2169 bypassing fsync(). The documentation is heavily laced with warnings.
2171 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2173 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2174 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2175 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2176 including adding "make clean"; (3) Added -fPIC when compiling the test
2177 dynamically loaded module, to get rid of a warning.
2179 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2180 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2181 = 0s. The bug is otherwise harmless.
2183 PH/07 There was a bug in the dovecot authenticator such that the value of
2184 $auth1 could be overwritten, and so not correctly preserved, after a
2185 successful authentication. This usually meant that the value preserved by
2186 the server_setid option was incorrect.
2188 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2190 PH/09 Installed PCRE release 7.0.
2192 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2193 run for batched SMTP input. It is now run at the start of every message
2194 in the batch. While fixing this I discovered that the process information
2195 (output by running exiwhat) was not always getting set for -bs and -bS
2196 input. This is fixed, and it now also says "batched" for BSMTP.
2198 PH/11 Added control=no_pipelining.
2200 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2201 patch, slightly modified), and move the expansion of helo_data till after
2202 the connection is made in the smtp transport (so it can use these
2205 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2207 PH/14 Added log_selector = +pid.
2209 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2211 PH/16 Add ${if forany and ${if forall.
2213 PH/17 Added dsn_from option to vary the From: line in DSNs.
2215 PH/18 Flush SMTP output before performing a callout, unless control =
2216 no_callout_flush is set.
2218 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2219 was true (the default) a successful delivery failed to delete the retry
2220 item, thus causing premature timeout of the address. The bug is now
2223 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2225 PH/21 Long custom messages for fakedefer and fakereject are now split up
2226 into multiline responses in the same way that messages for "deny" and
2227 other ACL rejections are.
2229 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2230 with slight modification.
2232 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2233 draft, changing the syntax and factoring some duplicate code".
2235 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2236 for deliveries of the second and subsequent messages over the same SMTP
2239 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2240 ${reduce, with only minor "tidies".
2242 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2244 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2245 expansion side effects.
2247 PH/27 When a message times out after an over-quota error from an Exim-imposed
2248 quota, the bounce message says "mailbox is full". This message was not
2249 being given when it was a system quota that was exceeded. It now should
2252 MH/03 Made $recipients available in local_scan(). local_scan() already has
2253 better access to the recipient list through recipients_list[], but
2254 $recipients can be useful in postmaster-provided expansion strings.
2256 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2257 in the case of a MAIL command with additional options following the
2258 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2259 were accidentally chopped off.
2261 PH/29 SMTP synchronization checks are implemented when a command is read -
2262 there is a check that no more input is waiting when there shouldn't be
2263 any. However, for some commands, a delay in an ACL can mean that it is
2264 some time before the response is written. In this time, more input might
2265 arrive, invalidly. So now there are extra checks after an ACL has run for
2266 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2267 pipelining has not been advertised.
2269 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2271 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2272 returned in the SMTP response when smtp_return_error_details was false.
2273 This has been fixed.
2275 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2276 instead of the C I/O that was originally supplied, because problems were
2277 reported on Solaris.
2279 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2280 Exim which did not show up earlier: it was assuming that a call to
2281 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2282 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2283 was a macro that became an assignment, so it seemed to work. This has
2284 changed to a proper function call with a void return, hence the compile
2285 error. Exim's code has been fixed.
2287 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2290 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2292 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2294 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2295 because it turns out to be too big or there is a timeout) there is no
2296 "Completed" line in the log. When some messages of this type were
2297 selected by exigrep, they were listed as "not completed". Others were
2298 picked up by some special patterns. I have improved the selection
2299 criteria to be more general.
2301 PH/38 The host_find_failed option in the manualroute router can now be set
2302 to "ignore", to completely ignore a host whose IP address cannot be
2303 found. If all hosts are ignored, the behaviour is controlled by the new
2304 host_all_ignored option.
2306 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2307 homing or because of multiple MX records with /mx) generated more than
2308 one IP address, and the following item turned out to be the local host,
2309 all the secondary addresses of the first item were incorrectly removed
2310 from the list, along with the local host and any following hosts (which
2311 is what is supposed to happen).
2313 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2314 whoever called Exim into the -H file. In the case of the daemon it was
2315 behaving confusingly. When first started, it used values for whoever
2316 started the daemon, but after a SIGHUP it used the Exim user (because it
2317 calls itself on a restart). I have changed the code so that it now always
2320 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2321 message are rejected with the same error (e.g. no authentication or bad
2322 sender address), and a DATA command is nevertheless sent (as can happen
2323 with PIPELINING or a stupid MUA), the error message that was given to the
2324 RCPT commands is included in the rejection of the DATA command. This is
2325 intended to be helpful for MUAs that show only the final error to their
2328 PH/42 Another patch from the Sieve maintainer.
2330 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2331 Eximstats - Fixed some broken HTML links and added missing column headers
2333 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2334 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2336 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2338 PH/43 Yet another patch from the Sieve maintainer.
2340 PH/44 I found a way to check for a TCP/IP connection going away before sending
2341 the response to the final '.' that terminates a message, but only in the
2342 case where the client has not sent further data following the '.'
2343 (unfortunately, this is allowed). However, in many cases there won't be
2344 any further data because there won't be any more messages to send. A call
2345 to select() can be used: if it shows that the input is "ready", there is
2346 either input waiting, or the socket has been closed. An attempt to read
2347 the next input character can distinguish the two cases. Previously, Exim
2348 would have sent an OK response which the client would never have see.
2349 This could lead to message repetition. This fix should cure that, at
2350 least in a lot of common cases.
2352 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2353 advertised in response to EHLO.
2359 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2360 fixed by 4.65/MH/01 (is this a record?) are fixed:
2362 (i) An empty string was always treated as zero by the numeric comparison
2363 operators. This behaviour has been restored.
2365 (ii) It is documented that the numeric comparison operators always treat
2366 their arguments as decimal numbers. This was broken in that numbers
2367 starting with 0 were being interpreted as octal.
2369 While fixing these problems I realized that there was another issue that
2370 hadn't been noticed. Values of message_size_limit (both the global option
2371 and the transport option) were treated as octal if they started with 0.
2372 The documentation was vague. These values are now always treated as
2373 decimal, and I will make that clear in the documentation.
2379 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2380 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2383 MH/01 Don't check that the operands of numeric comparison operators are
2384 integers when their expansion is in "skipping" mode (fixes bug
2385 introduced by 4.64-PH/07).
2387 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2388 child addresses, Exim now panics and dies. Previously, because the count
2389 is held in a short int, deliveries were likely to be lost. As such a
2390 large number of recipients for a single message is ridiculous
2391 (performance will be very, very poor), I have chosen to impose a limit
2392 rather than extend the field.
2398 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2399 leftover -K file (the existence of which was triggered by #402).
2400 While we were at it, introduced process PID as part of the -K
2401 filename. This should rule out race conditions when creating
2404 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2405 processing considerably. Previous code took too long for large mails,
2406 triggering a timeout which in turn triggers #401.
2408 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2409 in the DK code in transports.c. sendfile() is not really portable,
2410 hence the _LINUX specificness.
2412 TF/01 In the add_headers option to the mail command in an Exim filter,
2413 there was a bug that Exim would claim a syntax error in any
2414 header after the first one which had an odd number of characters
2417 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2418 callout verification, Exim cached a "reject" for the entire domain. This
2419 is correct for most verifications, but it is not correct for a recipient
2420 verification with use_sender or use_postmaster set, because in that case
2421 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2422 case of MAIL FROM:<> rejection from other early rejections (e.g.
2423 rejection of HELO). When verifying a recipient using a non-null MAIL
2424 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2425 Whatever the result of the callout, the value of the domain cache is
2426 left unchanged (for any other kind of callout, getting as far as trying
2427 RCPT means that the domain itself is ok).
2429 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2432 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2433 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2434 session, the daemon ignores ECONNECT errors and logs others; it now
2435 ignores EPIPE as well.
2437 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2438 (quoted-printable decoding).
2440 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2441 later the small subsequent patch to fix an introduced bug.
2443 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2445 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2447 PH/08 An error is now given if message_size_limit is specified negative.
2449 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2450 to be given (somewhat) arbitrary names.
2452 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2455 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2456 miscellaneous code fixes
2458 PH/10 Added the log_reject_target ACL modifier to specify where to log
2461 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2462 hostname. This is wrong, because it relates to the incoming message (and
2463 probably the interface on which it is arriving) and not to the outgoing
2464 callout (which could be using a different interface). This has been
2465 changed to use the value of the helo_data option from the smtp transport
2466 instead - this is what is used when a message is actually being sent. If
2467 there is no remote transport (possible with a router that sets up host
2468 addresses), $smtp_active_hostname is used.
2470 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2471 tweaks were necessary in order to get it to work (see also 21 below):
2472 (a) The code assumed that strncpy() returns a negative number on buffer
2473 overflow, which isn't the case. Replaced with Exim's string_format()
2475 (b) There were several signed/unsigned issues. I just did the minimum
2476 hacking in of casts. There is scope for a larger refactoring.
2477 (c) The code used strcasecmp() which is not a standard C function.
2478 Replaced with Exim's strcmpic() function.
2479 (d) The code set only $1; it now sets $auth1 as well.
2480 (e) A simple test gave the error "authentication client didn't specify
2481 service in request". It would seem that Dovecot has changed its
2482 interface. Fortunately there's a specification; I followed it and
2483 changed what the client sends and it appears to be working now.
2485 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2488 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2489 address A is aliased to B and C, where B exists and C does not. Without
2490 -v the output is "A verified" because verification stops after a
2491 successful redirection if more than one address is generated. However,
2492 with -v the child addresses are also verified. Exim was outputting "A
2493 failed to verify" and then showing the successful verification for C,
2494 with its parentage. It now outputs "B failed to verify", showing B's
2495 parentage before showing the successful verification of C.
2497 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2498 look up a TXT record in a specific list after matching in a combined
2501 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2502 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2503 they consult the DNS. I had assumed they would set it the way they
2504 wanted; and indeed my experiments on Linux seem to show that in some
2505 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2506 To be on the safe side, however, I have now made the interface to
2507 host_find_byname() similar to host_find_bydns(), with an argument
2508 containing the DNS resolver options. The host_find_byname() function now
2509 sets these options at its start, just as host_find_bydns() does. The smtp
2510 transport options dns_qualify_single and dns_search_parents are passed to
2511 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2512 of host_find_byname() use the default settings of RES_DEFNAMES
2513 (qualify_single) but not RES_DNSRCH (search_parents).
2515 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2516 spool_read_header() do less string testing, by means of a preliminary
2517 switch on the second character of optional "-foo" lines. (This is
2518 overdue, caused by the large number of possibilities that now exist.
2519 Originally there were few.) While I was there, I also converted the
2520 str(n)cmp tests so they don't re-test the leading "-" and the first
2521 character, in the hope this might squeeze out yet more improvement.
2523 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2524 flag allowing group syntax was set by the header_syntax check but not
2525 turned off, possible causing trouble later; (2) The flag was not being
2526 set at all for the header_verify test, causing "group"-style headers to
2527 be rejected. I have now set it in this case, and also caused header_
2528 verify to ignore an empty address taken from a group. While doing this, I
2529 came across some other cases where the code for allowing group syntax
2530 while scanning a header line wasn't quite right (mostly, not resetting
2531 the flag correctly in the right place). These bugs could have caused
2532 trouble for malformed header lines. I hope it is now all correct.
2534 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2535 with the "reply" argument non-NULL. The code, however (which originally
2536 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2537 but it didn't always do it. This confused somebody who was copying the
2538 code for some other use. I have removed all the tests.
2540 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2541 feature that was used to support insecure browsers during the U.S. crypto
2542 embargo. It requires special client support, and Exim is probably the
2543 only MTA that supported it -- and would never use it because real RSA is
2544 always available. This code has been removed, because it had the bad
2545 effect of slowing Exim down by computing (never used) parameters for the
2546 RSA_EXPORT functionality.
2548 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2549 authenticator to fail if there's a tab character in the incoming data
2550 (there should never be unless someone is messing about, as it's supposed
2551 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2552 if the connection is using TLS or if the remote IP is the same as the
2553 local IP, and the "valid-client-cert option" if a client certificate has
2556 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2557 authenticators. This can be used for authorization after authentication
2558 succeeds. (In the case of plaintext, it servers for both authentication
2561 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2562 if any retry times were supplied.
2564 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2565 connection, where there is no client IP address to check. In this
2566 situation, the verify now always succeeds.
2568 PH/25 Applied John Jetmore's -Mset patch.
2570 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2572 PH/27 In a string expansion for a processed (not raw) header when multiple
2573 headers of the same name were present, leading whitespace was being
2574 removed from all of them, but trailing whitespace was being removed only
2575 from the last one. Now trailing whitespace is removed from each header
2576 before concatenation. Completely empty headers in a concatenation (as
2577 before) are ignored.
2579 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2580 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2582 PH/29 [Removed. This was a change that I later backed out, and forgot to
2583 correct the ChangeLog entry (that I had efficiently created) before
2584 committing the later change.]
2586 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2587 address errors (4xx response to RCPT) over the same connection as other
2588 messages routed to the same hosts. Such deliveries are always "forced",
2589 so retry times are not inspected. This resulted in far too many retries
2590 for the affected addresses. The effect occurred only when there were more
2591 hosts than the hosts_max_try setting in the smtp transport when it had
2592 the 4xx errors. Those hosts that it had tried were not added to the list
2593 of hosts for which the message was waiting, so if all were tried, there
2594 was no problem. Two fixes have been applied:
2596 (i) If there are any address or message errors in an SMTP delivery, none
2597 of the hosts (tried or untried) are now added to the list of hosts
2598 for which the message is waiting, so the message should not be a
2599 candidate for sending over the same connection that was used for a
2600 successful delivery of some other message. This seems entirely
2601 reasonable: after all the message is NOT "waiting for some host".
2602 This is so "obvious" that I'm not sure why it wasn't done
2603 previously. Hope I haven't missed anything, but it can't do any
2604 harm, as the worst effect is to miss an optimization.
2606 (ii) If, despite (i), such a delivery is accidentally attempted, the
2607 routing retry time is respected, so at least it doesn't keep
2608 hammering the server.
2610 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2611 in ${readsocket because some servers need this prod.
2613 PH/32 Added some extra debug output when updating a wait-xxx database.
2615 PH/33 The hint "could be header name not terminated by colon", which has been
2616 given for certain expansion errors for a long time, was not being given
2617 for the ${if def:h_colon_omitted{... case.
2619 PH/34 The spec says: "With one important exception, whenever a domain list is
2620 being scanned, $domain contains the subject domain." There was at least
2621 one case where this was not true.
2623 PH/35 The error "getsockname() failed: connection reset by peer" was being
2624 written to the panic log as well as the main log, but it isn't really
2625 panic-worthy as it just means the connection died rather early on. I have
2626 removed the panic log writing for the ECONNRESET error when getsockname()
2629 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2630 runs only) independently of the message's sender address. This meant
2631 that, if the 4xx error was in fact related to the sender, a different
2632 message to the same recipient with a different sender could confuse
2633 things. In particular, this can happen when sending to a greylisting
2634 server, but other circumstances could also provoke similar problems.
2635 I have changed the default so that the retry time for these errors is now
2636 based a combination of the sender and recipient addresses. This change
2637 can be overridden by setting address_retry_include_sender=false in the
2640 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2641 remote server are returned as part of bounce messages. This was not
2642 happening for LMTP over a pipe (the lmtp transport), but now it is the
2643 same for both kinds of LMTP.
2645 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2646 in header lines that were in fact CNAMEs. This is no longer the case.
2648 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2649 and queue runs started by the daemon processed all messages. This has
2650 been fixed so that -R and -S can now usefully be given with -q<time>.
2652 PH/40 Import PCRE release 6.7 (fixes some bugs).
2654 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2656 PH/42 Give an error if -q is specified more than once.
2658 PH/43 Renamed the variables $interface_address and $interface_port as
2659 $received_ip_address and $received_port, to make it clear that these
2660 values apply to message reception, and not to the outgoing interface when
2661 a message is delivered. (The old names remain recognized, of course.)
2663 PH/44 There was no timeout on the connect() call when using a Unix domain
2664 socket in the ${readsocket expansion. There now is.
2666 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2667 be meaningful with "accept".
2669 SC/01 Eximstats V1.43
2670 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2672 SC/02 Eximstats V1.44
2673 Use a glob alias rather than an array ref in the generated
2674 parser. This improves both readability and performance.
2676 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2677 Collect SpamAssassin and rejection statistics.
2678 Don't display local sender or destination tables unless
2679 there is data to show.
2680 Added average volumes into the top table text output.
2682 SC/04 Eximstats V1.46
2683 Collect data on the number of addresses (recipients)
2684 as well as the number of messages.
2686 SC/05 Eximstats V1.47
2687 Added 'Message too big' to the list of mail rejection
2688 reasons (thanks to Marco Gaiarin).
2690 SC/06 Eximstats V1.48
2691 Mainlog lines which have GMT offsets and are too short to
2692 have a flag are now skipped.
2694 SC/07 Eximstats V1.49 (Alain Williams)
2695 Added the -emptyok flag.
2697 SC/08 Eximstats V1.50
2698 Fixes for obtaining the IP address from reject messages.
2700 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2701 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2702 whitespace changes from 4.64-PH/27
2704 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2707 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2708 are found, allow negative numbers in numeric criteria)
2710 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2712 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2713 to match changes made in 4.64-PH/43
2715 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2717 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2718 "domain" is actually an IP address, and force a failure. This locks out
2719 those revolvers/nameservers that support "A-for-A" lookups, in
2720 contravention of the specifications.
2722 PH/48 When a host name was looked up from an IP address, and the subsequent
2723 forward lookup of the name timed out, the host name was left in
2724 $sender_host_name, contrary to the specification.
2726 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2727 restricted to single-key lookups, Exim was not diagnosing an error if
2728 * or *@ was used with a query-style lookup.
2730 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2732 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2733 long ago, but noone interested enough thought of it. Let's just say that
2734 the "1.1" means that there are some new functions that weren't there at
2735 some point in the past.
2737 PH/51 Error processing for expansion failure of helo_data from an smtp
2738 transport during callout processing was broken.
2740 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2741 tested/used via the -bh/-bhc/-bs options.
2743 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2744 bug, fixed in subsequent PCRE releases).
2746 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2747 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2749 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2755 SC/01 Use a glob alias rather than an array ref in eximstats generated
2756 parser. This improves both readability and performance.
2758 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2759 Don't display local sender or destination tables in eximstats unless
2760 there is data to show.
2761 Added average volumes into the eximstats top table text output.
2763 SC/03 Collect data on the number of addresses (recipients) as well
2764 as the number of messages in eximstats.
2766 TF/01 Correct an error in the documentation for the redirect router. Exim
2767 does (usually) call initgroups() when daemonizing.
2769 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2770 with consistent privilege compared to when running as a daemon.
2772 TF/03 Note in the spec that $authenticated_id is not set for local
2773 submissions from trusted users.
2775 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2776 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2778 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2779 by adding some example configuration directives to the default
2780 configuration file. A little bit of work is required to uncomment the
2781 directives and define how usernames and passwords are checked, but
2782 there is now a framework to start from.
2784 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2785 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2786 without this. I don't know how relevant this is to other LDAP libraries.
2788 PH/02 Add the verb name to the "unknown ACL verb" error.
2790 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2792 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2794 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2795 directory not expanded when it should be if an expanded home directory
2796 was set for the address (which is overridden by the transport).
2798 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2801 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2802 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2803 because it is too late at that time, and has no effect.
2805 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2806 security issue with \' (bugzilla #107). I could not use the
2807 PQescapeStringConn() function, because it needs a PGconn value as one of
2810 PH/08 When testing addresses using -bt, indicate those final addresses that
2811 are duplicates that would not cause an additional delivery. At least one
2812 person was confused, thinking that -bt output corresponded to deliveries.
2813 (Suppressing duplicates isn't a good idea as you lose the information
2814 about possibly different redirections that led to the duplicates.)
2816 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2817 systems where poll() doesn't work, in particular OS X.
2819 PH/10 Added more information to debugging output for retry time not reached.
2821 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2822 operations in malware.c.
2824 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2827 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2828 syslog_duplication set false, log lines that would normally be written
2829 both the the main log and to the reject log were not written to syslog at
2832 PH/14 In the default configuration, change the use of "message" in ACL warn
2833 statements to "add_header".
2835 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2836 not followed by a command (e.g. "seen endif").
2838 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2839 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2842 PH/17 Added extra conditions to the default value of delay_warning_condition
2846 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2847 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2848 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2851 The Auto-Submitted: and various List- headers are standardised, whereas I
2852 don't think Precedence: ever was.
2854 PH/18 Refactored debugging code in route_finduser() to show more information,
2855 in particular, the error code if getpwnam() issues one.
2857 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2858 This is apparently needed in addition to the PH/07 change above to avoid
2859 any possible encoding problems.
2861 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2862 but not after initializing Perl.
2864 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2865 output them only if debugging. By default they are written stderr,
2866 apparently, which is not desirable.
2868 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2871 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2874 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2876 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2877 authenticated or an ident call has been made. Suppress the default
2878 values for $authenticated_id and $authenticated_sender (but permit -oMai
2879 and -oMas) when testing with -bh.
2881 PH/24 Re-jigged the order of the tests in the default configuration so that the
2882 tests for valid domains and recipients precede the DNS black list and CSA
2883 tests, on the grounds that those ones are more expensive.
2885 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2886 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2887 This bug exists in every version of Exim that I still have, right back to
2890 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2891 However, an attempt to turn on case-sensitivity in a regex key by
2892 including (?-i) didn't work because the subject string was already
2893 lowercased, and the effects were non-intuitive. It turns out that a
2894 one-line patch can be used to allow (?-i) to work as expected.
2900 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2901 other effects) broke the use of negated acl sub-conditions.
2903 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2906 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2907 "Deny" causes Exim to reject the incoming connection with a 554 error.
2908 Unfortunately, if there is a major crisis, such as a disk failure,
2909 tcp-wrappers gives "deny", whereas what one would like would be some
2910 kind of temporary error. A kludge has been added to help with this.
2911 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2912 554 error is used if errno is still zero or contains ENOENT (which occurs
2913 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2916 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2918 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2919 errors. Otherwise a message that provokes a temporary error (when other
2920 messages do not) can cause a whole host to time out.
2922 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2923 addresses were routed directly to files or pipes from a redirect router.
2924 File deliveries just didn't batch; pipe deliveries might have suffered
2927 PH/06 A failure to get a lock for a hints database would erroneously always say
2928 "Failed to get write lock", even when it was really a read lock.
2930 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2931 of 0600. This has been changed to use the value of the lockfile_mode
2932 option (which defaults to 0600).
2934 PH/08 Applied small patch from the Sieve maintainer.
2936 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2937 folder from quota calculations, a direct delivery into this folder messed
2938 up the contents of the maildirsize file. This was because the regex was
2939 used only to exclude .Trash (or whatever) when the size of the mailbox
2940 was calculated. There was no check that a delivery was happening into an
2941 excluded directory. This bug has been fixed by ignoring all quota
2942 processing for deliveries into excluded directories.
2944 PH/10 Added the maildirfolder_create_regex option to appendfile.
2950 PH/01 The code for finding all the local interface addresses on a FreeBSD
2951 system running IPv6 was broken. This may well have applied to all BSD
2952 systems, as well as to others that have similar system calls. The broken
2953 code found IPv4 interfaces correctly, but gave incorrect values for the
2954 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2955 that it would not match correctly against @[] and not recognize the IPv6
2958 PH/02 The ipliteral router was not recognizing addresses of the form user@
2959 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2961 PH/03 Added disable_ipv6.
2963 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2964 decoded form, because it is most often used to construct To: headers
2965 lines in autoreplies, and the decoded form may well be syntactically
2966 invalid. However, $reply_address has leading white space removed, and all
2967 newlines turned into spaces so that the autoreply transport does not
2970 PH/05 If group was specified without a user on a router, and no group or user
2971 was specified on a transport, the group from the router was ignored.
2973 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2974 for visible compile-time settings that can be used to change these
2975 numbers, for those that want even more. Backwards compatibility with old
2976 spool files has been maintained. However, going back to a previous Exim
2977 release will lost any variables that are in spool files.
2979 PH/07 Two small changes when running in the test harness: increase delay when
2980 passing a TCP/IP connection to a new process, in case the original
2981 process has to generate a bounce, and remove special handling of
2982 127.0.0.2 (sic), which is no longer necessary.
2984 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2985 be the same on different OS.
2987 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2990 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2991 whether --show-vars was specified or not
2993 JJ/02 exipick: Added support for new ACL variable spool format introduced
2996 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2997 syntactically invalid From: or Reply-to: line, and a filter used this to
2998 generate an autoreply, and therefore failed to obtain an address for the
2999 autoreply, Exim could try to deliver to a non-existent relative file
3000 name, causing unrelated and misleading errors. What now happens is that
3001 it logs this as a hard delivery error, but does not attempt to create a
3004 PH/11 The exinext utility has a -C option for testing purposes, but although
3005 the given file was scanned by exinext itself; it wasn't being passed on
3006 when Exim was called.
3008 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3009 an end-of-file indication when reading a command response.
3011 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3012 compiled. In many other places in Exim, IPv6 addresses are always
3013 recognized, so I have changed this. It also means that IPv4 domain
3014 literals of the form [IPV4:n.n.n.n] are now always recognized.
3016 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3017 used if the router is not running as root, for example, when verifying at
3018 ACL time, or when using -bh. The debugging output from this situation was
3019 non-existent - all you got was a failure to exec. I have made two
3022 (a) Failures to set uid/gid, the current directory, or a process leader
3023 in a subprocess such as that created by queryprogram now generate
3024 suitable debugging output when -d is set.
3026 (b) The queryprogram router detects when it is not running as root,
3027 outputs suitable debugging information if -d is set, and then runs
3028 the subprocess without attempting to change uid/gid.
3030 PH/15 Minor change to Makefile for building test_host (undocumented testing
3033 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3034 additional section of a DNS packet that returns MX or SRV records.
3035 Instead, it always explicitly searches for A/AAAA records. This avoids
3036 major problems that occur when a DNS server includes only records of one
3037 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3038 fixed another bug: if SRV records were looked up and the corresponding
3039 address records were *not* found in the additional section, the port
3040 values from the SRV records were lost.
3042 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3043 using the correct key (the original address) when searching the retry
3044 rules in order to find which one to use for generating the retry hint.
3046 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3047 adding the default one. Similarly, if it contains a Reply-To: header, the
3048 errors_reply_to option, if set, is not used.
3050 PH/19 When calculating a retry time, Exim used to measure the "time since
3051 failure" by looking at the "first failed" field in the retry record. Now
3052 it does not use this if it is later than than the arrival time of the
3053 message. Instead it uses the arrival time. This makes for better
3054 behaviour in cases where some deliveries succeed, thus re-setting the
3055 "first failed" field. An example is a quota failure for a huge message
3056 when small messages continue to be delivered. Without this change, the
3057 "time since failure" will always be short, possible causing more frequent
3058 delivery attempts for the huge message than are intended.
3059 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3061 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3062 $1, $2, $3) because the numerical variables can be reset during some
3063 expansion items (e.g. "match"), thereby losing the authentication data.
3065 PH/21 Make -bV show the size of off_t variables so that the test suite can
3066 decide whether to run tests for quotas > 2G.
3068 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3069 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3070 filecount value is greater than 2G or if a quota value is greater than 2G
3071 on a system where the size of off_t is not greater than 4, a panic error
3074 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3075 never match. The debug and -bh output now contains an explicit error
3076 message indicating a malformed IPv4 address or mask.
3078 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3079 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3080 PH/23 above applies.
3082 PH/25 Do not write to syslog when running in the test harness. The only
3083 occasion when this arises is a failure to open the main or panic logs
3084 (for which there is an explicit test).
3086 PH/26 Added the /no_tell option to "control=freeze".
3088 PH/27 If a host name lookup failed very early in a connection, for example, if
3089 the IP address matched host_lookup and the reverse lookup yielded a name
3090 that did not have a forward lookup, an error message of the form "no IP
3091 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3092 could be logged. Now it outputs the IP address instead of "NULL".
3094 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3095 allows the sender and the authenticated sender to be set when
3096 submitting a message from within Exim. Since child_open_exim() is
3097 documented for local_scan(), the new function should be too.
3099 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3100 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3101 results in an empty string is now treated as unset.
3103 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3105 PH/31 Added sender_verify_fail logging option.
3107 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3108 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3109 tidied the source and removed it altogether.
3111 PH/33 When a queue run was abandoned because the load average was too high, a
3112 log line was always written; now it is written only if the queue_run log
3113 selector is set. In addition, the log line for abandonment now contains
3114 information about the queue run such as the pid. This is always present
3115 in "start" and "stop" lines but was omitted from the "abandon" line.
3117 PH/34 Omit spaces between a header name and the colon in the error message that
3118 is given when verify = headers_syntax fails (if there are lots of them,
3119 the message gets confusing).
3121 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3122 names, as there are now some PTR records that contain slashes. This check
3123 is only to protect against broken name servers that fall over on strange
3124 characters, so the fact that it applies to all lookups doesn't matter.
3126 PH/36 Now that the new test suite is complete, we can remove some of the
3127 special code in Exim that was needed for the old test suite. For example,
3128 sorting DNS records because real resolvers return them in an arbitrary
3129 order. The new test suite's fake resolver always returns records in the
3132 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3133 bounces) except when queue_only is set, to avoid logging races between
3134 the different processes.
3136 PH/38 Panic-die if .include specifies a non-absolute path.
3138 PH/39 A tweak to the "H" retry rule from its user.
3140 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3141 a label. They prevented compilation on older perls.
3143 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3144 a warning to be raised on newish perls.
3146 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3147 on queue. Changes to match documented behaviour of showing count of
3148 messages matching specified criteria.
3150 PH/40 Changed the default ident timeout from 30s to 5s.
3152 PH/41 Added support for the use of login_cap features, on those BSD systems
3153 that have them, for controlling the resources used by pipe deliveries.
3155 PH/42 The content-scanning code uses fopen() to create files in which to put
3156 message data. Previously it was not paying any attention to the mode of
3157 the files. Exim runs with umask(0) because the rest of the code creates
3158 files with open(), and sets the required mode explicitly. Thus, these
3159 files were ending up world-writeable. This was not a big issue, because,
3160 being within the spool directory, they were not world-accessible. I have
3161 created a function called modefopen, which takes an additional mode
3162 argument. It sets umask(777), creates the file, chmods it to the required
3163 mode, then resets the umask. All the relevant calls to fopen() in the
3164 content scanning code have been changed to use this function.
3166 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3167 to 24 hours. This avoids potential overflow problems when processing G
3168 and H retry rules. I suspect nobody ever tinkers with this value.
3170 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3172 PH/45 When the plaintext authenticator is running as a client, the server's
3173 challenges are checked to ensure they are valid base64 strings. By
3174 default, the authentication attempt is cancelled if an invalid string is
3175 received. Setting client_ignore_invalid_base64 true ignores these errors.
3176 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3177 they are received. Thus, the responses can be made to depend on the
3178 challenges. If an invalid string is ignored, an empty string is placed in
3181 PH/46 Messages that are created by the autoreply transport now contains a
3182 References: header, in accordance with RFCs 2822 and 3834.
3184 PH/47 Added authenticated_sender_force to the smtp transport.
3186 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3188 PH/49 Installed latest patch from the Sieve maintainer.
3190 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3191 was also set, the appendfile transport was unnecessarily scanning a
3192 directory of message files (e.g. for maildir delivery) to find the count
3193 of files (along with the size), even though it did not need this
3194 information. It now does the scan only if it needs to find either the
3195 size of the count of files.
3197 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3199 PH/52 Two bugs concerned with error handling when the smtp transport is
3202 (i) Exim was not creating retry information for temporary errors given
3203 for individual recipients after the DATA command when the smtp transport
3204 was used in LMTP mode. This meant that they could be retried too
3205 frequently, and not timed out correctly.
3207 (ii) Exim was setting the flag that allows error details to be returned
3208 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3209 recipients that were returned after the DATA command.
3211 PH/53 This is related to PH/52, but is more general: for any failing address,
3212 when detailed error information was permitted to be returned to the
3213 sender, but the error was temporary, then after the final timeout, only
3214 "retry timeout exceeded" was returned. Now it returns the full error as
3215 well as "retry timeout exceeded".
3217 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3218 do this, and (what is worse) MTAs that accept it.
3220 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3221 will now be deprecated.
3223 PH/56 New os.c-cygwin from the Cygwin maintainer.
3225 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3226 formats (previously only available in exim formats via -bpr, -bpru,
3227 and -bpra. Now also available in native and exiqgrep formats)
3229 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3230 with very large, slow to parse queues
3232 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3234 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3236 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3237 responses to authentication challenges, though it was showing the
3238 challenges; (ii) I've removed the CR characters from the debug output for
3241 PH/58 Allow for the insertion of a newline as well as a space when a string
3242 is turned into more than one encoded-word during RFC 2047 encoding. The
3243 Sieve code now uses this.
3245 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3246 data_4xx, lost_connection, tls_required.
3248 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3249 message was being sent as an SMTP response.
3251 PH/61 Add -l and -k options to exicyclog.
3253 PH/62 When verifying, if an address was redirected to one new address, so that
3254 verification continued, and the new address failed or deferred after
3255 having set something in $address_data, the value of $address_data was not
3256 passed back to the ACL. This was different to the case when no
3257 redirection occurred. The value is now passed back in both cases.
3259 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3260 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3261 use login_cap.h, so on its own it isn't the distinguishing feature. The
3262 new name refers directly to the setclassresources() function.
3264 PH/65 Added configuration files for NetBSD3.
3266 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3268 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3269 is preferred over IPv4.
3271 PH/68 The bounce_return_message and bounce_return_body options were not being
3272 honoured for bounces generated during the reception of non-SMTP messages.
3273 In particular, this applied to messages rejected by the ACL. This bug has
3274 been fixed. However, if bounce_return_message is true and bounce_return_
3275 body is false, the headers that are returned for a non-SMTP message
3276 include only those that have been read before the error was detected.
3277 (In the case of an ACL rejection, they have all been read.)
3279 PH/69 The HTML version of the specification is now built in a directory called
3280 spec_html instead of spec.html, because the latter looks like a path with
3281 a MIME-type, and this confuses some software.
3283 PH/70 Catch two compiler warnings in sieve.c.
3285 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3286 function verify_get_ident() calls ip_connect() to connect a socket, but
3287 if the "connect()" function timed out, ip_connect() used to close the
3288 socket. However, verify_get_ident() also closes the socket later, and in
3289 between Exim writes to the log, which may get opened at this point. When
3290 the socket was closed in ip_connect(), the log could get the same file
3291 descriptor number as the socket. This naturally causes chaos. The fix is
3292 not to close the socket in ip_connect(); the socket should be closed by
3293 the function that creates it. There was only one place in the code where
3294 this was missing, in the iplookup router, which I don't think anybody now
3295 uses, but I've fixed it anyway.
3297 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3298 well as to direct DNS lookups. Otherwise the handling of names in host
3299 lists is inconsistent and therefore confusing.
3305 PH/01 Two changes to the default runtime configuration:
3307 (1) Move the checks for relay_from_hosts and authenticated clients from
3308 after to before the (commented out) DNS black list checks.
3310 (2) Add control=submission to the relay_from_hosts and authenticated
3311 clients checks, on the grounds that messages accepted by these
3312 statements are most likely to be submissions.
3314 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3316 (1) Generate an error if the third argument for the ${prvs expansion is
3319 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3322 (3) Reset the variables that are obtained from the first argument of
3323 ${prvscheck and used in the second argument before leaving the code,
3324 because their memory is reclaimed, so using them afterwards may do
3327 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3328 one (it's much easier than Tom thought :-).
3330 (5) Because of (4), we can now allow for the use of $prvscheck_result
3331 inside the third argument.
3333 PH/03 For some reason, the default setting of PATH when running a command from
3334 a pipe transport was just "/usr/bin". I have changed it to
3337 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3338 anything to be listed in the output from -bV.
3340 PH/05 When a filter generated an autoreply, the entire To: header line was
3341 quoted in the delivery log line, like this:
3343 => >A.N.Other <ano@some.domain> <original@ddress> ...
3345 This has been changed so that it extracts the operative address. There
3346 may be more than one such address. If so, they are comma-separated, like
3349 => >ano@some.domain,ona@other.domain <original@ddress> ...
3351 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3352 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3353 not being looked up in the rDNS to get a host name, Exim was showing the
3354 IP address twice in Received: lines, even though the IP addresses were
3355 identical. For example:
3357 Received: from [1.2.3.4] (helo=[1.2.3.4])
3359 However, if the real host name was known, it was omitting the HELO data
3360 if it matched the actual IP address. This has been tidied up so that it
3361 doesn't show the same IP address twice.
3363 PH/07 When both +timestamp and +memory debugging was on, the value given by
3364 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3365 called by the debug printing, thereby overwriting the timestamp buffer.
3366 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3368 PH/08 When the original message was included in an autoreply transport, it
3369 always said "this is a copy of the message, including all the headers",
3370 even if body_only or headers_only was set. It now gives an appropriate
3373 PH/09 Applied a patch from the Sieve maintainer which:
3375 o fixes some comments
3376 o adds the (disabled) notify extension core
3377 o adds some debug output for the result of if/elsif tests
3378 o points to the current vacation draft in the documentation
3379 and documents the missing references header update
3383 o fixes a bug in processing the envelope test (when testing
3384 multiple envelope elements, the last element determined the
3387 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3388 Electronic Mail") by including:
3390 Auto-submitted: auto-generated
3392 in the messages that it generates (bounce messages and others, such as
3393 warnings). In the case of bounce messages for non-SMTP messages, there was
3394 also a typo: it was using "Auto_submitted" (underscore instead of
3395 hyphen). Since every message generated by Exim is necessarily in response
3396 to another message, thes have all been changed to:
3398 Auto-Submitted: auto-replied
3400 in accordance with these statements in the RFC:
3402 The auto-replied keyword:
3404 - SHOULD be used on messages sent in direct response to another
3405 message by an automatic process,
3407 - MUST NOT be used on manually-generated messages,
3409 - MAY be used on Delivery Status Notifications (DSNs) and Message
3410 Disposition Notifications (MDNs),
3412 - MUST NOT be used on messages generated by automatic or periodic
3413 processes, except for messages which are automatic responses to
3416 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3417 to the default Received: header definition.
3419 PH/12 Added log selector acl_warn_skipped (default on).
3421 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3422 variables because (a) they are in the wrong storage pool and (b) even if
3423 they were copied, it wouldn't work properly because of the caching.
3425 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3426 checking when decoding. Apparently there are clients that generate
3427 overlong encoded strings. Why am I not surprised?
3429 PH/15 If the first argument of "${if match_address" was not empty, but did not
3430 contain an "@" character, Exim crashed. Now it writes a panic log message
3431 and treats the condition as false.
3433 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3435 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3436 extension "envelope-auth". The code is finished and in agreement with
3437 other implementations, but there is no documentation so far and in fact,
3438 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3439 not changing the active code.
3441 Print executed "if" and "elsif" statements when debugging is used. This
3442 helps a great deal to understand what a filter does.
3444 Document more things not specified clearly in RFC3028. I had all this
3445 sorted out, when out of a sudden new issues came to my mind. Oops."
3447 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3450 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3451 canonical form (as documented). However, after a host name lookup from
3452 the IP address, check_host() was doing a simple string comparison with
3453 addresses acquired from the DNS when checking that the found name did
3454 have the original IP as one of its addresses. Since any found IPv6
3455 addresses are likely to be in abbreviated form, the comparison could
3456 fail. Luckily, there already exists a function for doing the comparison
3457 by converting both addresses to binary, so now that is used instead of
3458 the text comparison.
3460 PH/20 There was another similar case to PH/19, when a complete host name was
3461 given in a host list; looking up its IP address could give an abbreviated
3462 form, whereas the current host's name might or might not be abbreviated.
3463 The same fix has been applied.
3469 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3470 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3473 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3474 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3476 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3478 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3479 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3480 Also, for "Auto-Submitted", treat anything other than "no" as
3481 non-personal, in accordance with RFC 3834. (Previously it treated
3482 anything starting "auto-" as non-personal.)
3484 TF/01 The control=submission/name=... option had a problem with syntax
3485 errors if the name included a slash character. The /name= option
3486 now slurps the rest of the string, so it can include any characters
3487 but it must come last in the list of options (after /sender_retain
3490 PH/05 Some modifications to the interface to the fake nameserver for the new
3498 TK/01 Added the "success_on_redirect" address verification option. See
3499 NewStuff for rationale and an example.
3501 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3503 PH/02 Patch to exigrep to allow it to work on syslog lines.
3505 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3506 fread() to skip over the body file's header line, because in Cygwin the
3507 header line is locked and is inaccessible.
3509 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3510 co-exist for some time) to make it clear that it is the Exim ID that is
3511 referenced, not the Message-ID: header line.
3513 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3514 string_format() function, because snprintf() does not exist on all
3517 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3518 ${stat: expansion item.
3520 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3521 protocol synchronization error", to keep the pedants happy.
3523 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3524 well as for IRIX systems, when gcc is being used. See the host.c source
3527 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3529 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3532 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3533 transport and to the smtp transport in LMTP mode.
3535 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3537 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3538 run a filter in a subprocess. This could lead to confusion in subsequent
3539 lookups in the parent process. There should also be a search_tidyup() at
3540 the end of the subprocess.
3542 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3543 only if the host matched helo_try_verify_hosts, which caused the
3544 verification to occur when the EHLO/HELO command was issued. The ACL just
3545 tested the remembered result. Now, if a previous verification attempt has
3546 not happened, "verify = helo" does it there and then.
3548 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3550 TK/03 Fix log output including CR from clamd.
3552 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3553 exist provoked a memory error which could cause a segfault.
3555 PH/15 Installed PCRE 6.2
3557 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3559 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3560 of the problem. Specifically, suggested +O2 rather than +O1 for the
3563 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3565 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3566 record, the port was not correctly logged when the outgoing_port log
3567 selector was set (it logged the transort's default port).
3569 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3570 fallback_hosts, and "hosts" in the smtp transport.
3572 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3573 host errors such as "Connection refused".
3575 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3576 authentication with radiusclient 0.4.9:
3578 - Error returned from rc_read_config was caught wrongly
3579 - Username/password not passed on to radius server due to wrong length.
3581 The presumption is that some radiusclient API changes for 4.51/PH/17
3582 were not taken care of correctly. The code is still untested by me (my
3583 Linux distribution still has 0.3.2 of radiusclient), but it was
3584 contributed by a Radius user.
3586 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3587 expanding the "port" option of the smtp transport.
3589 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3590 while reading a MIME header. Thanks to Tom Hughes for a patch.
3592 PH/24 Include config.h inside local_scan.h so that configuration settings are
3595 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3596 that in an ACL for RCPT (for example), you can examine exactly what was
3599 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3600 commands, but it was not correctly comparing the address with the actual
3601 client host address. Thus, it would show the EHLO address in Received:
3602 header lines when this was not necessary.
3604 PH/27 Added the % operator to ${eval:}.
3606 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3607 it should be ignoring failures (because with -C, for example, it has lost
3608 privilege). It wasn't ignoring creation failures other than "already
3611 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3614 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3615 because an input line was too long, either on its own, or by virtue of
3616 too many continuations, the temporary file was not being removed, and the
3617 return code was incorrect.
3619 PH/31 Missing "BOOL" in function definition in filtertest.c.
3621 PH/32 Applied Sieve patches from the maintainer.
3623 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3625 PH/33 Added "verify = not_blind".
3627 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3628 Local/Makefile (with some defaults set). These are used in built scripts
3629 such as exicyclog, but they have never been used in the exim_install
3630 script (though there are many overriding facilities there). I have
3631 arranged that the exim_install script now takes note of these two
3634 PH/35 Installed configuration files for Dragonfly.
3636 PH/36 When a locally submitted message by a trusted user did not contain a
3637 From: header, and the sender address was obtained from -f or from an SMTP
3638 MAIL command, and the trusted user did not use -F to supply a sender
3639 name, $originator_name was incorrectly used when constructing a From:
3640 header. Furthermore, $originator_name was used for submission mode
3641 messages from external hosts without From: headers in a similar way,
3642 which is clearly wrong.
3644 PH/37 Added control=suppress_local_fixups.
3646 PH/38 When log_selector = +received_sender was set, and the addition of the
3647 sender made the log line's construction buffer exactly full, or one byte
3648 less than full, an overflow happened when the terminating "\n" was
3651 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3652 when the result of a list match is failure because a DNS lookup failed.
3654 PH/40 RM_COMMAND is now used in the building process.
3656 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3657 the "build-* directories that it finds.
3659 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3660 address in a domain literal was a prefix of an interface address.
3662 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3663 when verifying a sender address, unless rewrite_headers is false.
3665 PH/44 Wrote a long comment about why errors_to addresses are verified as
3666 recipients, not senders.
3668 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3669 the ratelimit ACL was added.
3671 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3673 PH/46 Added extra information about PostgreSQL errors to the error string.
3675 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3676 avoiding the need to install special zones in a real server. This is
3677 backwards compatible; if it can't find the fake resolver, it drops back.
3678 Thus, both old and new test suites can be run.
3680 TF/02 Added util/ratelimit.pl
3682 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3683 clock is set back in time.
3685 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3686 Candler <B.Candler@pobox.com>.
3688 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3689 OK for addresses that are the result of redirections.
3691 PH/48 A number of further additions for the benefit of the new test suite,
3692 including a fake gethostbyname() that interfaces to the fake DNS resolver
3695 TF/06 The fix for widen_domains has also been applied to qualify_single and
3696 search_parents which are the other dnslookup options that can cause
3699 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3702 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3704 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3705 service" if it encountered an MX record with an empty target hostname.
3706 The message is now "an MX or SRV record indicated no SMTP service".
3708 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3709 if the DNS of the sending site is misconfigured. This is quite a
3710 common situation. This change restores the behaviour of treating a
3711 helo verification defer as a failure.
3713 PH/51 If self=fail was set on a router, the bounce message did not include the
3714 actual error message.
3720 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3722 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3723 command itself ended in error, the underlying message about the transport
3724 filter timeout was being overwritten with the pipe command error. Now the
3725 underlying error message should be appended to the second error message.
3727 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3729 PH/02 Reduce the amount of output that "make" produces by default. Full output
3730 can still be requested.
3732 PH/03 The warning log line about a condition test deferring for a "warn" verb
3733 was being output only once per connection, rather than after each
3734 occurrence (because it was using the same function as for successful
3735 "warn" verbs). This seems wrong, so I have changed it.
3737 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3738 it should not have, which might have caused a crash in the right
3739 circumstances, but probably never did.
3741 PH/04 Installed a modified version of Tony Finch's patch to make submission
3742 mode fix the return path as well as the Sender: header line, and to
3743 add a /name= option so that you can make the user's friendly name appear
3746 TF/03 Added the control = fakedefer ACL modifier.
3748 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3749 Mark Lowes for thorough testing.
3751 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3753 TK/03 Merged latest SRS patch from Miles Wilton.
3755 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3756 with the definition in sysexits.h (which is #included earlier).
3757 Fortunately, Exim does not actually use EX_OK. The code used to try to
3758 preserve the sysexits.h value, by assuming that macro definitions were
3759 scanned for macro replacements. I have been disabused of this notion,
3760 so now the code just undefines EX_OK before #including unistd.h.
3762 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3763 in the smtp transport. When a block could not be written in a single
3764 write() function, the timeout was being re-applied to each part-write.
3765 This seems wrong - if the receiver was accepting one byte at a time it
3766 would take for ever. The timeout is now adjusted when this happens. It
3767 doesn't have to be particularly precise.
3769 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3770 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3772 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3773 without a domain if the check to <postmaster@domain> fails.
3775 SC/01 Eximstats: added -xls and the ability to specify output files
3776 (patch written by Frank Heydlauf).
3778 SC/02 Eximstats: use FileHandles for outputting results.
3780 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3782 SC/04 Eximstats: fixed display of large numbers with -nvr option
3784 SC/05 Eximstats: fixed merging of reports with empty tables.
3786 SC/06 Eximstats: added the -include_original_destination flag
3788 SC/07 Eximstats: removed tabs and trailing whitespace.
3790 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3792 TK/06 MBOX spool code: Add real "From " MBOX separator line
3793 so the .eml file is really in mbox format (even though
3794 most programs do not really care). Patch from Alex Miller.
3796 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3797 The latter is generated from $received_to and is only set if the
3798 message has one envelope recipient. SA can use these headers,
3799 obviously out-of-the-box. Patch from Alex Miller.
3801 PH/08 The ${def test on a variable was returning false if the variable's
3802 value was "0", contrary to what the specification has always said!
3803 The result should be true unless the variable is empty.
3805 PH/09 The syntax error of a character other than { following "${if
3806 def:variable_name" (after optional whitespace) was not being diagnosed.
3807 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3808 accidental colon was present, for example, could give incorrect results.
3810 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3811 result is used (not including appendfile, where other changes are about
3814 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3815 This involved changing a lot of size variables from int to off_t. It
3816 should work with maildirs and everything.
3818 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3819 spamd dying while we are connected to it.
3821 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3824 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3825 function for BDB 4.3.
3827 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3829 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3830 the sources for PCRE have been re-arranged and more files are now
3833 PH/15 The code I had for printing potentially long long variables in PH/11
3834 above was not the best (it lost precision). The length of off_t variables
3835 is now inspected at build time, and an appropriate printing format (%ld
3836 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3837 to be "long long int" or "long int". This is needed for the internal
3838 formatting function string_vformat().
3840 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3841 the configuration file to be ":syslog", then the script "guesses" where
3842 the logs files are, rather than using the compiled in default. In our
3843 case the guess is not the same as the compiled default, so the script
3844 suddenly stopped working when I started to use syslog. The patch checks
3845 to see if log_file_path is "". If so, it attempts to read it from exim
3846 with no configuration file to get the compiled in version, before it
3847 falls back to the previous guessing code."
3849 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3850 implementing BATV in an Exim configuration. See NewStuff for the gory
3853 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3854 Makefile that are specific to HP-UX.
3856 PH/18 If the "use_postmaster" option was set for a recipient callout together
3857 with the "random" option, the postmaster address was used as the MAIL
3858 FROM address for the random test, but not for the subsequent recipient
3859 test. It is now used for both.
3861 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3862 patch removes a few documentation additions to RFC 3028, because the
3863 latest draft now contains them. It adds the new en;ascii-case comparator
3864 and a new error check for 8bit text in MIME parts. Comparator and
3865 require names are now matched exactly. I enabled the subaddress
3866 extension, but it is not well tested yet (read: it works for me)."
3868 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3869 rework some of the code of TK/09 above to avoid the hardwired use of
3870 "%lld" and "long long". Replaced the call to snprintf() with a call to
3873 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3874 records point to non-existent hosts", "retry timeout exceeded", and
3875 "retry time not reached for any host after a long failure period".
3877 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3878 experimental DomainKeys support:
3880 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3881 (2) On an error such as an illegally used "control", the wrong name for
3882 the control was given.
3884 These problems did NOT occur unless DomainKeys support was compiled.
3886 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3888 PH/24 Added ${if match_ip condition.
3890 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3891 databases so that it will be absolutely obvious if a crash occurs in the
3892 DB library. This is a regular occurrence (often caused by mis-matched
3895 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3896 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3897 picked up on a user's system that detects such things. There doesn't seem
3898 to be a gcc warning option for this - only an attribute that has to be
3899 put on the function's prototype. It seems that in Fedora Core 4 they have
3900 set this on a number of new functions. No doubt there will be more in due
3903 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3904 specify a transport. However, if an address that was verified by such a
3905 router was the subject of a callout, Exim crashed because it tried to
3906 read the rcpt_include_affixes from the non-existent transport. Now it
3907 just assumes that the setting of that option is false. This bug was
3908 introduced by 4.51/PH/31.
3910 PH/28 Changed -d+all to exclude +memory, because that information is very
3911 rarely of interest, but it makes the output a lot bigger. People tend to
3912 do -d+all out of habit.
3914 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3915 code in os-type was giving problems when libc.so lives in lib64, like on
3918 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3919 aren't the modern standard, and it seems that some systems' include files
3920 don't always have them. Exim was already checking for some of the newer
3921 ones like T_AAAA, and defining it itself. I've added checks for all the
3922 record types that Exim uses.
3924 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3925 not automatically generating a new one, as it is supposed to. This
3926 prevented TLS from working. If the file did exist, but contained invalid
3927 data, a new version was generated, as expected. It was only the case of a
3928 non-existent file that was broken.
3930 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3931 with a change in libdomainkeys > 0.64.
3933 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3934 from DNS. If the selector record carries the flag, it now has
3935 precedence over the domain-wide flag.
3937 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3939 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3940 the use of a port name that isn't defined in /etc/services) occurred, the
3941 message was deferred as in a normal delivery, and thus remained on the
3942 spool, instead of being failed because of the mua_wrapper setting. This
3943 is now fixed, and I tidied up some of the mua_wrapper messages at the
3946 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3947 the messages in a hash of arrays rather than using individual hashes.
3948 This is a bit cleaner and results in dramatic memory savings, albeit
3949 at a slight CPU cost.
3951 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3952 as requested by Marc Sherman.
3954 SC/10 Eximstats: added histograms for user specified patterns as requested
3957 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3959 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3960 fopen() in the content-scanning modules that did not already have it.
3966 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3967 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3969 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3971 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3973 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3974 to test Sieve filters that use "vacation".
3976 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3977 that changes the way the GnuTLS parameters are stored in the cache file.
3978 The new format can be generated externally. For backward compatibility,
3979 if the data in the cache doesn't make sense, Exim assumes it has read an
3980 old-format file, and it generates new data and writes a new file. This
3981 means that you can't go back to an older release without removing the
3984 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3985 work if there are any delivery delays because "one_time" forces the
3986 parent to be marked "delivered", so its unseen clone is never tried
3987 again. For this reason, Exim now forbids the simultaneous setting of
3990 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3991 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3992 entry if you want to know the details. The fix, however, neglected to
3993 consider the case when local delivery batching is involved. The test for
3994 "previously delivered" was not happening when checking to see if an
3995 address could be batched with a previous (undelivered) one; under
3996 certain circumstances this could lead to multiple deliveries to the same
3999 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4000 in its include files, and this causes problems building Exim.
4002 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4003 header_syntax) but Exim was just ignoring anything given after a slash.
4004 In particular, this caused confusion with an attempt to use "verify =
4005 reverse_host_lookup/defer_ok". An error is now given when options are
4006 supplied for verify items that do not have them. (Maybe reverse_host_
4007 lookup should have a defer_ok option, but that's a different point.)
4009 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4010 defined by RFC 821) to 2048, because there were problems with some AUTH
4011 commands, and RFC 1869 says the size should be increased for extended
4012 SMTP commands that take arguments.
4014 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4017 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4018 "unknown" error; now it says that the functionality isn't in the binary.
4020 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4021 an address' error message when a string expansion fails (syntax or
4022 whatever). Otherwise the password may appear in the log. Following change
4023 PH/42 below, there is no longer a chance of it appearing in a bounce
4026 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4028 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4029 of its addresses was ever tried. (Bugzilla bug #2.)
4031 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4032 the result incorrectly in the debug output. (It correctly added a newline
4033 to what was transported.)
4035 TF/01 Added $received_time.
4037 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4038 commented out examples of how to interface to a virus scanner and to
4039 SpamAssassin. Also added commented examples of av_scanner and
4040 spamd_address settings.
4042 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4043 and controls are allowed in which ACLs. There were a couple of minor
4044 errors. Some of the entries in the conditions table (which is a table of
4045 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4046 negation of where the condition IS allowed.
4048 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4050 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4051 header file does not have a version number, so I've had to invent a new
4052 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4053 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4054 radiusclient), but it was contributed by a Radius user.
4056 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4057 files or directories, for OpenSSL.
4059 PH/19 When an Exim process that is running as root has to create an Exim log
4060 file, it does so in a subprocess that runs as exim:exim so as to get the
4061 ownership right at creation (otherwise, other Exim processes might see
4062 the file with the wrong ownership). There was no test for failure of this
4063 fork() call, which would lead to the process getting stuck as it waited
4064 for a non-existent subprocess. Forks do occasionally fail when resources
4065 run out. I reviewed all the other calls to fork(); they all seem to check
4068 PH/20 When checking for unexpected SMTP input at connect time (before writing
4069 the banner), Exim was not dealing correctly with a non-positive return
4070 from the read() function. If the client had disconnected by this time,
4071 the result was a log entry for a synchronization error with an empty
4072 string after "input=" when read() returned zero. If read() returned -1
4073 (an event I could not check), uninitialized data bytes were printed.
4074 There were reports of junk text (parts of files, etc) appearing after
4077 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4079 PH/22 Added support for macro redefinition, and (re)definition in between
4080 driver and ACL definitions.
4082 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4083 forgetting to use the resulting value; it was using the unexpanded value.
4085 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4086 hadn't been configured. The fix is from Juergen Kreileder, who
4087 understands it better than I do:
4089 "Here's what I see happening with three configured cyrus_sasl
4090 authenticators configured (plain, login, cram-md5):
4092 On startup auth_cyrus_sasl_init() gets called for each of these.
4093 This means three calls to sasl_listmech() without a specified mech_list.
4094 => SASL tests which mechs of all available mechs actually work
4095 => three warnings about OTP not working
4096 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4098 With the patch, sasl_listmech() also gets called three times. But now
4099 SASL's mech_list option is set to the server_mech specified in the the
4100 authenticator. Or in other words, the answer from sasl_listmech()
4101 gets limited to just the mech you're testing for (which is different
4103 => the return list contains just 'plain' or 'login', 'cram-md5' or
4104 nothing depending on the value of ob->server_mech.
4106 I've just tested the patch: Authentication still works fine,
4107 unavailable mechs specified in the exim configuration are still
4108 caught, and the auth.log warnings about OTP are gone."
4110 PH/25 When debugging is enabled, the contents of the command line are added
4111 to the debugging output, even when log_selector=+arguments is not
4114 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4115 answer is "GNU", and only if the return is "GNU/something" is the answer
4118 PH/27 $acl_verify_message is now set immediately after the failure of a
4119 verification in an ACL, and so is available in subsequent modifiers. In
4120 particular, the message can be preserved by coding like this:
4122 warn !verify = sender
4123 set acl_m0 = $acl_verify_message
4125 Previously, $acl_verify_message was set only while expanding "message"
4126 and "log_message" when a very denied access.
4128 PH/28 Modified OS/os.c-Linux with
4130 -#ifndef OS_LOAD_AVERAGE
4131 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4133 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4134 nomenclature these days.)
4136 PH/29 Installed patch from the Sieve maintainer that adds the options
4137 sieve_useraddress and sieve_subaddress to the redirect router.
4139 PH/30 In these circumstances:
4140 . Two addresses routed to the same list of hosts;
4141 . First host does not offer TLS;
4142 . First host accepts first address;
4143 . First host gives temporary error to second address;
4144 . Second host offers TLS and a TLS session is established;
4145 . Second host accepts second address.
4146 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4147 and peerdn, if requested) that were in fact used only for the second
4150 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4151 attention to any local part prefix or suffix that was matched by the
4152 router that accepted the address. It now behaves in the same way as it
4153 does for delivery: the affixes are removed from the local part unless
4154 rcpt_include_affixes is set on the transport.
4156 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4157 timeout during the DATA phase of an incoming message.
4159 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4160 applied a patch sanctioned by the Sieve maintainer.
4162 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4163 the uid or gid is negative. A case of a negative gid caused this to be
4164 noticed. The fix allows for either to be negative.
4166 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4167 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4170 PH/36 The change PH/12 above was broken. Fixed it.
4172 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4173 the grounds that routing the same address twice would always produce the
4174 same answer. This might have been true once, but it is certainly no
4175 longer true now. Routing a child address may depend on the previous
4176 routing that produced that child. Some complicated redirection strategies
4177 went wrong when messages had multiple recipients, and made Exim's
4178 behaviour dependent on the order in which the addresses were given.
4180 I have moved the duplicate checking until after the routing is complete.
4181 Exim scans the addresses that are assigned to local and remote
4182 transports, and removes any duplicates. This means that more work will be
4183 done, as duplicates will always all be routed, but duplicates are
4184 presumably rare, so I don't expect this is of any significance.
4186 For deliveries to pipes, files, and autoreplies, the duplicate checking
4187 still happens during the routing process, since they are not going to be
4190 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4191 It corrects a timeout issue with spamd. This is Ian's comment: "The
4192 background is that sometimes spamd either never reads data from a
4193 connection it has accepted, or it never writes response data. The exiscan
4194 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4195 blindly assumes that writes won't block so it may never time out."
4197 PH/39 Allow G after quota size as well as K and M.
4199 PH/40 The value set for $authenticated_id in an authenticator may not contain
4200 binary zeroes or newlines because the value is written to log lines and
4201 to spool files. There was no check on this. Now the value is run through
4202 the string_printing() function so that such characters are converted to
4203 printable escape sequences.
4205 PH/41 $message_linecount is a new variable that contains the total number of
4206 lines in the message. Compare $body_linecount, which is the count for the
4209 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4210 bounce and delay warning messages, except in certain special cases, which
4213 (a) An SMTP error message from a remote host;
4214 (b) A message specified in a :fail: redirection;
4215 (c) A message specified in a "fail" command in a system filter;
4216 (d) A message specified in a FAIL return from the queryprogram router;
4217 (e) A message specified by the cannot_route_message router option.
4219 In these cases only, Exim does include the error details in bounce and
4220 warning messages. There are also a few cases where bland messages such
4221 as "unrouteable address" or "local delivery error" are given.
4223 PH/43 $value is now also set for the "else" part of a ${run expansion.
4225 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4226 being worked on, but at least Exim now implements the latest version to
4229 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4230 process to complete was treated as a delivery failure, a timeout while
4231 writing the message to the pipe was logged, but erroneously treated as a
4232 successful delivery. Such timeouts include transport filter timeouts. For
4233 consistency with the overall process timeout, these timeouts are now
4234 treated as errors, giving rise to delivery failures by default. However,
4235 there is now a new Boolean option for the pipe transport called
4236 timeout_defer, which, if set TRUE, converts the failures into defers for
4237 both kinds of timeout. A transport filter timeout is now identified in
4240 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4241 systems where "make" and "gmake" are different, calling "gmake" at top
4242 level broke things. I've arranged for the value of $(MAKE) to be passed
4243 from the Makefile to this script so that it can call the same version of
4247 A note about Exim versions 4.44 and 4.50
4248 ----------------------------------------
4250 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4251 changes of various kinds. As a consequence, a big documentation update was
4252 needed. This delayed the release for rather longer than seemed good, especially
4253 in the light of a couple of (minor) security issues. Therefore, the changes
4254 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4255 release. So 4.44 and 4.50 are in effect two different branches that both start
4258 I have left the 4.50 change log unchanged; it contains all the changes since
4259 4.43. The change log for 4.44 is below; many of its items are identical to
4260 those for 4.50. This seems to be the most sensible way to preserve the
4261 historical information.
4267 1. Minor wording change to the doc/README.SIEVE file.
4269 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4270 computation of the current number of files was incorrect.
4272 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4273 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4276 4. Give more explanation in the error message when the command for a transport
4277 filter fails to execute.
4279 5. There are several places where Exim runs a non-Exim command in a
4280 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4281 was being done only for the command run by the queryprogram router. It is
4282 now done for all such subprocesses. The other cases are: ${run, transport
4283 filters, and the commands run by the lmtp and pipe transports.
4285 6. Added CONFIGURE_GROUP build-time option.
4287 7. Some older OS have a limit of 256 on the maximum number of file
4288 descriptors. Exim was using setrlimit() to set 1000 as a large value
4289 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4290 systems. I've change it so that if it can't get 1000, it tries for 256.
4292 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4293 was an oversight, and furthermore, ever since the addition of extra
4294 controls (e.g. 4.43/32), the checks on when to allow different forms of
4295 "control" were broken. There should now be diagnostics for all cases when a
4296 control that does not make sense is encountered.
4298 9. Added the /retain_sender option to "control=submission".
4300 10. $recipients is now available in the predata ACL (oversight).
4302 11. Tidy the search cache before the fork to do a delivery from a message
4303 received from the command line. Otherwise the child will trigger a lookup
4304 failure and thereby defer the delivery if it tries to use (for example) a
4305 cached ldap connection that the parent has called unbind on.
4307 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4308 of $address_data from the recipient verification was clobbered by the
4309 sender verification.
4311 13. The value of address_data from a sender verification is now available in
4312 $sender_address_data in subsequent conditions in the ACL statement.
4314 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4316 15. Added a new option "connect=<time>" to callout options, to set a different
4319 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4320 was its contents. (It was OK if the option was not defined at all.)
4322 17. A "Completed" log line is now written for messages that are removed from
4323 the spool by the -Mrm option.
4325 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4326 information about exactly what failed.
4328 19. Added -dd to debug only the daemon process.
4330 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4331 handles timeouts, both on the server side and network timeouts. Renamed the
4332 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4334 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4335 It is now set to "smtps".
4337 22. $host_address is now set to the target address during the checking of
4338 ignore_target_hosts.
4340 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4341 being passed; this would have caused $sender_host_name to have been used if
4342 matching the list had actually called for a host name (not very likely,
4343 since this list is usually IP addresses). A host name is now passed as
4346 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4347 code that specifies a non-restarting handler (typically sigaction() in
4348 modern systems) in an attempt to fix a rare and obscure crash bug.
4350 25. Narrowed the window for a race in the daemon that could cause it to ignore
4351 SIGCHLD signals. This is not a major problem, because they are used only to
4352 wake it up if nothing else does.
4354 26. A malformed maildirsize file could cause Exim to calculate negative values
4355 for the mailbox size or file count. Odd effects could occur as a result.
4356 The maildirsize information is now recalculated if the size or filecount
4359 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4360 support for a long time. Removed HAVE_SYS_VFS_H.
4362 28. Installed the latest version of exipick from John Jetmore.
4364 29. In an address list, if the pattern was not a regular expression, an empty
4365 subject address (from a bounce message) matched only if the pattern was an
4366 empty string. Non-empty patterns were not even tested. This was the wrong
4367 because it is perfectly reasonable to use an empty address as part of a
4368 database query. An empty address is now tested by patterns that are
4369 lookups. However, all the other forms of pattern expect the subject to
4370 contain a local part and a domain, and therefore, for them, an empty
4371 address still always fails if the pattern is not itself empty.
4373 30. Exim went into a mad DNS loop when attempting to do a callout where the
4374 host was specified on an smtp transport, and looking it up yielded more
4375 than one IP address.
4377 31. Re-factored the code for checking spool and log partition space into a
4378 function that finds that data and another that does the check. The former
4379 is then used to implement four new variables: $spool_space, $log_space,
4380 $spool_inodes, and $log_inodes.
4382 32. The RFC2047 encoding function was originally intended for short strings
4383 such as real names; it was not keeping to the 75-character limit for
4384 encoded words that the RFC imposes. It now respects the limit, and
4385 generates multiple encoded words if necessary. To be on the safe side, I
4386 have increased the buffer size for the ${rfc2047: expansion operator from
4389 33. It is now permitted to omit both strings after an "if" condition; if the
4390 condition is true, the result is "true". As before, when the second string
4391 is omitted, a false condition yields an empty string. This makes it less
4392 cumbersome to write custom ACL and router conditions.
4394 34. Failure to deliver a bounce message always caused it to be frozen, even if
4395 there was an errors_to setting on the router. The errors_to setting is now
4398 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4399 canonical form (fully expanded) before being placed in
4400 $sender_host_address.
4402 36. The table in the code that translates DNS record types into text (T_A to
4403 "A" for instance) was missing entries for NS and CNAME. It is just possible
4404 that this could have caused confusion if both these types were looked up
4405 for the same domain, because the text type is used as part of Exim's
4406 per-process caching. But the chance of anyone hitting this buglet seems
4409 37. The dnsdb lookup has been extended in a number of ways.
4411 (1) There is a new type, "zns", which walks up the domain tree until it
4412 finds some nameserver records. It should be used with care.
4414 (2) There is a new type, "mxh", which is like "mx" except that it returns
4415 just the host names, not the priorities.
4417 (3) It is now possible to give a list of domains (or IP addresses) to be
4418 looked up. The behaviour when one of the lookups defers can be
4419 controlled by a keyword.
4421 (4) It is now possible to specify the separator character for use when
4422 multiple records are returned.
4424 38. The dnslists ACL condition has been extended: it is now possible to supply
4425 a list of IP addresses and/or domains to be looked up in a particular DNS
4428 39. Added log_selector=+queue_time_overall.
4430 40. When running the queue in the test harness, wait just a tad after forking a
4431 delivery process, to get repeatability of debugging output.
4433 41. Include certificate and key file names in error message when GnuTLS fails
4434 to set them up, because the GnuTLS error message doesn't include the name
4435 of the failing file when there is a problem reading it.
4437 42. Allow both -bf and -bF in the same test run.
4439 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4441 44. The "Exiscan patch" is now merged into the mainline Exim source.
4443 45. Sometimes the final signoff response after QUIT could fail to get
4444 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4445 < 0 before doing a fflush(). This bug looks as though it goes back to the
4446 introduction of TLS in release 3.20, but "sometimes" must have been rare
4447 because the tests only now provoked it.
4449 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4450 (this can affect the format of dates).
4452 47. exim_tidydb, when checking for the continued existence of a message for
4453 which it has found a message-specific retry record, was not finding
4454 messages that were in split spool directories. Consequently, it was
4455 deleting retry records that should have stayed in existence.
4457 48. Steve fixed some bugs in eximstats.
4459 49. The SPA authentication driver was not abandoning authentication and moving
4460 on to the next authenticator when an expansion was forced to fail,
4461 contradicting the general specification for all authenticators. Instead it
4462 was generating a temporary error. It now behaves as specified.
4464 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4465 (the order specifies the preference for clients). The order is now AES256,
4466 AES128, 3DES, ARCFOUR128.
4468 51. Small patch to Sieve code - explicitly set From: when generating an
4471 52. Exim crashed if a remote delivery caused a very long error message to be
4472 recorded - for instance if somebody sent an entire SpamAssassin report back
4473 as a large number of 550 error lines. This bug was coincidentally fixed by
4474 increasing the size of one of Exim's internal buffers (big_buffer) that
4475 happened as part of the Exiscan merge. However, to be on the safe side, I
4476 have made the code more robust (and fixed the comments that describe what
4479 53. Now that there can be additional text after "Completed" in log lines (if
4480 the queue_time_overall log selector is set), a one-byte patch to exigrep
4481 was needed to allow it to recognize "Completed" as not the last thing in
4484 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4485 patch that reportedly fixes this has been added. I am not expert enough to
4486 create a test for it. This is what the patch creator wrote:
4488 "I found a little strange behaviour of ldap code when working with
4489 Windows 2003 AD Domain, where users was placed in more than one
4490 Organization Units. When I tried to give exim partial DN, the exit code
4491 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4492 But simultaneously result of request was absolutely normal ldap result,
4493 so I produce this patch..."
4495 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4496 REFERENCE, so I have modified the code to exclude the patch when that macro
4499 55. Some experimental protocols are using DNS PTR records for new purposes. The
4500 keys for these records are domain names, not reversed IP addresses. The
4501 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4502 leaves it alone. Component reversal etc. now happens only for IP addresses.
4505 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4507 57. Double the size of the debug message buffer (to 2048) so that more of very
4508 long debug lines gets shown.
4510 58. The exicyclog utility now does better if the number of log files to keep
4511 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4513 59. Two changes related to the smtp_active_hostname option:
4515 (1) $smtp_active_hostname is now available as a variable.
4516 (2) The default for smtp_banner uses $smtp_active_hostname instead
4517 of $primary_hostname.
4519 60. The host_aton() function is supposed to be passed a string that is known
4520 to be a valid IP address. However, in the case of IPv6 addresses, it was
4521 not checking this. This is a hostage to fortune. Exim now panics and dies
4522 if the condition is not met. A case was found where this could be provoked
4523 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4524 components; fortuitously, this particular loophole had already been fixed
4525 by change 4.50/55 above.
4527 If there are any other similar loopholes, the new check in host_aton()
4528 itself should stop them being exploited. The report I received stated that
4529 data on the command line could provoke the exploit when Exim was running as
4530 exim, but did not say which command line option was involved. All I could
4531 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4532 running as the user.
4535 61. There was a buffer overflow vulnerability in the SPA authentication code
4536 (which came originally from the Samba project). I have added a test to the
4537 spa_base64_to_bits() function which I hope fixes it.
4540 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4541 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4543 63. The daemon start-up calls getloadavg() while still root for those OS that
4544 need the first call to be done as root, but it missed one case: when
4545 deliver_queue_load_max is set with deliver_drop_privilege. This is
4546 necessary for the benefit of the queue runner, because there is no re-exec
4547 when deliver_drop_privilege is set.
4549 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4550 This has been fixed.
4552 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4553 was in use, was not putting the data itself into the right store pool;
4554 consequently, it could be overwritten for a subsequent message in the same
4555 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4558 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4560 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4561 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4562 the function were treating the return as a boolean value, which happened to
4563 work because 0=false and not-0=true, but is not correct code.
4565 68. The host_aton() function was not handling scoped IPv6 addresses (those
4566 with, for example, "%eth0" on the end) correctly.
4568 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4569 negated items (that is, ~something) in unsigned ints. Some compilers
4570 apparently mutter when there is no cast.
4572 70. If an address verification called from an ACL failed, and did not produce a
4573 user-specific message (i.e. there was only a "system" message), nothing was
4574 put in $acl_verify_message. In this situation, it now puts the system
4577 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4578 session; change 4.31/43 added the unwanted input to the log line - except
4579 that it did not do this in the start of session case. It now does.
4581 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4582 This is wrong and can cause the other end to generate a synchronization
4583 error if it is another Exim or anything else that does the synchronization
4584 check. A QUIT command is no longer sent after a timeout.
4586 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4587 during host lookups.
4589 74. The defer_ok option of callout verification was not working if it was used
4590 when verifying addresses in header lines, that is, for this case:
4592 verify = header_sender/callout=defer_ok
4594 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4595 those file descriptors could be used for SMTP connections. If anything
4596 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4597 could be sent to the SMTP client, causing chaos. The daemon now opens
4598 stdin, stdout, and stderr to /dev/null when it puts itself into the
4601 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4602 log by default. The user can override this with suitable Perl magic.
4604 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4605 the log message when discard triggers, was not working for the DATA ACL or
4606 for the non-SMTP ACL.
4608 78. Error message wording change in sieve.c.
4610 79. If smtp_accept_max_per_host was set, the number of connections could be
4611 restricted to fewer than expected, because the daemon was trying to set up
4612 a new connection before checking whether the processes handling previous
4613 connections had finished. The check for completed processes is now done
4614 earlier. On busy systems, this bug wouldn't be noticed because something
4615 else would have woken the daemon, and it would have reaped the completed
4618 80. If a message was submitted locally by a user whose login name contained one
4619 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4620 It caused a spool format error. I have fixed the spool reading code. A
4621 related problem was that the "from" clause in the Received: line became
4622 illegal because of the space(s). It is now covered by ${quote_local_part.
4624 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4627 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4629 83. Patch from Sieve maintainer to fix unterminated string problem in
4630 "vacation" handling.
4632 84. Some minor changes to the Linux configuration files to help with other
4633 OS variants using glibc.
4635 85. One more patch for Sieve to update vacation handling to latest spec.
4638 ----------------------------------------------------
4639 See the note above about the 4.44 and 4.50 releases.
4640 ----------------------------------------------------
4646 1. Change 4.43/35 introduced a bug that caused file counts to be
4647 incorrectly computed when quota_filecount was set in an appendfile
4650 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4651 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4654 3. Give more explanation in the error message when the command for a transport
4655 filter fails to execute.
4657 4. There are several places where Exim runs a non-Exim command in a
4658 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4659 was being done only for the command run by the queryprogram router. It is
4660 now done for all such subprocesses. The other cases are: ${run, transport
4661 filters, and the commands run by the lmtp and pipe transports.
4663 5. Some older OS have a limit of 256 on the maximum number of file
4664 descriptors. Exim was using setrlimit() to set 1000 as a large value
4665 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4666 systems. I've change it so that if it can't get 1000, it tries for 256.
4668 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4669 was an oversight, and furthermore, ever since the addition of extra
4670 controls (e.g. 4.43/32), the checks on when to allow different forms of
4671 "control" were broken. There should now be diagnostics for all cases when a
4672 control that does not make sense is encountered.
4674 7. $recipients is now available in the predata ACL (oversight).
4676 8. Tidy the search cache before the fork to do a delivery from a message
4677 received from the command line. Otherwise the child will trigger a lookup
4678 failure and thereby defer the delivery if it tries to use (for example) a
4679 cached ldap connection that the parent has called unbind on.
4681 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4682 of $address_data from the recipient verification was clobbered by the
4683 sender verification.
4685 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4686 was its contents. (It was OK if the option was not defined at all.)
4688 11. A "Completed" log line is now written for messages that are removed from
4689 the spool by the -Mrm option.
4691 12. $host_address is now set to the target address during the checking of
4692 ignore_target_hosts.
4694 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4695 being passed; this would have caused $sender_host_name to have been used if
4696 matching the list had actually called for a host name (not very likely,
4697 since this list is usually IP addresses). A host name is now passed as
4700 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4701 code that specifies a non-restarting handler (typically sigaction() in
4702 modern systems) in an attempt to fix a rare and obscure crash bug.
4704 15. Narrowed the window for a race in the daemon that could cause it to ignore
4705 SIGCHLD signals. This is not a major problem, because they are used only to
4706 wake it up if nothing else does.
4708 16. A malformed maildirsize file could cause Exim to calculate negative values
4709 for the mailbox size or file count. Odd effects could occur as a result.
4710 The maildirsize information is now recalculated if the size or filecount
4713 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4714 support for a long time. Removed HAVE_SYS_VFS_H.
4716 18. Updated exipick to current release from John Jetmore.
4718 19. Allow an empty sender to be matched against a lookup in an address list.
4719 Previously the only cases considered were a regular expression, or an
4722 20. Exim went into a mad DNS lookup loop when doing a callout where the
4723 host was specified on the transport, if the DNS lookup yielded more than
4726 21. The RFC2047 encoding function was originally intended for short strings
4727 such as real names; it was not keeping to the 75-character limit for
4728 encoded words that the RFC imposes. It now respects the limit, and
4729 generates multiple encoded words if necessary. To be on the safe side, I
4730 have increased the buffer size for the ${rfc2047: expansion operator from
4733 22. Failure to deliver a bounce message always caused it to be frozen, even if
4734 there was an errors_to setting on the router. The errors_to setting is now
4737 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4738 canonical form (fully expanded) before being placed in
4739 $sender_host_address.
4741 24. Updated eximstats to version 1.33
4743 25. Include certificate and key file names in error message when GnuTLS fails
4744 to set them up, because the GnuTLS error message doesn't include the name
4745 of the failing file when there is a problem reading it.
4747 26. Expand error message when OpenSSL has problems setting up cert/key files.
4750 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4751 (this can affect the format of dates).
4753 28. exim_tidydb, when checking for the continued existence of a message for
4754 which it has found a message-specific retry record, was not finding
4755 messages that were in split spool directories. Consequently, it was
4756 deleting retry records that should have stayed in existence.
4758 29. eximstats updated to version 1.35
4759 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4760 1.35 - bugfix such that pie charts by volume are generated correctly
4762 30. The SPA authentication driver was not abandoning authentication and moving
4763 on to the next authenticator when an expansion was forced to fail,
4764 contradicting the general specification for all authenticators. Instead it
4765 was generating a temporary error. It now behaves as specified.
4767 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4768 (the order specifies the preference for clients). The order is now AES256,
4769 AES128, 3DES, ARCFOUR128.
4771 31. Small patch to Sieve code - explicitly set From: when generating an
4774 32. Exim crashed if a remote delivery caused a very long error message to be
4775 recorded - for instance if somebody sent an entire SpamAssassin report back
4776 as a large number of 550 error lines. This bug was coincidentally fixed by
4777 increasing the size of one of Exim's internal buffers (big_buffer) that
4778 happened as part of the Exiscan merge. However, to be on the safe side, I
4779 have made the code more robust (and fixed the comments that describe what
4782 33. Some experimental protocols are using DNS PTR records for new purposes. The
4783 keys for these records are domain names, not reversed IP addresses. The
4784 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4785 leaves it alone. Component reversal etc. now happens only for IP addresses.
4788 34. The host_aton() function is supposed to be passed a string that is known
4789 to be a valid IP address. However, in the case of IPv6 addresses, it was
4790 not checking this. This is a hostage to fortune. Exim now panics and dies
4791 if the condition is not met. A case was found where this could be provoked
4792 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4793 components; fortuitously, this particular loophole had already been fixed
4794 by change 4.50/55 or 4.44/33 above.
4796 If there are any other similar loopholes, the new check in host_aton()
4797 itself should stop them being exploited. The report I received stated that
4798 data on the command line could provoke the exploit when Exim was running as
4799 exim, but did not say which command line option was involved. All I could
4800 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4801 running as the user.
4804 35. There was a buffer overflow vulnerability in the SPA authentication code
4805 (which came originally from the Samba project). I have added a test to the
4806 spa_base64_to_bits() function which I hope fixes it.
4809 36. The daemon start-up calls getloadavg() while still root for those OS that
4810 need the first call to be done as root, but it missed one case: when
4811 deliver_queue_load_max is set with deliver_drop_privilege. This is
4812 necessary for the benefit of the queue runner, because there is no re-exec
4813 when deliver_drop_privilege is set.
4815 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4816 was in use, was not putting the data itself into the right store pool;
4817 consequently, it could be overwritten for a subsequent message in the same
4818 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4821 38. Sometimes the final signoff response after QUIT could fail to get
4822 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4823 < 0 before doing a fflush(). This bug looks as though it goes back to the
4824 introduction of TLS in release 3.20, but "sometimes" must have been rare
4825 because the tests only now provoked it.
4831 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4832 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4833 Now it returns an integer. A number of calls were still expecting a T/F
4834 return. Fortuitously, in all cases, the tests worked in OK situations,
4835 which is the norm. However, things would have gone wrong on any write
4836 failures on the smtp file descriptor. This function is used when sending
4837 messages over SMTP and also when doing verify callouts.
4839 2. When Exim is called to do synchronous delivery of a locally submitted
4840 message (the -odf or -odi options), it no longer closes stderr before doing
4843 3. Implemented the mua_wrapper option.
4845 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4847 5. Implemented the functions header_remove(), header_testname(),
4848 header_add_at_position(), and receive_remove_recipient(), and exported them
4851 6. If an ACL "warn" statement specified the addition of headers, Exim already
4852 inserted X-ACL-Warn: at the start if there was no header name. However, it
4853 was not making this test for the second and subsequent header lines if
4854 there were newlines in the string. This meant that an invalid header could
4855 be inserted if Exim was badly configured.
4857 7. Allow an ACL "warn" statement to add header lines at the start or after all
4858 the Received: headers, as well as at the end.
4860 8. Added the rcpt_4xx retry error code.
4862 9. Added postmaster_mailfrom=xxx to callout verification option.
4864 10. Added mailfrom=xxxx to the callout verification option, for verify=
4867 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4868 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4870 12. Inserted some casts to stop certain compilers warning when using pointer
4871 differences as field lengths or precisions in printf-type calls (mostly
4872 affecting debugging statements).
4874 13. Added optional readline() support for -be (dynamically loaded).
4876 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4877 same clock tick as a message's arrival, so that its received time was the
4878 same as the "first fail" time on the retry record, and that message
4879 remained on the queue past the ultimate address timeout, every queue runner
4880 would try a delivery (because it was past the ultimate address timeout) but
4881 after another failure, the ultimate address timeout, which should have then
4882 bounced the address, did not kick in. This was a "< instead of <=" error;
4883 in most cases the first failure would have been in the next clock tick
4884 after the received time, and all would be well.
4886 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4887 being recognized when the domain list was tested by the match_domain
4888 condition in an expansion string.
4890 16. Added the ${str2b64: operator.
4892 17. Exim was always calling setrlimit() to set a large limit for the number of
4893 processes, without checking whether the existing limit was already
4894 adequate. (It did check for the limit on file descriptors.) Furthermore,
4895 errors from getrlimit() and setrlimit() were being ignored. Now they are
4896 logged to the main and panic logs, but Exim does carry on, to try to do its
4897 job under whatever limits there are.
4899 18. Imported PCRE 5.0.
4901 19. Trivial typo in log message " temporarily refused connection" (the leading
4904 20. If the log selector return_path_on_delivery was set and an address was
4905 redirected to /dev/null, the delivery process crashed because it assumed
4906 that a return path would always be set for a "successful" delivery. In this
4907 case, the whole delivery is bypassed as an optimization, and therefore no
4910 21. Internal re-arrangement: the function for sending a challenge and reading
4911 a response while authentication was assuming a zero-terminated challenge
4912 string. It's now changed to take a pointer and a length, to allow for
4913 binary data in such strings.
4915 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4917 23. Exim was not respecting finduser_retries when seeking the login of the
4918 uid under which it was called; it was always trying 10 times. (The default
4919 setting of finduser_retries is zero.) Also, it was sleeping after the final
4920 failure, which is pointless.
4922 24. Implemented tls_on_connect_ports.
4924 25. Implemented acl_smtp_predata.
4926 26. If the domain in control=submission is set empty, Exim assumes that the
4927 authenticated id is a complete email address when it generates From: or
4928 Sender: header lines.
4930 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4931 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4932 chown and chgrp in /bin and hostname in /usr/bin.
4934 28. Exim was keeping the "process log" file open after each use, just as it
4935 does for the main log. This opens the possibility of it remaining open for
4936 long periods when the USR1 signal hits a daemon. Occasional processlog
4937 errors were reported, that could have been caused by this. Anyway, it seems
4938 much more sensible not to leave this file open at all, so that is what now
4941 29. The long-running daemon process does not normally write to the log once it
4942 has entered its main loop, and it closes the log before doing so. This is
4943 so that log files can straightforwardly be renamed and moved. However,
4944 there are a couple of unusual error situations where the daemon does write
4945 log entries, and I had neglected to close the log afterwards.
4947 30. The text of an SMTP error response that was received during a remote
4948 delivery was being truncated at 512 bytes. This is too short for some of
4949 the long messages that one sometimes sees. I've increased the limit to
4952 31. It is now possible to make retry rules that apply only when a message has a
4953 specific sender, in particular, an empty sender.
4955 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4956 it possible to be selective about when SMTP synchronization is enforced.
4958 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4960 32. Implemented hosts_connection_nolog.
4962 33. Added an ACL for QUIT.
4964 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4967 35. Added mailbox_size and mailbox_filecount to appendfile.
4969 36. Added control = no_multiline_responses to ACLs.
4971 37. There was a bug in the logic of the code that waits for the clock to tick
4972 in the case where the clock went backwards by a substantial amount such
4973 that the microsecond fraction of "now" was more than the microsecond
4974 fraction of "then" (but the whole seconds number was less).
4976 38. Added support for the libradius Radius client library this is found on
4977 FreeBSD (previously only the radiusclient library was supported).
4983 1. When certain lookups returned multiple values in the form name=value, the
4984 quoting of the values was not always being done properly. Specifically:
4985 (a) If the value started with a double quote, but contained no whitespace,
4987 (b) If the value contained whitespace other than a space character (i.e.
4988 tabs or newlines or carriage returns) it was not quoted.
4989 This fix has been applied to the mysql and pgsql lookups by writing a
4990 separate quoting function and calling it from the lookup code. The fix
4991 should probably also be applied to nisplus, ibase and oracle lookups, but
4992 since I cannot test any of those, I have not disturbed their existing code.
4994 2. A hit in the callout cache for a specific address caused a log line with no
4995 reason for rejecting RCPT. Now it says "Previous (cached) callout
4996 verification failure".
4998 3. There was an off-by-one bug in the queryprogram router. An over-long
4999 return line was truncated at 256 instead of 255 characters, thereby
5000 overflowing its buffer with the terminating zero. As well as fixing this, I
5001 have increased the buffer size to 1024 (and made a note to document this).
5003 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5004 when Exim is waiting for an SMTP response from a remote server, Exim
5005 restarts its select() call on the socket, thereby resetting its timeout.
5006 This is not a problem when such interrupts are rare. Somebody set up a cron
5007 job to run exiwhat every 2 minutes, which is less than the normal select()
5008 timeout (5 or 10 minutes). This meant that the select() timeout never
5009 kicked in because it was always reset. I have fixed this by comparing the
5010 time when an interrupt arrives with the time at the start of the first call
5011 to select(). If more time than the timeout has elapsed, the interrupt is
5012 treated as a timeout.
5014 5. Some internal re-factoring in preparation for the addition of Sieve
5015 extensions (by MH). In particular, the "personal" test is moved to a
5016 separate function, and given an option for scanning Cc: and Bcc: (which is
5017 not set for Exim filters).
5019 6. When Exim created an email address using the login of the caller as the
5020 local part (e.g. when creating a From: or Sender: header line), it was not
5021 quoting the local part when it contained special characters such as @.
5023 7. Installed new OpenBSD configuration files.
5025 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5026 try to make them clearer.
5028 9. Callout options, other than the timeout value, were being ignored when
5029 verifying sender addresses in header lines. For example, when using
5031 verify = header_sender/callout=no_cache
5033 the cache was (incorrectly) being used.
5035 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5036 only the Cygwin environment.
5038 11. When return_path_on_delivery was set as a log selector, if different remote
5039 addresses in the same message used different return paths and parallel
5040 remote delivery occurred, the wrong values would sometimes be logged.
5041 (Whenever a remote delivery process finished, the return path value from
5042 the most recently started remote delivery process was logged.)
5044 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5045 lines when AUTH and/or TLS are in use. This is the "received protocol"
5046 field. Exim used to use "asmtp" for authenticated SMTP, without any
5047 indication (in the protocol name) for TLS use. Now it follows the RFC and
5048 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5049 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5050 names appear in log lines as well as in Received: header lines.
5052 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5053 extensions, and comparison tests, and to fix some bugs.
5055 14. Changes to the "personal" filter test:
5057 (1) The test was buggy in that it was just doing the equivalent of
5058 "contains" tests on header lines. For example, if a user's address was
5059 anne@some.where, the "personal" test would incorrectly be true for
5061 To: susanne@some.where
5063 This test is now done by extracting each address from the header in turn,
5064 and checking the entire address. Other tests that are part of "personal"
5065 are now done using regular expressions (for example, to check local parts
5066 of addresses in From: header lines).
5068 (2) The list of non-personal local parts in From: addresses has been
5069 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5070 taken from the Sieve specification recommendations.
5072 (3) If the message contains any header line starting with "List-" it is
5073 treated as non-personal.
5075 (4) The test for "circular" in the Subject: header line has been removed
5076 because it now seems ill-conceived.
5078 15. Minor typos in src/EDITME comments corrected.
5080 16. Installed latest exipick from John Jetmore.
5082 17. If headers_add on a router specified a text string that was too long for
5083 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5084 of string_sprintf() is now avoided.
5086 18. $message_body_size was not set (it was always zero) when running the DATA
5087 ACL and the local_scan() function.
5089 19. For the "mail" command in an Exim filter, no default was being set for
5090 the once_repeat time, causing a random time value to be used if "once" was
5091 specified. (If the value happened to be <= 0, no repeat happened.) The
5092 default is now 0s, meaning "never repeat". The "vacation" command was OK
5093 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5094 (I found it when inspecting the code).
5096 20. There is now an overall timeout for performing a callout verification. It
5097 defaults to 4 times the callout timeout, which applies to individual SMTP
5098 commands during the callout. The overall timeout applies when there is more
5099 than one host that can be tried. The timeout is checked before trying the
5100 next host. This prevents very long delays if there are a large number of
5101 hosts and all are timing out (e.g. when the network connections are timing
5102 out). The value of the overall timeout can be changed by specifying an
5103 additional sub-option for "callout", called "maxwait". For example:
5105 verify = sender/callout=5s,maxwait=20s
5107 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5108 to the end before writing, but this should make it even safer).
5110 22. Exim was forgetting that it had advertised PIPELINING for the second and
5111 subsequent messages on an SMTP connection. It was also not resetting its
5112 memory on STARTTLS and an internal HELO.
5114 23. When Exim logs an SMTP synchronization error within a session, it now
5115 records whether PIPELINING has been advertised or not.
5117 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5118 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5119 rather than long int.
5121 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5123 26. Added the never_mail option to autoreply.
5129 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5130 crash if the getsockname() call failed; this can happen if a connection is
5131 closed very soon after it is established. The problem was simply in the
5132 order in which certain operations were done, causing Exim to try to write
5133 to the SMTP stream before it had set up the file descriptor. The bug has
5134 been fixed by making things happen in the correct order.
5140 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5141 before the connection was closed, thus losing the rejection response.
5143 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5144 some early Solaris releases, but causes trouble in current releases where
5145 socklen_t is defined.
5147 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5150 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5153 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5154 that releasing the top of it at the end releases what was used for sub-
5155 expansions (unless the block got too big). However, discard this block if
5156 the first thing is a variable or header, so that we can use its block when
5157 it is dynamic (useful for very large $message_headers, for example).
5159 6. Lookups now cache *every* query, not just the most recent. A new, separate
5160 store pool is used for this. It can be recovered when all lookup caches are
5161 flushed. Lookups now release memory at the end of their result strings.
5162 This has involved some general refactoring of the lookup sources.
5164 7. Some code has been added to the store_xxx() functions to reduce the amount
5165 of flapping under certain conditions.
5167 8. log_incoming_interface used to affect only the <= reception log lines. Now
5168 it causes the local interface and port to be added to several more SMTP log
5169 lines, for example "SMTP connection from", and rejection lines.
5171 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5173 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5175 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5176 could be overwritten at the end of the current message (or the start of a
5177 new message if it was set in a HELO ACL). The value is now preserved for
5178 the duration of the SMTP connection.
5180 12. If a transport had a headers_rewrite setting, and a matching header line
5181 contained an unqualified address, that address was qualified, even if it
5182 did not match any rewriting rules. The underlying bug was that the values
5183 of the flags that permit the existence of unqualified sender and recipient
5184 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5185 non-local messages, and by -bnq for local messages) were not being
5186 preserved with the message after it was received.
5188 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5189 "next input=" as part of the text comprising the host identity instead of
5190 the correct text. The code was using the same buffer for two different
5191 strings. However, depending on which order the printing function evaluated
5192 its arguments, the bug did not always show up. Under Linux, for example, my
5193 test suite worked just fine.
5195 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5196 doesn't work with some older versions of Perl. It has been changed to "my",
5197 which in any case is probably the better facility to use.
5199 15. A really picky compiler found some instances of statements for creating
5200 error messages that either had too many or two few arguments for the format
5203 16. The size of the buffer for calls to the DNS resolver has been increased
5204 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5205 for addresses that have a lot of PTR records. This alleviates a problem; it
5206 does not fully solve it.
5208 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5209 buffer now truncates the list and logs the incident, which is the same
5210 action as happens when Exim is looking up a host name and its aliases.
5211 Previously in this situation something unpredictable would happen;
5212 sometimes it was "internal error: store_reset failed".
5214 18. If a server dropped the connection unexpectedly when an Exim client was
5215 using GnuTLS and trying to read a response, the client delivery process
5216 crashed while trying to generate an error log message.
5218 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5219 string, for example:
5221 warn message = H1: something\nH2: something
5223 the text was added as a single header line from Exim's point of view
5224 though it ended up OK in the delivered message. However, searching for the
5225 second and subsequent header lines using $h_h2: did not work. This has been
5226 fixed. Similarly, if a system filter added multiple headers in this way,
5227 the routers could not see them.
5229 20. Expanded the error message when iplsearch is called with an invalid key to
5230 suggest using net-iplsearch in a host list.
5232 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5233 ACLs are no longer actually imposed (and a message to that effect is
5236 22. If a "gecos" field in a passwd entry contained escaped characters, in
5237 particular, if it contained a \" sequence, Exim got it wrong when building
5238 a From: or a Sender: header from that name. A second bug also caused
5239 incorrect handling when an unquoted " was present following a character
5240 that needed quoting.
5242 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5243 was not being matched caselessly.
5245 24. Arranged for all hyphens in the exim.8 source to be escaped with
5248 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5249 properly. Recipient callout cache records were still being keyed to include
5250 the sender, even when use_sender was set false. This led to far more
5251 callouts that were necessary. The sender is no longer included in the key
5252 when use_sender is false.
5254 26. Added "control = submission" modifier to ACLs.
5256 27. Added the ${base62d: operator to decode base 62 numbers.
5258 28. dnsdb lookups can now access SRV records.
5260 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5261 the configuration file.
5263 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5264 (-v) mode. This makes the output for a verbose queue run more intelligible.
5266 31. Added a use_postmaster feature to recipient callouts.
5268 32. Added the $body_zerocount variable, containing the number of binary zero
5269 bytes in the message body.
5271 33. The time of last modification of the "new" subdirectory is now used as the
5272 "mailbox time last read" when there is a quota error for a maildir
5275 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5277 35. Added +ignore_unknown as a special item in host lists.
5279 36. Code for decoding IPv6 addresses in host lists is now included, even if
5280 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5281 address was recognized as an IP address, but was then not correctly decoded
5282 into binary, causing unexpected and incorrect effects when compared with
5289 1. Very minor rewording of debugging text in manualroute to say "list of
5290 hosts" instead of "hostlist".
5292 2. If verify=header_syntax was set, and a header line with an unqualified
5293 address (no domain) and a large number of spaces between the end of the
5294 name and the colon was received, the reception process suffered a buffer
5295 overflow, and (when I tested it) crashed. This was caused by some obsolete
5296 code that should have been removed. The fix is to remove it!
5298 3. When running in the test harness, delay a bit after writing a bounce
5299 message to get a bit more predictability in the log output.
5301 4. Added a call to search_tidyup() just before forking a reception process. In
5302 theory, someone could use a lookup in the expansion of smtp_accept_max_
5303 per_host which, without the tidyup, could leave open a database connection.
5305 5. Added the variables $recipient_data and $sender_data which get set from a
5306 lookup success in an ACL "recipients" or "senders" condition, or a router
5307 "senders" option, similar to $domain_data and $local_part_data.
5309 6. Moved the writing of debug_print from before to after the "senders" test
5312 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5313 problems for message scanning, either using a data ACL, or using
5314 local_scan() because the Received: header was not generated till after they
5315 were called (in order to set the time as the time of reception completion).
5316 I have revised the way this works. The header is now generated after the
5317 body is received, but before the ACL or local_scan() are called. After they
5318 are run, the timestamp in the header is updated.
5324 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5325 before starting a queue runner without re-exec. This happened only when
5326 deliver_drop_privilege was set or when the Exim user was set to root. The
5327 effect of the bug was that timeouts during subsequent deliveries caused
5328 crashes instead of being properly handled. The handler is now left at its
5329 default (and expected) setting.
5331 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5332 message, again when deliver_drop_privilege is set or Exim is run as root.
5333 The bug described in (1) was not present in this case, but the tidying up
5334 of the other signals was missing. I have made the two cases consistent.
5336 3. The ignore_target_hosts setting on a manualroute router was being ignored
5337 for hosts that were looked up using the /MX notation.
5339 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5342 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5343 operated on the sender address. After changing the $sender_address to <>
5344 for the sender address verify, Exim was re-instated it as the original
5345 (before rewriting) address, but remembering that it had rewritten it, so it
5346 wasn't rewriting it again. This bug also had the effect of breaking the
5347 sender address verification caching when the sender address was rewritten.
5349 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5350 This has been changed so that if the ip literal address matches
5351 ignore_target_hosts, the router declines.
5353 7. Added expansion conditions match_domain, match_address, and match_local_
5354 part (NOT match_host).
5356 8. The placeholder for the Received: header didn't have a length field set.
5358 9. Added code to Exim itself and to exim_lock to test for a specific race
5359 condition that could lead to file corruption when using MBX delivery. The
5360 issue is with the lockfile that is created in /tmp. If this file is removed
5361 after a process has opened it but before that process has acquired a lock,
5362 there is the potential for a second process to recreate the file and also
5363 acquire a lock. This could lead to two Exim processes writing to the file
5364 at the same time. The added code performs the same test as UW imapd; it
5365 checks after acquiring the lock that its file descriptor still refers to
5366 the same named file.
5368 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5369 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5372 11. Added the smtp_active_hostname option. If used, this will typically be made
5373 to depend on the incoming interface address. Because $interface_address is
5374 not set up until the daemon has forked a reception process, error responses
5375 that can happen earlier (such as "too many connections") no longer contain
5378 12. If an expansion in a condition on a "warn" statement fails because a lookup
5379 defers, the "warn" statement is abandoned, and the next ACL statement is
5380 processed. Previously this caused the whole ACL to be aborted.
5382 13. Added the iplsearch lookup type.
5384 14. Added ident_timeout as a log selector.
5386 15. Added tls_certificate_verified as a log selector.
5388 16. Added a global option tls_require_ciphers (compare the smtp transport
5389 option of the same name). This controls incoming TLS connections.
5391 17. I finally figured out how to make tls_require_ciphers do a similar thing
5392 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5393 before starting the TLS session.
5395 18. Tabs are now shown as \t in -bP output.
5397 19. If the log selector return_path_on_delivery was set, Exim crashed when
5398 bouncing a message because it had too many Received: header lines.
5400 20. If two routers both had headers_remove settings, and the first one included
5401 a superfluous trailing colon, the final name in the first list and the
5402 first name in the second list were incorrectly joined into one item (with a
5403 colon in the middle).
5409 1. Added -C and -D options to the exinext utility, mainly to make it easier
5410 to include in the automated testing, but these could be helpful when
5411 multiple configurations are in use.
5413 2. The exinext utility was not formatting the output nicely when there was
5414 an alternate port involved in the retry record key, nor when there was a
5415 message id as well (for retries that were specific to a specific message
5416 and a specific host). It was also confused by IPv6 addresses, because of
5417 the additional colons they contain. I have fixed the IPv4 problem, and
5418 patched it up to do a reasonable job for IPv6.
5420 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5421 delivery, the log line now contains "pipelined" if PIPELINING was used.
5423 4. An SMTP transport process used to panic and die if the bind() call to set
5424 an explicit outgoing interface failed. This has been changed; it is now
5425 treated in the same way as a connect() failure.
5427 5. A reference to $sender_host_name in the part of a conditional expansion
5428 that was being skipped was still causing a DNS lookup. This no longer
5431 6. The def: expansion condition was not recognizing references to header lines
5432 that used bh_ and bheader_.
5434 7. Added the _cache feature to named lists.
5436 8. The code for checking quota_filecount in the appendfile transport was
5437 allowing one more file than it should have been.
5439 9. For compatibility with Sendmail, the command line option
5447 and sets the incoming protocol and host name (for trusted callers). The
5448 host name and its colon can be omitted when only the protocol is to be set.
5449 Note the Exim already has two private options, -pd and -ps, that refer to
5450 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5451 "s", but I don't think that's a major issue.
5453 10. A number of refactoring changes to the code, none of which should affect
5456 (a) The number of logging options was getting close to filling up the
5457 32-bit word that was used as a bit map. I have split them into two classes:
5458 those that are passed in the argument to log_write(), and those that are
5459 only ever tested independently outside of that function. These are now in
5460 separate 32-bit words, so there is plenty of room for expansion again.
5461 There is no change in the user interface or the logging behaviour.
5463 (b) When building, for example, log lines, the code previously used a
5464 macro that called string_cat() twice, in order to add two strings. This is
5465 not really sufficiently general. Furthermore, there was one instance where
5466 it was actually wrong because one of the argument was used twice, and in
5467 one call a function was used. (As it happened, calling the function twice
5468 did not affect the overall behaviour.) The macro has been replaced by a
5469 function that can join an arbitrary number of extra strings onto a growing
5472 (c) The code for expansion conditions now uses a table and a binary chop
5473 instead of a serial search (which was left over from when there were very
5474 few conditions). Also, it now recognizes conditions like "pam" even when
5475 the relevant support is not compiled in: a suitably worded error message is
5476 given if an attempt is made to use such a condition.
5478 11. Added ${time_interval:xxxxx}.
5480 12. A bug was causing one of the ddress fields not to be passed back correctly
5481 from remote delivery subprocesses. The field in question was not being
5482 subsequently used, so this caused to problems in practice.
5484 13. Added new log selectors queue_time and deliver_time.
5486 14. Might have fixed a bug in maildirsizefile handling that threw up
5487 "unexpected character" debug warnings, and recalculated the data
5488 unnecessarily. In any case, I expanded the warning message to give more
5491 15. Added the message "Restricted characters in address" to the statements in
5492 the default ACL that block characters like @ and % in local parts.
5494 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5495 Three changes have been made:
5497 (a) There was a serious bug; a negative response to MAIL caused the whole
5498 recipient domain to be cached as invalid, thereby blocking all messages
5499 to all local parts at the same domain, from all senders. This bug has
5500 been fixed. The domain is no longer cached after a negative response to
5501 MAIL if the sender used is not empty.
5503 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5506 (c) A new callout option, "use_sender" has been added for people who want
5507 the modified behaviour.
5513 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5516 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5517 indeed breaks things for older releases.
5519 3. Added additional logging to the case where there is a problem reading data
5520 from a filter that is running in a subprocess using a pipe, in order to
5521 try to track down a specific problem.
5523 4. Testing facility fudge: when running in the test harness and attempting
5524 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5525 getting "No route to host". Convert this to a timeout.
5527 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5530 6. Some OS don't have socklen_t but use size_t instead. This affects the
5531 fifth argument of getsockopt() amongst other things. This is now
5532 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5533 can be set for individual OS. I have set it for SunOS5, OSF1, and
5534 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5535 some earlier ones do not.
5537 7. Change 4.30/15 was not doing the test caselessly.
5539 8. The standard form for an IPv6 address literal was being rejected by address
5540 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5541 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5542 this, as well as the form without the "IPv6" on the front (but only when
5543 address literals are enabled, of course).
5545 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5547 10. Exim crashed if a message with an empty sender address specified by -f
5548 encountered a router with an errors_to setting. This could be provoked only
5549 by a command such as
5553 where an empty string was supplied; "<>" did not hit this bug.
5555 11. Installed PCRE release 4.5.
5557 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5558 remained set. It is now erased.
5560 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5561 times from message ids (which are base 36 rather than the normal 62).
5563 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5564 were being counted as actual protocol errors, and logged if the log
5565 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5566 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5567 and DATA following a set of rejected RCPTs do not count as protocol errors.
5568 In other words, Exim assumes they were pipelined, though this may not
5569 actually be the case. Of course, in all cases the client gets an
5570 appropriate error code.
5572 15. If a lookup fails in an ACL condition, a message about the failure may
5573 be available; it is used if testing the ACL cannot continue, because most
5574 such messages specify what the cause of the deferral is. However, some
5575 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5576 that caused an old message to be retained and used if a later statement
5577 caused a defer, replacing the real cause of the deferral.
5579 16. If an IP address had so many PTR records that the DNS lookup buffer
5580 was not large enough to hold them, Exim could crash while trying to process
5581 the truncated data. It now detects and logs this case.
5583 17. Further to 4.21/58, another change has been made: if (and only if) the
5584 first line of a message (the first header line) ends with CRLF, a bare LF
5585 in a subsequent header line has a space inserted after it, so as not to
5586 terminate the header.
5588 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5589 unnecessarily, used atoi() instead of strtol(), and didn't check the
5590 termination when getting file sizes from file names by regex.
5592 19. Completely re-implemented the support for maildirsize files, in the light
5593 of a number of problems with the previous contributed implementation
5594 (4.30/29). In particular:
5596 . If the quota is zero, the maildirsize file is maintained, but no quota is
5599 . If the maildir directory does not exist, it is created before any attempt
5600 to write a maildirsize file.
5602 . The quota value in the file is just a cache; if the quota is changed in
5603 the transport, the new value overrides.
5605 . A regular expression is available for excluding directories from the
5608 20. The autoreply transport checks the characters in options that define the
5609 message's headers; it allows continued headers, but it was checking with
5610 isspace() after an embedded newline instead of explicitly looking for a
5613 21. If all the "regular" hosts to which an address was routed had passed their
5614 expiry times, and had not reached their retry times, the address was
5615 bounced, even if fallback hosts were defined. Now Exim should go on to try
5618 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5619 equivalent code in the SMTP transport. Some hosts send humungous responses
5620 to HELO/EHLO, more than 1024 it seems.
5622 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5623 gives compiler warnings in some environments. I've now done it "properly",
5626 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5627 (because of problems with the built-in one) was declared to return uschar *
5628 instead of char *, causing compiler failure.
5630 25. Fixed a file descriptor leak when processing alias/forward files.
5632 26. Fixed a minor format string issue in dbfn.c.
5634 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5636 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5637 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5640 29. When a sender address is verified, it is cached, to save repeating the test
5641 when there is more than one recipient in a message. However, when the
5642 verification involves a callout, it is possible for different callout
5643 options to be set for different recipients. It is too complicated to keep
5644 track of this in the cache, so now Exim always runs a verification when a
5645 callout is required, relying on the callout cache for the optimization.
5646 The overhead is duplication of the address routing, but this should not be
5649 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5650 to be verified with callout=postmaster, and the main callout worked but the
5651 postmaster check failed, the verification correctly failed. However, if a
5652 subsequent RCPT command asked for sender verification *without* the
5653 postmaster check, incorrect caching caused this verification also to fail,
5656 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5657 it was not caching the DNS options (qualify_single, search_parents) that
5658 were used when the lookup failed. A subsequent lookup with different
5659 options therefore always gave the same answer, though there were cases
5660 where it should not have. (Example: a "domains = !$mx_any" option on a
5661 dnslookup router: the "domains" option is always processed without any
5662 widening, but the router might have qualify_single set.) Now Exim uses the
5663 cached value only when the same options are set.
5665 32. Added John Jetmore's "exipick" utility to the distribution.
5667 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5668 than a timeout (e.g. a certificate is required, and is not provided), an
5669 Exim server now closes the connection immediately. Previously it waited for
5670 the client to close - but if the client is SSL, it seems that they each
5671 wait for each other, leading to a delay before one of them times out.
5673 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5674 maintained 0.8.x compatibility because I don't think many are using it, and
5675 it is clearly obsolete.
5677 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5680 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5681 client certificate was expired. A simple patch fixes this, though I don't
5682 understand the full logic of why the verify callback is called multiple
5685 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5686 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5687 which causes problems with some clients (such as the Certicom SSL Plus
5688 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5689 disables the coutermeasure allowing Eudora to connect."
5691 38. Exim was not checking that a write() to a log file succeeded. This could
5692 lead to Bad Things if a log got too big, in particular if it hit a file
5693 size limit. Exim now panics and dies if it cannot write to a log file, just
5694 as it does if it cannot open a log file.
5696 39. Modified OS/Makefile-Linux so that it now contains
5698 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5700 The two -D definitions ensure that Exim is compiled with large file
5701 support, which makes it possible to handle log files that are bigger than
5704 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5705 instance) a domain was checked against a named list that involved a lookup,
5706 causing $domain_data to be set, then another domain was checked against the
5707 same list, then the first domain was re-checked, the value of $domain_data
5708 after the final check could be wrong. In particular, if the second check
5709 failed, it could be set empty. This bug probably also applied to
5712 41. The strip_trailing_dot option was not being applied to the address given
5713 with the -f command-line option.
5715 42. The code for reading a message's header from the spool was incrementing
5716 $received_count, but never initializing it. This meant that the value was
5717 incorrect (doubled) while delivering a message in the same process in which
5718 it was received. In the most common configuration of Exim, this never
5719 happens - a fresh exec is done - but it can happen when
5720 deliver_drop_privilege is set.
5722 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5723 it now includes up to 150 characters of the unexpected data in the log
5726 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5727 and building data strings. The size of both of these buffers was 10 000
5728 bytes - far larger than anybody would *ever* want, thought I. Needless to
5729 say, somebody hit the limit. I have increased the maximum line length to
5730 20 000 and the maximum data length of concatenated lines to 100 000. I have
5731 also fixed two bugs, because there was no checking on these buffers. Tsk,
5732 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5733 buffer is too small.
5735 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5736 lsearch lookups. Now it does.
5738 46. When parsing a route_list item in a manualroute router, a fixed-length
5739 buffer was used for the list of hosts. I made this 1024 bytes long,
5740 thinking that nobody would ever have a list of hosts that long. Wrong.
5741 Somebody had a whole pile of complicated expansion conditions, and the
5742 string was silently truncated, leading to an expansion error. It turns out
5743 that it is easier to change to an unlimited length (owing to other changes
5744 that have happened since this code was originally written) than to build
5745 structure for giving a limitation error. The length of the item that
5746 expands into the list of hosts is now unlimited.
5748 47. The lsearch lookup could not handle data where the length of text line was
5749 more than 4095 characters. Such lines were truncated, leading to shortened
5750 data being returned. It should now handle lines of any length.
5752 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5753 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5756 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5757 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5758 no longer happens in comment lines. A list of replacements is now placed
5759 at the head of all of the source files, except those whose only change is
5760 to replace PERL_COMMAND in the very first #! line.
5762 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5763 messages on the queue, with a bottom-up merge sort, using code contributed
5764 by Michael Haardt. This should make operations like -bp somewhat faster on
5765 large queues. It won't affect queue runners, except when queue_run_in_order
5768 51. Installed eximstats 1.31 in the distribution.
5770 52. Added support for SRV lookups to the dnslookup router.
5772 53. If an ACL referred to $message_body or $message_body_end, the value was not
5773 reset for any messages that followed in the same SMTP session.
5775 54. The store-handling optimization for building very long strings was not
5776 differentiating between the different store pools. I don't think this
5777 actually made any difference in practice, but I've tidied it.
5779 55. While running the routers to verify a sender address, $sender_address
5780 was still set to the sender address. This is wrong, because when routing to
5781 send a bounce to the sender, it would be empty. Therefore, I have changed
5782 it so that, while verifying a sender address, $sender_address is set to <>.
5783 (There is no change to what happens when verifying a recipient address.)
5785 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5786 target A or AAAA records (if not already returned) without resetting the
5787 qualify_single or search_parents options of the DNS resolver. These are
5788 inappropriate in this case because the targets of MX and SRV records must
5789 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5790 target that, when qualified, matched something in the local domain. These
5791 two options are now turned off when doing these lookups.
5793 57. It seems that at least some releases of Reiserfs (which does not have the
5794 concept of a fixed number of inodes) returns zero and not -1 for the
5795 number of available inodes. This interacted badly with check_spool_inodes,
5796 which assumed that -1 was the "no such thing" setting. What I have done is
5797 to check that the total number of inodes is greater than zero before doing
5798 the test of how many are available.
5800 58. When a "warn" ACL statement has a log_message modifier, the message is
5801 remembered, and not repeated. This is to avoid a lot of repetition when a
5802 message has many recipients that cause the same warning to be written.
5803 However, Exim was preserving the list of already written lines for an
5804 entire SMTP session, which doesn't seem right. The memory is now reset if a
5805 new message is started.
5807 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5808 debugging output unless log_rewrite was also set.
5810 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5811 of (int)(handle) when we know that handle contains (void *)(-1).
5813 61. The Exim daemon panic-logs an error return when it closes the incoming
5814 connection. However "connection reset by peer" seems to be common, and
5815 isn't really an error worthy of noting specially, so that particular error
5818 62. When Exim is trying to find all the local interfaces, it used to panic and
5819 die if the ioctl to get the interface flags failed. However, it seems that
5820 on at least one OS (Solaris 9) it is possible to have an interface that is
5821 included in the list of interfaces, but for which you get a failure error
5822 for this call. This happens when the interface is not "plumbed" into a
5823 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5824 failure of the "get flags" call assumes that the interface is down.
5826 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5827 makes life easier for people who are doing arithmetic on fields extracted
5828 from dates, where you often get leading zeros that should not be
5829 interpreted as octal.
5831 64. Added qualify_domain to the redirect router, to override the global
5834 65. If a pathologically long header line contained very many addresses (the
5835 report of this problem mentioned 10 000) and each of them was rewritten,
5836 Exim could use up a very large amount of memory. (It kept on making new
5837 copies of the header line as it rewrote, and never released the old ones.)
5838 At the expense of a bit more processing, the header rewriting function has
5839 been changed so that it no longer eats memory in this way.
5841 66. The generation of the Received: header has been moved from the time that a
5842 message starts to be received, to the time that it finishes. The timestamp
5843 in the Received: header should now be very close to that of the <= log
5844 line. There are two side-effects of this change:
5846 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5847 logged header lines no longer include the local Received: line, because
5848 it has not yet been created. The same applies to a copy of the message
5849 that is returned to a non-SMTP sender when a message is rejected.
5851 (b) When a filter file is tested using -bf, no additional Received: header
5852 is added to the test message. After some thought, I decided that this
5855 This change does not affect the value of $received_for. It is still set
5856 after address rewriting, but before local_scan() is called.
5858 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5860 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5861 gave an unhelpful panic error message, and a defer error. I have managed to
5862 change this behaviour so that it now rejects any supplied certificate,
5863 which seems right, as the list of acceptable certificates is empty.
5865 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5866 gave an unhelpful defer error. I have not managed to make this reject any
5867 supplied certificates, but the error message it gives is "no certificate
5868 supplied", which is not helpful.
5870 70. exigrep's output now also includes lines that are not associated with any
5871 message, but which match the given pattern. Implemented by a patch from
5872 Martin Sluka, which also tidied up the Perl a bit.
5874 71. Recipient callout verification, like sender verification, was using <> in
5875 the MAIL FROM command. This isn't really the right thing, since the actual
5876 sender may affect whether the remote host accepts the recipient or not. I
5877 have changed it to use the actual sender in the callout; this means that
5878 the cache record is now keyed on a recipient/sender pair, not just the
5879 recipient address. There doesn't seem to be a real danger of callout loops,
5880 since a callout by the remote host to check the sender would use <>.
5881 [SEE ABOVE: changed after hitting problems.]
5883 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5884 temporary errors. However, in the case of such a code being given after
5885 the end of a data transmission (i.e. after ".") Exim was failing to write
5886 a retry record for the message. (Yes, there was some broken host that was
5887 actually sending 8xx at this point.)
5889 73. An unknown lookup type in a host list could cause Exim to panic-die when
5890 the list was checked. (An example that provoked this was putting <; in the
5891 middle of a list instead of at the start.) If this happened during a DATA
5892 ACL check, a -D file could be left lying around. This kind of configuration
5893 error no longer causes Exim to die; instead it causes a defer error. The
5894 incident is still logged to the main and panic logs.
5896 74. Buglet left over from Exim 3 conversion. The message "too many messages
5897 in one connection" was written to the rejectlog but not the mainlog, except
5898 when address rewriting (yes!) was being logged.
5900 75. Added write_rejectlog option.
5902 76. When a system filter was run not as root (that is, when system_filter_user
5903 was set), the values of the $n variables were not being returned to the
5904 main process; thus, they were not subsequently available in the $sn
5907 77. Added +return_path_on_delivery log selector.
5909 78. A connection timeout was being treated differently from recipients deferred
5910 when testing hosts_max_try with a message that was older than the host's
5911 retry timeout. (The host should not be counted, thus allowing all hosts to
5912 be tried at least once before bouncing.) This may have been the cause of an
5913 occasionally reported bug whereby a message would remain on the queue
5914 longer than the retry timeout, but would be bounced if a delivery was
5915 forced. I say "may" because I never totally pinned down the problem;
5916 setting up timeout/retry tests is difficult. See also the next item.
5918 79. The ultimate address timeout was not being applied to errors that involved
5919 a combination of host plus message (for example, a timeout on a MAIL
5920 command). When an address resolved to a number of possible hosts, and they
5921 were not all tried for each delivery (e.g. because of hosts_max_try), a
5922 message could remain on the queue longer than the retry timeout.
5924 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5927 81. Fixed an obscure SMTP outgoing bug which required at least the following
5928 conditions: (a) there was another message waiting for the same server;
5929 (b) the server returned 5xx to all RCPT commands in the first message so
5930 that the message was not completed; (c) the server dropped the connection
5931 or gave a negative response to the RSET that Exim sends to abort the
5932 transaction. The observed case was a dropped connection after DATA that had
5933 been sent in pipelining mode. That is, the server had advertised PIPELINING
5934 but was not implementing it correctly. The effect of the bug was incorrect
5935 behaviour, such as trying another host, and this could lead to a crash.
5941 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5942 and daemon.c were passed as pointers to ints; they should have been
5943 pointers to socklen_t variables (which are typically unsigned ints).
5945 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5948 3. Fixed a really odd bug that affected only the testing scheme; patching a
5949 certain fixed string in the binary changed the value of another string that
5950 happened to be identical to the end of the original first string.
5952 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5953 name", it returns that address as the IP address. On some operating
5954 systems (e.g. Solaris), it also passes back the IP address string as the
5955 "host name". However, on others (e.g. Linux), it passes back an empty
5956 string. Exim wasn't checking for this, and was changing the host name to an
5957 empty string, assuming it had been canonicalized.
5959 5. Although rare, it is permitted to have more than one PTR record for a given
5960 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5961 all the names associated with an address, because they do in Solaris.
5962 However, it seems that they do not in Linux for data that comes from the
5963 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5964 I found this out when I moved to a new Linux workstation and tried to run
5965 the Exim test suite.
5967 To get round this problem I have changed the code so that it now does its
5968 own call to the DNS to look up PTR records when searching for a host name.
5969 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5970 addresses that are only in /etc/hosts are still found.
5972 This behaviour is, however, controlled by an option called host_lookup_
5973 order, which defaults to "bydns:byaddr". If people want to use the other
5974 order, or indeed, just use one or the other means of lookup, they can
5975 specify it in this variable.
5977 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5978 some operating systems, this comes back from gethostbyaddr() as an empty
5979 string, and this is what Exim used to test for. However, it seems that in
5980 other systems, "." is yielded. Exim now tests for this case too.
5982 7. The values of check_spool_space and check_log_space are now held internally
5983 as a number of kilobytes instead of an absolute number of bytes. If a
5984 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5985 kilobyte. This means that much larger values can be stored.
5987 8. Exim monitor: an attempt to get the action menu when not actually pointing
5988 at a message produces an empty menu entitled "No message selected". This
5989 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5990 no entries in it ("Shell widget menu has zero width and/or height"). So I
5991 have added a single, blank menu entry in this case.
5993 9. Added ${quote_local_part.
5995 10. MIME decoding is now applied to the contents of Subject: header lines when
5998 11. Now that a reference to $sender_host_address automatically causes a reverse
5999 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6000 host lookup before query-style lookups in lists that might use this
6001 variable. This has therefore been abolished, and the "net-" prefix is no
6002 longer necessary for query-style lookups.
6004 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6005 have been a typo for LFLAGS, so it has been changed.
6007 13. The install script calls Exim with "-C /dev/null" in order to find the
6008 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6009 to be output. However, since Exim outputs its version number before the
6010 error, it didn't break the script. It just looked ugly. I fixed this by
6011 always allowing "-C /dev/null" if the caller is root.
6013 14. Ignore overlarge ACL variable number when reading spool file - insurance
6014 against a later release with more variables having written the file.
6016 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6017 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6018 this, as well as the form without the "IPv6" on the front.
6020 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6021 OS/Makefile-Darwin file.
6023 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6024 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6026 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6027 variable. However, some OS use ints for inodes. I've added cast to long int
6028 to get rid of the compiler warning.
6030 19. I had forgotten to lock out "/../" in configuration file names when
6031 ALT_CONFIG_PREFIX was set.
6033 20. Routers used for verification do not need to specify transports. However,
6034 if such a router generated a host list, and callout was configured, Exim
6035 crashed, because it could not find a port number from the (non-existent)
6036 transport. It now assumes port 25 in this circumstance.
6038 21. Added the -t option to exigrep.
6040 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6041 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6042 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6043 them is set, LOOKUP_LSEARCH is forced.
6045 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6047 24. Added sender and host information to the "rejected by local_scan()" log
6048 line; previously there was no indication of these.
6050 25. Added .include_if_exists.
6052 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6053 Linux. It turns out that not all file systems support this. Apparently some
6054 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6055 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6056 not supported on the file descriptor, is now ignored when Exim is trying to
6057 sync a directory. This applies only to Linux.
6059 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6061 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6062 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6065 29. Added support for maildirsize files from supplied patch (modified a bit).
6067 30. The use of :fail: followed by an empty string could lead Exim to respond to
6068 sender verification failures with (e.g.):
6070 550 Verification failed for <xxx>
6071 550 Sender verify failed
6073 where the first response line was missing the '-' that indicates it is not
6074 the final line of the response.
6076 31. The loop for finding the name of the user that called Exim had a hardwired
6077 limit of 10; it now uses the value of finduser_retries, which is used for
6078 all other user lookups.
6080 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6083 33. Exim was neglecting to zero errno before one call of strtol() when
6084 expanding a string and expecting an integer value. On some systems this
6085 resulted in spurious "integer overflow" errors. Also, it was casting the
6086 result into an int without checking.
6088 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6089 did not work. The code looks as if it has *never* worked, though it appears
6090 to have been documented since at least release 1.62. I have made it work.
6092 35. The "timeout_DNS" error in retry rules, also documented since at least
6093 1.62, also never worked. As it isn't clear exactly what this means, and
6094 clearly it isn't a major issue, I have abolished the feature by treating it
6095 as "timeout", and writing a warning to the main and panic logs.
6097 36. The display of retry rules for -brt wasn't always showing the error code
6100 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6101 timeout_connect_A, timeout_connect_MX.
6103 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6104 to the empty sender.
6106 39. The daemon was not analysing the content of -oX till after it had closed
6107 stderr and disconnected from the controlling terminal. This meant that any
6108 syntax errors were only noted on the panic log, and the return code from
6109 the command was 0. By re-arranging the code a little, I've made the
6110 decoding happen first, so such errors now appear on stderr, and the return
6111 code is 1. However, the actual setting up of the sockets still happens in
6112 the disconnected process, so errors there are still only recorded on the
6115 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6116 connections (as happens on some IP stacks) was logged at start up time as
6117 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6118 it from "IPv6 and IPv4", which means that two separate sockets are being
6121 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6122 says whether AF_INET or AF_INET6 was passed as an argument.
6124 42. Exiwhat output was messed up when time zones were included in log
6127 43. Exiwhat now gives more information about the daemon's listening ports,
6128 and whether -tls-on-connect was used.
6130 44. The "port" option of the smtp transport is now expanded.
6132 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6133 silently ignored. Now an error message is written to the main and panic
6136 46. There's a new ACL modifier called "logwrite" which writes to a log file
6137 as soon as it is encountered.
6139 47. Added $local_user_uid and $local_user_gid at routing time.
6141 48. Exim crashed when trying to verify a sender address that was being
6144 49. Exim was recognizing only a space character after ".include". It now also
6145 recognizes a tab character.
6147 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6148 extracting the relevant information from the specification. The man page no
6149 longer contains scrambled data for the -d option, and I've added a section
6150 at the front about calling Exim under different names.
6152 51. Added "extra_headers" argument to the "mail" command in filter files.
6154 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6157 53. Installed eximstats 1.29.
6159 54. Added transport_filter_timeout as a generic transport option.
6161 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6162 Cc: header lines. This was required by RFC 822, but it not required by RFC
6165 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6166 incoming messages that did not have them. Now it does so only if the
6167 message originates locally, that is, if there is no associated remote host
6168 address. When Resent- header lines are present, this applies to the Resent-
6169 lines rather than the non-Resent- lines.
6171 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6172 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6174 58. Messages for configuration errors now include the name of the main
6175 configuration file - useful now that there may be more than one file in a
6176 list (.included file names were always shown).
6178 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6179 for those rare installations that do not start the daemon as root or run it
6180 setuid root. I've cut out the call to initgroups() if the daemon is not
6183 60. The Exim user and group can now be bound into the binary as text strings
6184 that are looked up at the start of Exim's processing.
6186 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6188 62. Added $mailstore_basename variable.
6190 63. Installed patch to sieve.c from Michael Haardt.
6192 64. When Exim failed to open the panic log after failing to open the main log,
6193 the original message it was trying to log was written to stderr and debug
6194 output, but if they were not available (the usual case in production), it
6195 was lost. Now it is written to syslog before the two lines that record the
6196 failures to open the logs.
6198 65. Users' Exim filters run in subprocesses under the user's uid. It is
6199 possible for a "deliver" command or an alias in a "personal" command to
6200 provoke an address rewrite. If logging of address rewriting is configured,
6201 this fails because the process is not running as root or exim. There may be
6202 a better way of dealing with this, but for the moment (because 4.30 needs
6203 to be released), I have disabled address rewrite logging when running a
6204 filter in a non-root, non-exim process.
6210 1. The buildconfig auxiliary program wasn't quoting the value set for
6211 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6212 not defined. This bug was masked in 4.22 by the effect that was fixed in
6215 2. Some messages that were rejected after a message id was allocated were
6216 shown as "incomplete" by exigrep. It no longer does this for messages that
6217 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6219 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6220 have allow_domain_literals set, the ID did not get logged in the <= line.
6221 Domain literals are now always recognized in Message-ID: header lines.
6223 4. The first argument for a ${extract expansion item is the key name or field
6224 number. Leading and trailing spaces in this item were not being ignored,
6225 causing some misleading effects.
6227 5. When deliver_drop_privilege was set, single queue runner processes started
6228 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6229 same command in the process it spins off) were not dropping privilege.
6231 6. When the daemon running as "exim" started a queue runner, it always
6232 re-executed Exim in the spun-off process. This is a waste of effort when
6233 deliver_drop_privilege is set. The new process now just calls the
6234 queue-runner function directly.
6240 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6243 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6244 ignored. Though the use of -oP was forcing the writing of a pid file, it
6245 was always written to the default place.
6247 3. If the message "no IP address found for host xxxx" is generated during
6248 incoming verification, it is now followed by identification of the incoming
6249 connection (so you can more easily find what provoked it).
6251 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6253 5. Added some features to "harden" Exim a bit more against certain attacks:
6255 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6256 be put in Local/Makefile. This is like the never_users runtime option,
6257 but it cannot be overridden. The default setting is "root".
6259 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6260 prefix string with which any file named in a -C command line option
6263 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6264 is retained for -C and -D only if the caller of Exim is root. Without
6265 it, the exim user may also use -C and -D and retain privilege.
6267 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6268 command line option is disabled.
6270 6. Macro names set by the -D option must start with an upper case letter, just
6271 like macro names defined in the configuration file.
6273 7. Added "dereference=" facility to LDAP.
6275 8. Two instances of the typo "uknown" in the source files are fixed.
6277 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6278 the Configure-Makefile script screwed up while processing it.
6280 10. Incorporated PCRE 4.4.
6282 11. The SMTP synchronization check was not operating right at the start of an
6283 SMTP session. For example, it could not catch a HELO sent before the client
6284 waited for the greeting. There is now a check for outstanding input at the
6285 point when the greeting is written. Because of the duplex, asynchronous
6286 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6287 way, but not yet received, when the check is performed.
6289 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6290 on TCP/IP sockets, because this apparently causes some broken clients to
6293 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6294 unchanged) from the Cygwin maintainer.
6296 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6297 was supported instead of testing for mbx. Effectively a typo.
6299 15. The spa authenticator server code was not checking that the input it
6300 received was valid base64.
6302 16. The debug output line for the "set" modifier in ACLs was not showing the
6303 name of the variable that was being set.
6305 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6307 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6308 lookup on its own. Something else was needed to trigger the lookup. For
6309 example, a match in host_lookup or the need for a host name in a host list.
6310 Now, if $sender_host_name is referenced and the host name has not yet been
6311 looked up, a lookup is performed. If the lookup fails, the variable remains
6312 empty, and $host_lookup_failed is set to "1".
6314 19. Added "eqi" as a case-independent comparison operator.
6316 20. The saslauthd authentication condition could segfault if neither service
6317 nor realm was specified.
6319 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6320 error message that was logged was misleading, and incoming SMTP
6321 connections were dropped. The message is now more accurate, and temporary
6322 errors are given to SMTP connections.
6324 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6325 (except RSET) with a 421 error, until QUIT is received. However, it was
6326 failing to send a response to QUIT.
6328 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6329 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6330 ACL is now run after the helo_try_verify_hosts code.
6332 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6333 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6334 case-independent, so other case variants are also recognized). Apparently
6335 some systems use these upper case variants.
6337 25. If more than two messages were waiting for the same host, and a transport
6338 filter was specified for the transport, Exim sent two messages over the
6339 same TCP/IP connection, and then failed with "socket operation on non-
6340 socket" when it tried to send the third.
6342 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6344 27. The extern definition of crypt16() in expand.c was not being excluded when
6345 the OS had its own crypt16() function.
6347 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6348 as a preferred synonym for return_size_limit, both as an option and as an
6351 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6353 30. Changed the default configuration ACL to relax the local part checking rule
6354 for addresses that are not in any local domains. For these addresses,
6355 slashes and pipe symbols are allowed within local parts, but the sequence
6356 /../ is explicitly forbidden.
6358 31. SPA server authentication was not clearing the challenge buffer before
6361 32. log_message in a "warn" ACL statement was writing to the reject log as
6362 well as to the main log, which contradicts the documentation and doesn't
6363 seem right (because no rejection is happening). So I have stopped it.
6365 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6366 However, I am unable to do any testing of this.
6368 34. Fixed an infelicity in the appendfile transport. When checking directories
6369 for a mailbox, to see if any needed to be created, it was accidentally
6370 using path names with one or more superfluous leading slashes; tracing
6371 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6373 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6374 contents are added to the log line that is written for every discarded
6375 recipient. (Previously a log_message setting was ignored.)
6377 36. The ${quote: operator now quotes the string if it is empty.
6379 37. The install script runs exim in order to find its version number. If for
6380 some reason other than non-existence or emptiness, which it checks, it
6381 could not run './exim', it was installing it with an empty version number,
6382 i.e. as "exim-". This error state is now caught, and the installation is
6385 38. An argument was missing from the function that creates an error message
6386 when Exim fails to connect to the socket for saslauthd authentication.
6387 This could cause Exim to crash, or give a corrupted message.
6389 39. Added isip, isip4, and isip6 to ${if conditions.
6391 40. The ACL variables $acl_xx are now saved with the message, and can be
6392 accessed later in routers, transports, and filters.
6394 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6395 strings in the file are not string-expanded.
6397 42. If a MAIL command specified a SIZE value that was too large to fit into an
6398 int variable, the check against message_size_limit failed. Such values are
6399 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6400 day this will have to be increased, but I don't think I want to be around
6401 when emails are that large.
6408 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6409 iconv() is not standard in FreeBSD.
6411 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6412 IPv6 enabled. The observed symptom was a segmentation fault on return from
6413 the function os_common_find_running_interfaces() in src/os.c.
6415 3. In the check_special_case() function in daemon.c I had used "errno" as an
6416 argument name, which causes warnings on some systems. This was basically a
6417 typo, since it was named "eno" in the comments!
6419 4. The code that waits for the clock to tick (at a resolution of some fraction
6420 of a second) so as to ensure message-id uniqueness was always waiting for
6421 at least one whole tick, when it could have waited for less. [This is
6422 almost certainly not relevant at current processor speeds, where it is
6423 unlikely to ever wait at all. But we try to future-proof.]
6425 5. The function that sleeps for a time interval that includes fractions of a
6426 second contained a race. It did not block SIGALRM between setting the
6427 timer, and suspending (a couple of lines later). If the interval was short
6428 and the sigsuspend() was delayed until after it had expired, the suspension
6429 never ended. On busy systems this could lead to processes getting stuck for
6432 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6433 process, before it forks any delivery processes. The open lookup caching
6434 mechanism meant that the open file or database connection was passed into
6435 the delivery process. The problem was that delivery processes always tidy
6436 up cached lookup data. This could cause a problem for the next delivery
6437 process started by the queue runner, because the external queue runner
6438 process does not know about the closure. So the next delivery process
6439 still has data in the lookup cache. In the case of a file lookup, there was
6440 no problem because closing a file descriptor in a subprocess doesn't affect
6441 the parent. However, if the lookup was caching a connection to a database,
6442 the connection was closed, and the second delivery process was likely to
6443 see errors such as "PGSQL: query failed: server closed the connection
6444 unexpectedly". The problem has been fixed by closing all cached lookups
6445 in a queue runner before running a delivery process.
6447 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6448 seem to have the "const" qualifier which it has on other OS. I've
6451 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6452 *of the same type* (client or server) with the same public name that an
6453 error should be diagnosed.
6455 9. When Exim looked up a host name for an IP address, but failed to find the
6456 original IP address when looking up the host name (a safety check), it
6457 output the message "<ip address> does not match any IP for NULL", which was
6458 confusing, to say the least. The bug was that the host name should have
6459 appeared instead of "NULL".
6461 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6462 user that is built into the binary, and the -C or -D options is used, root
6463 privilege is dropped before the configuration file is read. In addition,
6464 logging is switched to stderr instead of the normal log files. If the
6465 configuration then re-defines the Exim user, the unprivileged environment
6466 is probably not what is expected, so Exim logs a panic warning message (but
6469 However, if deliver_drop_privilege is set, the unprivileged state may well
6470 be exactly what is intended, so the warning has been cut out in that case,
6471 and Exim is allowed to try to write to its normal log files.
6477 1. smtp_return_error_details was not giving details for temporary sender
6478 or receiver verification errors.
6480 2. Diagnose a configuration error if two authenticators have the same public
6483 3. Exim used not to create the message log file for a message until the first
6484 delivery attempt. This could be confusing when incoming messages were held
6485 for policy or load reasons. The message log file is now created at the time
6486 the message is received, and an initial "Received" line is written to it.
6488 4. The automatically generated man page for command line options had a minor
6489 bug that caused no ill effects; however, a more serious problem was that
6490 the procedure for building the man page automatically didn't always
6491 operate. Consequently, release 4.20 contains an out-of-date version. This
6492 shouldn't happen again.
6494 5. When building Exim with embedded Perl support, the script that builds the
6495 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6496 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6498 6. The freeze_tell option was not being used for messages that were frozen on
6499 arrival, either by an ACL or by local_scan().
6501 7. Added the smtp_incomplete_transaction log selector.
6503 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6504 was accepting AUTH without a new EHLO.
6506 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6507 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6510 10. Logging of TCP/IP connections (when configured) now happens in the main
6511 daemon process instead of the child process, so that the TCP/IP connection
6512 count is more accurate (but it can never be perfect).
6514 11. The use of "drop" in a nested ACL was not being handled correctly in the
6515 outer ACL. Now, if condition failure induced by the nested "drop" causes
6516 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6517 or "require"), the connection is dropped.
6519 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6520 that yield "discard" can now be used with an "accept" or a "discard" verb,
6521 but an error is generated for any others (because I can't see a useful way
6522 to define what should happen).
6524 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6525 are now processed in the same way as lines in the Exim configuration file.
6526 In particular, continuation lines are supported.
6528 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6530 15. Added -ti meaning -t -i.
6532 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6533 domains, and warn by logging if others are found.
6535 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6536 returned from GIFCONF when Exim is trying to find the list of interfaces on
6537 a host. The code in os.c has been modified to copy each ifreq to an aligned
6538 structure in all cases.
6540 Also, in some cases, the returned ifreq's were being copied to a 'struct
6541 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6542 means the last couple of bytes of an IPv6 address could be chopped if the
6543 ifreq contained only a normal sockaddr (14 bytes storage).
6545 18. Named domain lists were not supported in the hosts_treat_as_local option.
6546 An entry such as +xxxx was not recognized, and was treated as a literal
6549 19. Ensure that header lines added by a DATA ACL are included in the reject log
6550 if the ACL subsequently rejects the message.
6552 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6553 MD5 (which is deprecated).
6555 21. When testing a filter file using -bf, Exim was writing a message when it
6556 took the sender from a "From " line in the message, but it was not doing so
6557 when it took $return_path from a Return-Path: header line. It now does.
6559 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6560 with a valid header line field name (a series of printing characters
6561 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6563 23. Changed "disc" in the source to "disk" to conform to the documentation and
6564 the book and for uniformity.
6566 24. Ignore Sendmail's -Ooption=value command line item.
6568 25. When execve() failed while trying to run a command in a pipe transport,
6569 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6570 could be confused with a return value of 69 from the command itself. This
6571 has been changed to 127, the value the shell returns if it is asked to run
6572 a non-existent command. The wording for the related log line suggests a
6573 non-existent command as the problem.
6575 26. If received_header_text expands to an empty string, do not add a Received:
6576 header line to the message. (Well, it adds a token one on the spool, but
6577 marks it "old" so that it doesn't get used or transmitted.)
6579 27. Installed eximstats 1.28 (addition of -nt option).
6581 28. There was no check for failure on the call to getsockname() in the daemon
6582 code. This can fail if there is a shortage of resources on the system, with
6583 ENOMEM, for example. A temporary error is now given on failure.
6585 29. Contrary to the C standard, it seems that in some environments, the
6586 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6587 program. Exim now does this explicitly; it affects the formatting of
6588 timestamps using strftime().
6590 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6591 complaints. I've now initialized all the variables, to avoid this.
6593 32. Header lines added by a system filter were not being "seen" during
6594 transport-time rewrites.
6596 33. The info_callback() function passed to OpenSSL is set up with type void
6597 (*)(SSL *, int, int), as described somewhere. However, when calling the
6598 function (actually a macro) that sets it up, the type void(*)() is
6599 expected. I've put in a cast to prevent warnings from picky compilers.
6601 34. If a DNS black list lookup found a CNAME record, but there were no A
6602 records associated with the domain it pointed at, Exim crashed.
6604 35. If a DNS black list lookup returned more than one A record, Exim ignored
6605 all but the first. It now scans all returned addresses if a particular IP
6606 value is being sought. In this situation, the contents of the
6607 $dnslist_value variable are a list of all the addresses, separated by a
6610 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6611 to accept a host name and all its aliases if the forward lookup for any of
6612 them yielded the IP address of the incoming connection. Now it accepts only
6613 those names whose forward lookup yields the correct IP address. Any other
6614 names are discarded. This closes a loophole whereby a rogue DNS
6615 administrator could create reverse DNS records to break through a
6616 wildcarded host restriction in an ACL.
6618 37. If a user filter or a system filter that ran in a subprocess used any of
6619 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6620 the wrong values were passed to the pipe command ($thisaddress had the
6621 value of $0, $0 had the value of $1, etc). This bug was introduced by
6622 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6624 38. Improved the line breaking for long SMTP error messages from ACLs.
6625 Previously, if there was no break point between 40 and 75 characters, Exim
6626 left the rest of the message alone. Two changes have been made: (a) I've
6627 reduced the minimum length to 35 characters; (b) if it can't find a break
6628 point between 35 and 75 characters, it looks ahead and uses the first one
6629 that it finds. This may give the occasional overlong line, but at least the
6630 remaining text gets split now.
6632 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6633 file descriptors might be low, and that setting 1000 would always raise it.
6634 It turns out that in some environments, the limit is already over 1000 and
6635 that lowering it causes trouble. So now Exim takes care not to decrease it.
6637 40. When delivering a message, the value of $return_path is set to $sender_
6638 address at the start of routing (routers may change the value). By an
6639 oversight, this default was not being set up when an address was tested by
6640 -bt or -bv, which affected the outcome if any router or filter referred to
6643 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6644 log only when "message" or "log_message" are set. However, if one of the
6645 conditions was an address verification, or a call to a nested ACL, the
6646 messages generated by the underlying test were being passed through. This
6647 no longer happens. The underlying message is available in $acl_verify_
6648 message for both "message" and "log_message" expansions, so it can be
6649 passed through if needed.
6651 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6652 new expansion $bh_ to give the encoded byte string without charset
6653 translation. Translation happens only if iconv() is available; HAVE_ICONV
6654 indicates this at build time. HEADERS_CHARSET gives the charset to
6655 translate to; headers_charset can change it in the configuration, and
6656 "headers charset" can change it in an individual filter file.
6658 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6659 that creates RFC 2047 encoded "words" labels them as that charset instead
6660 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6661 expansion operator; (ii) when Exim creates a From: line for a local
6662 message; (iii) when a header line is rewritten to include a "phrase" part.
6664 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6665 buggy, causing it to skip the first lines of messages whose message ID
6666 ended in 'D'. This would not have bitten before Exim release 4.14, because
6667 message IDs were unlikely to end in 'D' before then. The effect was to have
6668 incorrect size information for certain domains.
6670 45. #include "config.h" was missing at the start of the crypt16.c module. This
6671 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6674 46. If there was a timeout during a "random" callout check, Exim treated it as
6675 a failure of the random address, and carried on sending RSET and the real
6676 address. If the delay was just some slowness somewhere, the response to the
6677 original RCPT would be taken as a response to RSET and so on, causing
6678 mayhem of various kinds.
6680 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6681 when I implemented it. It didn't allow for the fact that some option values
6682 may legitimately be negative (e.g. size_addition), and it didn't even do
6683 the right test for positive values.
6685 48. Domain names in DNS records are case-independent. Exim always looks them up
6686 in lower case. Some resolvers return domain names in exactly the case they
6687 appear in the zone file, that is, they may contain uppercase letters. Not
6688 all resolvers do this - some return always lower case. Exim was treating a
6689 change of case by a resolver as a change of domain, similar to a widening
6690 of a domain abbreviation. This triggered its re-routing code and so it was
6691 trying to route what was effectively the same domain again. This normally
6692 caused routing to fail (because the router wouldn't handle the domain
6693 twice). Now Exim checks for this case specially, and just changes the
6694 casing of the domain that it ultimately uses when it transmits the message
6697 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6700 50. If a filter generated a file delivery with a non-absolute name (possible if
6701 no home directory exists for the router), the forbid_file option was not
6704 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6705 the existing equality matching.
6707 52. Exim was using ints instead of ino_t variables in some places where it was
6708 dealing with inode numbers.
6710 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6711 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6712 called TMPDIR, and if it finds it is different, it changes its value.
6714 54. The smtp_printf() function is now made available to local_scan() so
6715 additional output lines can be written before returning. There is also an
6716 smtp_fflush() function to enable the detection of a dropped connection.
6717 The variables smtp_input and smtp_batched_input are exported to
6720 55. Changed the default runtime configuration: the message "Unknown user"
6721 has been removed from the ACL, and instead placed on the localuser router,
6722 using the cannot_route_message feature. This means that any verification
6723 failures that generate their own messages won't get overridden. Similarly,
6724 the "Unrouteable address" message that was in the ACL for unverifiable
6725 relay addresses has also been removed.
6727 56. Added hosts_avoid_esmtp to the smtp transport.
6729 57. The exicyclog script was not checking for the esoteric option
6730 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6731 will work only if exicyclog is run under the appropriate euid.
6733 58. Following a discussion on the list, the rules by which Exim recognises line
6734 endings on incoming messages have been changed. The -dropcr and drop_cr
6735 options are now no-ops, retained only for backwards compatibility. The
6736 following line terminators are recognized: LF CRLF CR. However, special
6737 processing applies to CR:
6739 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6740 nor a local message in the state where . is a terminator.
6742 (ii) If a bare CR is encountered in a header line, an extra space is added
6743 after the line terminator so as not to end the header. The reasoning
6744 behind this is that bare CRs in header lines are most likely either
6745 to be mistakes, or people trying to play silly games.
6747 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6748 was being incorrectly given as 18 bytes larger than it should have been.
6749 This is a VOB (very old bug).
6751 60. This may never have affected anything current, but just in case it has:
6752 When the local host is found other than at the start of a list of hosts,
6753 the local host, those with the same MX, and any that follow, are discarded.
6754 When the list in question was part of a longer list of hosts, the following
6755 hosts (not currently being processed) were also being discarded. This no
6756 longer happens. I'm not sure if this situation could ever has previously
6759 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6762 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6763 header. This is something that is recommended in a new Internet Draft, and
6764 is something that is documented as being done by Sendmail. There are two
6765 possible values. For messages generated by the autoreply transport, Exim
6768 Auto-Submitted: auto-replied
6770 whereas for all other generated messages (e.g. bounces) it adds
6772 Auto-Submitted: auto-generated
6774 63. The "personal" condition in filters now includes a test for the
6775 Auto-Submitted: header. If it contains the string "auto-" the message it
6776 not considered personal.
6778 64. Added rcpt_include_affixes as a generic transport option.
6780 65. Added queue_only_override (default true).
6782 66. Added the syslog_duplication option.
6784 67. If what should have been the first header line of a message consisted of
6785 a space followed by a colon, Exim was mis-interpreting it as a header line.
6786 It isn't of course - it is syntactically invalid and should therefore be
6787 treated as the start of the message body. The misbehaviour could have
6788 caused a number of strange effects, including loss of data in subsequent
6789 header lines, and spool format errors.
6791 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6792 client host had authenticated. This control can now be exercised by an ACL
6793 for more flexibility.
6795 69. By default, callouts do not happen when testing with -bh. There is now a
6796 variant, -bhc, which does actually run the callout code, including
6797 consulting and updating the callout cache.
6799 70. Added support for saslauthd authentication, courtesy of Alexander
6802 71. If statvfs() failed on the spool or log directories while checking their
6803 size for availability, Exim confusingly gave the error "space shortage".
6804 Furthermore, in debugging mode it crashed with a floating point exception.
6805 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6806 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6807 serious problem, Exim now writes to the main and panic logs when this
6808 happens, with details of the failure. It then refuses to accept the
6809 incoming message, giving the message "spool directory problem" or "log
6810 directory problem" with a 421 code for SMTP messages.
6812 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6813 0, 1, and 2 exist, because some OS complain for execs without them (see
6814 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6815 descriptors. However, the code omitted to check that the open succeeded,
6816 causing mysterious errors if for some reason the permissions on /dev/null
6817 got screwed. Now Exim writes a message to the main and panic logs, and
6818 bombs out if it can't open /dev/null.
6820 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6821 interact so that it is all more flexible. It is supposed to remain
6822 backwards compatible. Also added extra_local_interfaces.
6824 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6825 to bomb out with an assertion failure - to the client this appears as a
6826 connection drop. This problem occurs in the part of the code that was taken
6827 from the Samba project. Fortunately, the assertion is in a very simple
6828 function, so I have fixed this by reproducing the function inline in the
6829 one place where it is called, and arranging for authentication to fail
6830 instead of killing the process with assert().
6832 75. The SPA client code was not working when the server requested OEM rather
6833 than Unicode encoding.
6835 76. Added code to make require_files with a specific uid setting more usable in
6836 the case where statting the file as root fails - usually a non-root-mounted
6837 NFS file system. When this happens and the failure is EACCES, Exim now
6838 forks a subprocess and does the per-uid checking as the relevant uid.
6840 77. Added process_log_path.
6842 78. If log_file_path was not explicitly set, a setting of check_log_space or
6843 check_log_inodes was ignored.
6845 79. If a space check for the spool or log partitions fails, the incident is now
6846 logged. Of course, in the latter case the data may get lost...
6848 80. Added the %p formatting code to string_format() so that it can be used to
6849 print addresses in debug_print(). Adjusted all the address printing in the
6850 debugging in store.c to use %p rather than %d.
6852 81. There was a concern that a line of code in smtp_in.c could overflow a
6853 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6854 initially expressed, the concern was not well-founded, because trailing
6855 spaces are removed early. However, if the trailing spaces were followed by
6856 a NULL, they did not get removed, so the overflow was possible. Two fixes
6859 (a) I re-wrote the offending code in a cleaner fashion.
6860 (b) If an incoming SMTP command contains a NULL character, it is rejected
6863 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6864 runs initgroups(), so that if the Exim user is in any additional groups,
6865 they will be used during message reception.
6871 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob