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 in any order, even if one
113 is the name of a transport.
116 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
118 JH/17 Change the list-building routines interface to use the expanding-string
119 triplet model, for better allocation and copying behaviour.
121 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
122 Previously it was constructed the first time a possibly-matching string
123 was met in the configuration file input during startup; now it is done
126 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
127 compatible one, to avoid the (poorly documented) possibility of a config
128 file in the working directory redirecting the DB files, possibly correpting
129 some existing file. CVE-2017-10140 assigned for BDB.
131 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
132 cache-hot. Previously, although the result was properly cached, the
133 initial verify call returned a defer.
135 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
136 the main verify for receipient in uncached-mode.
138 JH/22 Retire historical build files to an "unsupported" subdir. These are
139 defined as "ones for which we have no current evidence of testing".
141 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
142 if present. Previously it was ignored.
144 JH/24 Start using specified-initialisers in C structure init coding. This is
145 a C99 feature (it's 2017, so now considered safe).
147 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
148 if was a fixed-sized field and bitmask ops via macros; it is now more
155 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
156 than -2003 did; needs libidn2 in addition to libidn.
158 JH/02 The path option on a pipe transport is now expanded before use.
160 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
161 Patch provided by "Björn", documentation fix added too.
163 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
164 missing a wire-to-host endian conversion.
166 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
167 close after a BDAT command line could be taken as a following command,
168 giving a synch failure. Fix by only checking for synch immediately
169 before acknowledging the chunk.
171 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
172 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
173 macro. Patches provided by Josh Soref.
175 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
176 Previously we did not; the RFC seems ambiguous and VRFY is not listed
177 by IANA as a service extension. However, John Klensin suggests that we
180 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
181 the dkim code may be unix-mode line endings rather than smtp wire-format
182 CRLF, so prepend a CR to any bare LF.
184 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
185 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
187 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
188 the first were themselves being wrongly included in the feed into dkim
189 processing; with most chunk sizes in use this resulted in an incorrect
190 body hash calculated value.
192 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
193 DKIM signature block, for verification. Although advised against by
194 standards it is specifically not ruled illegal.
196 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
198 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
199 missing a body hash (the bh= tag).
201 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
202 It seems that HAProxy sends the Proxy Protocol information in clear and
203 only then does a TLS startup, so do the same.
205 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
206 TCP connections (such as for Spamd) unless the daemon successfully set
207 Fast Open mode on its listening sockets. This fixes breakage seen on
208 too-old kernels or those not configured for Fast Open, at the cost of
209 requiring both directions being enabled for TFO, and TFO never being used
210 by non-daemon-related Exim processes.
212 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
213 endings, at least on the first header line. Try to canonify any that get
214 past that check, despite the cost.
216 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
217 now limited to an arbitrary five deep, while parsing addresses with the
218 strip_excess_angle_brackets option enabled.
220 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
221 instead leave the unprompted TLS handshake in socket buffer for the
222 TLS library to consume.
224 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
226 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
228 JH/16 Drop variables when they go out of scope. Memory management drops a whole
229 region in one operation, for speed, and this leaves assigned pointers
230 dangling. Add checks run only under the testsuite which checks all
231 variables at a store-reset and panics on a dangling pointer; add code
232 explicitly nulling out all the variables discovered. Fixes one known
233 bug: a transport crash, where a dangling pointer for $sending_ip_address
234 originally assigned in a verify callout, is re-used.
236 PP/06 Drop '.' from @INC in various Perl scripts.
238 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
240 PP/08 Reduce a number of compilation warnings under clang; building with
241 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
242 should be warning-free.
244 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
246 HS/01 Fix portability problems introduced by PP/08 for platforms where
247 realloc(NULL) is not equivalent to malloc() [SunOS et al].
249 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
250 chunk. This allows us to accept broken chunked messages. We need a more
251 general solution here.
253 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
254 already-broken messages in the queue.
256 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
258 JH/19 Fix reference counting bug in routing-generated-address tracking.
264 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
265 supports it and a size is available (ie. the sending peer gave us one).
267 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
268 years of being deprecated). The replacements are the ACLs
269 acl_smtp_mime and acl_not_smtp_mime.
271 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
272 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
273 or even an in-clear connection were permitted. Now, if the host lookup
274 was dnssec and dane was requested then the host is only used if the
275 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
276 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
277 if one fails this test.
278 This means that a poorly-configured remote DNS will make it incommunicado;
279 but it protects against a DNS-interception attack on it.
281 JH/04 Bug 1810: make continued-use of an open smtp transport connection
282 non-noisy when a race steals the message being considered.
284 JH/05 If main configuration option tls_certificate is unset, generate a
285 self-signed certificate for inbound TLS connections.
287 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
288 in rewrites and routers.
290 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
291 and logged a warning sing 4.83; now they are a configuration file error.
293 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
294 (lacking @domain). Apply the same qualification processing as RCPT.
296 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
298 JH/10 Support ${sha256:} applied to a string (as well as the previous
301 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
302 a cutthrough deliver is pending, as we always want to make a connection.
303 This also avoids re-routing the message when later placing the cutthrough
304 connection after a verify cache hit.
305 Do not update it with the verify result either.
307 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
308 when routing results in more than one destination address.
310 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
311 signing (which inhibits the cutthrough capability). Previously only
312 the presence of an option was tested; now an expansion evaluating as
313 empty is permissible (obviously it should depend only on data available
314 when the cutthrough connection is made).
316 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
317 the relevant preceding SMTP command did not note the pipelining mode.
319 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
320 Previously they were not counted.
322 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
323 as one having no matching records. Previously we deferred the message
324 that needed the lookup.
326 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
327 distinguished as "(=".
329 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
330 for missing MX records. Previously it only worked for missing A records.
332 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
334 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
335 after the data-go-ahead and data-ack. Patch from Jason Betts.
337 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
338 even for a "none" policy. Patch from Tony Meyer.
340 JH/22 Fix continued use of a connection for further deliveries. If a port was
341 specified by a router, it must also match for the delivery to be
344 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
345 When one of the recipients of a message was unsuitable for the connection
346 (has no matching addresses), we lost track of needing to mark it
347 deferred. As a result mail would be lost.
349 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
351 JH/25 Decoding ACL controls is now done using a binary search; the source code
352 takes up less space and should be simpler to maintain. Merge the ACL
353 condition decode tables also, with similar effect.
355 JH/26 Fix problem with one_time used on a redirect router which returned the
356 parent address unchanged. A retry would see the parent address marked as
357 delivered, so not attempt the (identical) child. As a result mail would
360 JH/27 Fix a possible security hole, wherein a process operating with the Exim
361 UID can gain a root shell. Credit to http://www.halfdog.net/ for
362 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
365 JH/28 Enable {spool,log} filesystem space and inode checks as default.
366 Main config options check_{log,spool}_{inodes,space} are now
367 100 inodes, 10MB unless set otherwise in the configuration.
369 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
370 Previously it only applied to the main-section connection policy
373 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
375 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
376 by me. Added RFC7919 DH primes as an alternative.
378 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
379 are not in the system include path.
381 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
382 GnuTLS, when a session startup failed (eg because the client disconnected)
383 Exim did stdio operations after fclose. This was exposed by a recent
384 change which nulled out the file handle after the fclose.
386 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
387 signed directly by the cert-signing cert, rather than an intermediate
388 OCSP-signing cert. This is the model used by LetsEncrypt.
390 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
392 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
393 an incoming connection.
395 HS/02 Bug 1802: Do not half-close the connection after sending a request
398 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
399 fallback to "prime256v1".
401 JH/34 SECURITY: Use proper copy of DATA command in error message.
402 Could leak key material. Remotely exploitable. CVE-2016-9963.
408 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
409 and 3.4.4 - once the server is enabled to respond to an OCSP request
410 it does even when not requested, resulting in a stapling non-aware
411 client dropping the TLS connection.
413 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
414 support variable-length bit vectors. No functional change.
416 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
417 The I= interface field on outgoing lines is now after the H= remote
418 host field, same as incoming lines. There is a separate
419 outgoing_interface log selector which allows you to disable the
422 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
423 If not running log_selector +smtp_connection the mainlog would be held
424 open indefinitely after a "too many connections" event, including to a
425 deleted file after a log rotate. Leave the per net connection logging
426 leaving it open for efficiency as that will be quickly detected by the
427 check on the next write.
429 HS/01 Bug 1671: Fix post transport crash.
430 Processing the wait-<transport> messages could crash the delivery
431 process if the message IDs didn't exist for some reason. When
432 using 'split_spool_directory=yes' the construction of the spool
433 file name failed already, exposing the same netto behaviour.
435 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
436 mime_regex ACL conditions.
438 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
439 to DSN fail messages (bounces): remote IP, remote greeting, remote response
440 to HELO, local diagnostic string.
442 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
443 log line to debug. Even when configured with a tls authenticator many
444 client connections are expected to not authenticate in this way, so
445 an authenticate fail is not an error.
447 HS/02 Add the Exim version string to the process info. This way exiwhat
448 gives some more detail about the running daemon.
450 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
451 matter for fast-change records such as DNSBLs.
453 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
454 retry record, even if constant. There may be multiple transports with
455 different interface settings and the retry behaviour needs to be kept
458 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
460 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
462 JH/10 Bug 840: fix log_defer_output option of pipe transport
464 JH/11 Bug 830: use same host for all RCPTS of a message, even under
465 hosts_randomize. This matters a lot when combined with mua_wrapper.
467 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
468 ${quote_pgsql:<string>} operator.
470 JH/13 Bug 1708: avoid misaligned access in cached lookup.
472 JH/14 Change header file name for freeradius-client. Relevant if compiling
473 with Radius support; from the Gentoo tree and checked under Fedora.
475 JH/15 Bug 1712: Introduce $prdr_requested flag variable
477 JH/16 Bug 1714: Permit an empty string as expansion result for transport
478 option transport_filter, meaning no filtering.
480 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
482 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
483 defaults to "*" (all hosts). The variable is now available when not built
484 with TLS, default unset, mainly to enable keeping the testsuite sane.
485 If a server certificate is not supplied (via tls_certificate) an error is
486 logged, and clients will find TLS connections fail on startup. Presumably
487 they will retry in-clear.
488 Packagers of Exim are strongly encouraged to create a server certificate
489 at installation time.
491 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
492 with the $config_file variable.
494 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
495 in transport context, after the attempt, and per-recipient. The latter type
496 is per host attempted. The event data is the error message, and the errno
497 information encodes the lookup type (A vs. MX) used for the (first) host,
498 and the trailing two digits of the smtp 4xx response.
500 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
501 to write to mainlog (or rejectlog, paniclog) in the window between file
502 creation and permissions/ownership being changed. Particularly affects
503 installations where exicyclog is run as root, rather than exim user;
504 result is that the running daemon panics and dies.
506 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
508 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
509 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
510 "pri" and "weight". Note that the previous implicit priority given by the
511 list order is no longer honoured.
513 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
516 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
517 by defining SUPPORT_SOCKS.
519 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
520 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
521 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
522 variables are renamed to proxy_{local,external}_{address,port}.
524 JH/27 Move Internationalisation support from Experimental to mainline, enabled
525 for a build by defining SUPPORT_I18N
527 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
528 of the query string, and make ${quote_redis:} do that quoting.
530 JH/29 Move Events support from Experimental to mainline, enabled by default
531 and removable for a build by defining DISABLE_EVENT.
533 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
535 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
536 cached by the daemon.
538 JH/32 Move Redis support from Experimental to mainline, enabled for a build
539 by defining LOOKUP_REDIS. The libhiredis library is required.
541 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
542 keys are given for lookup.
544 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
545 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
546 only supported when built with TLS support. The PolarSSL SHA routines
547 are still used when the TLS library is too old for convenient support.
549 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
550 openssl_options), for security. OpenSSL forces this from version 1.1.0
551 server-side so match that on older versions.
553 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
554 allocation for $value could be released as the expansion processing
555 concluded, but leaving the global pointer active for it.
557 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
558 and to use the domains and local_parts ACL conditions.
560 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
561 incorrectly not doubled on cutthrough transmission, hence seen as a
562 body-termination at the receiving system - resulting in truncated mails.
563 Commonly the sender saw a TCP-level error, and retransmitted the message
564 via the normal store-and-forward channel. This could result in duplicates
565 received - but deduplicating mailstores were liable to retain only the
566 initial truncated version.
568 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
570 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
572 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
573 we're in there, support oversigning also; bug 1309.
575 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
577 HS/04 Add support for keep_environment and add_environment options.
579 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
580 either intentional arithmetic overflow during PRNG, or testing config-
583 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
584 delivery resulted in actual delivery. Cancel cutthrough before DATA
587 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
588 rejects a recipient: pass the reject to the originator.
590 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
591 Many were false-positives and ignorable, but it's worth fixing the
594 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
595 for the new environment-manipulation done at startup. Move the routines
596 from being local to tls.c to being global via the os.c file.
598 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
599 an extract embedded as result-arg for a map, the first arg for extract
600 is unavailable so we cannot tell if this is a numbered or keyed
601 extraction. Accept either.
607 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
610 JH/02 The smtp transport option "multi_domain" is now expanded.
612 JH/03 The smtp transport now requests PRDR by default, if the server offers
615 JH/04 Certificate name checking on server certificates, when exim is a client,
616 is now done by default. The transport option tls_verify_cert_hostnames
617 can be used to disable this per-host. The build option
618 EXPERIMENTAL_CERTNAMES is withdrawn.
620 JH/05 The value of the tls_verify_certificates smtp transport and main options
621 default to the word "system" to access the system default CA bundle.
622 For GnuTLS, only version 3.0.20 or later.
624 JH/06 Verification of the server certificate for a TLS connection is now tried
625 (but not required) by default. The verification status is now logged by
626 default, for both outbound TLS and client-certificate supplying inbound
629 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
632 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
633 Status Notification (bounce) messages are now MIME format per RFC 3464.
634 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
635 under the control of the dsn_advertise_hosts option, and routers may
636 have a dsn_lasthop option.
638 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
639 default, modifiable by a malware= option. The list separator for
640 the options can now be changed in the usual way. Bug 68.
642 JH/10 The smtp_receive_timeout main option is now expanded before use.
644 JH/11 The incoming_interface log option now also enables logging of the
645 local interface on delivery outgoing connections.
647 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
648 if the interface and destination host and port all match.
650 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
653 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
654 Patch from Andrew Lewis.
656 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
657 now supports optional time-restrictions, weighting, and priority
658 modifiers per server. Patch originally by <rommer@active.by>.
660 JH/16 The spamd_address main option now supports a mixed list of local
661 and remote servers. Remote servers can be IPv6 addresses, and
662 specify a port-range.
664 JH/17 Bug 68: The spamd_address main option now supports an optional
665 timeout value per server.
667 JH/18 Bug 1581: Router and transport options headers_add/remove can
668 now have the list separator specified.
670 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
673 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
676 JH/21 Support for the A6 type of dns record is withdrawn.
678 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
679 rather than the verbs used.
681 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
682 from 255 to 1024 chars.
684 JH/24 Verification callouts now attempt to use TLS by default.
686 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
687 are generic router options now. The defaults didn't change.
689 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
690 Original patch from Alexander Shikoff, worked over by JH.
692 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
693 files if ZCAT_COMMAND is not executable.
695 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
697 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
699 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
700 Normally benign, it bites when the pair was led to by a CNAME;
701 modern usage is to not canonicalize the domain to a CNAME target
702 (and we were inconsistent anyway for A-only vs AAAA+A).
704 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
706 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
707 when evaluating $sender_host_dnssec.
709 JH/31 Check the HELO verification lookup for DNSSEC, adding new
710 $sender_helo_dnssec variable.
712 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
714 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
716 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
718 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
719 documented as working, but never had. Support all but $spam_report.
721 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
722 added for tls authenticator.
724 HS/03 Add perl_taintmode main config option
730 TL/01 When running the test suite, the README says that variables such as
731 no_msglog_check are global and can be placed anywhere in a specific
732 test's script, however it was observed that placement needed to be near
733 the beginning for it to behave that way. Changed the runtest perl
734 script to read through the entire script once to detect and set these
735 variables, reset to the beginning of the script, and then run through
736 the script parsing/test process like normal.
738 TL/02 The BSD's have an arc4random API. One of the functions to induce
739 adding randomness was arc4random_stir(), but it has been removed in
740 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
741 function when detected.
743 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
744 cause callback expansion.
746 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
747 syntax errors in an expansion can be treated as a string instead of
748 logging or causing an error, due to the internal use of bool_lax
749 instead of bool when processing it.
751 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
752 server certificates when making smtp deliveries.
754 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
756 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
758 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
760 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
761 Merged patch from Sebastian Wiedenroth.
763 JH/05 Fix results-pipe from transport process. Several recipients, combined
764 with certificate use, exposed issues where response data items split
765 over buffer boundaries were not parsed properly. This eventually
766 resulted in duplicates being sent. This issue only became common enough
767 to notice due to the introduction of connection certificate information,
768 the item size being so much larger. Found and fixed by Wolfgang Breyha.
770 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
771 size buffer was used, resulting in syntax errors when an expansion
774 JH/07 Add support for directories of certificates when compiled with a GnuTLS
775 version 3.3.6 or later.
777 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
778 is EXPERIMENTAL_EVENT, the main-configuration and transport options
779 both become "event_action", the variables become $event_name, $event_data
780 and $event_defer_errno. There is a new variable $verify_mode, usable in
781 routers, transports and related events. The tls:cert event is now also
782 raised for inbound connections, if the main configuration event_action
785 TL/06 In test suite, disable OCSP for old versions of openssl which contained
786 early OCSP support, but no stapling (appears to be less than 1.0.0).
788 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
789 server certificate names available under the smtp transport option
790 "tls_verify_cert_hostname" now do not permit multi-component wildcard
793 JH/10 Time-related extraction expansions from certificates now use the main
794 option "timezone" setting for output formatting, and are consistent
795 between OpenSSL and GnuTLS compilations. Bug 1541.
797 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
798 encoded parameter in the incoming message. Bug 1558.
800 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
801 include certificate info, eximon was claiming there were spoolfile
804 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
806 JH/14 Log delivery-related information more consistently, using the sequence
807 "H=<name> [<ip>]" wherever possible.
809 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
810 are problematic for Debian distribution, omit them from the release
813 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
815 JH/16 Fix string representation of time values on 64bit time_t architectures.
818 JH/17 Fix a null-indirection in certextract expansions when a nondefault
819 output list separator was used.
824 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
825 checkers that were complaining about end of non-void function with no
828 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
829 This was a regression introduced in 4.83 by another bugfix.
831 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
833 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
834 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
840 TF/01 Correctly close the server side of TLS when forking for delivery.
842 When a message was received over SMTP with TLS, Exim failed to clear up
843 the incoming connection properly after forking off the child process to
844 deliver the message. In some situations the subsequent outgoing
845 delivery connection happened to have the same fd number as the incoming
846 connection previously had. Exim would try to use TLS and fail, logging
847 a "Bad file descriptor" error.
849 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
850 utilities have not been installed.
852 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
853 temporary space as the ACL may create new global variables.
855 TL/01 LDAP support uses per connection or global context settings, depending
856 upon the detected version of the libraries at build time.
858 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
859 to extract and use the src ip:port in logging and expansions as if it
860 were a direct connection from the outside internet. PPv2 support was
861 updated based on HAProxy spec change in May 2014.
863 JH/02 Add ${listextract {number}{list}{success}{fail}}.
865 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
866 Properly escape header and check for NULL return.
868 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
871 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
873 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
874 characters in header names, implemented as a verify condition.
875 Contributed by Michael Fischer v. Mollard.
877 TL/05 Rename SPF condition results err_perm and err_temp to standardized
878 results permerror and temperror. Previous values are deprecated but
879 still accepted. In a future release, err_perm and err_temp will be
880 completely removed, which will be a backward incompatibility if the
881 ACL tests for either of these two old results. Patch contributed by
882 user bes-internal on the mailing list.
884 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
886 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
887 selectors, in both main and reject logs.
889 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
892 JH/07 Add malware type "sock" for talking to simple daemon.
894 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
896 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
897 routers/transports under cutthrough routing.
899 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
900 numbers. Touch up "bool" conditional to keep the same definition.
902 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
904 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
906 JH/12 Expand items in router/transport headers_add or headers_remove lists
907 individually rather than the list as a whole. Bug 1452.
909 Required for reasonable handling of multiple headers_ options when
910 they may be empty; requires that headers_remove items with embedded
911 colons must have them doubled (or the list-separator changed).
913 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
914 view the policy declared in the DMARC record. Currently, $dmarc_status
915 is a combined value of both the record presence and the result of the
918 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
920 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
921 dnslookup router and the smtp transport (applying to the forward
924 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
925 of ldap servers used for a specific lookup. Patch provided by Heiko
928 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
929 New variable $lookup_dnssec_authenticated for observability.
931 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
932 Patch submitted by Lars Timman.
934 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
936 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
937 Requires trusted mode and valid format message id, aborts otherwise.
938 Patch contributed by Heiko Schlichting.
940 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
941 certextract with support for various fields. Bug 1358.
943 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
944 is requested by default, modifiable by smtp transport option
947 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
948 operate on certificate variables to give certificate fingerprints
949 Also new ${sha256:cert_variable}.
951 JH/23 The PRDR feature is moved from being Experimental into the mainline.
953 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
954 Christian Aistleitner.
956 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
958 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
959 file. Patch from Wolfgang Breyha.
961 JH/25 Expand the coverage of the delivery $host and $host_address to
962 client authenticators run in verify callout. Bug 1476.
964 JH/26 Port service names are now accepted for tls_on_connect_ports, to
965 align with daemon_smtp_ports. Bug 72.
967 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
968 support and error reporting did not work properly.
970 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
971 and is readable. Patch from Andrew Colin Kissa.
973 TL/14 Enhance documentation of ${run expansion and how it parses the
974 commandline after expansion, particularly in the case when an
975 unquoted variable expansion results in an empty value.
977 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
979 PP/02 Fix internal collision of T_APL on systems which support RFC3123
980 by renaming away from it. Addresses GH issue 15, reported by
983 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
985 TL/15 SECURITY: prevent double expansion in math comparison functions
986 (can expand unsanitized data). Not remotely exploitable.
993 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
995 PP/02 Make -n do something, by making it not do something.
996 When combined with -bP, the name of an option is not output.
998 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1001 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1002 $sender_host_name and config options to manage this, and basic check
1005 PP/05 DSCP support for outbound connections and control modifier for inbound.
1007 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1008 (Only plugin which currently uses this is kerberos4, which nobody should
1009 be using, but we should make it available and other future plugins might
1010 conceivably use it, even though it would break NAT; stuff *should* be
1011 using channel bindings instead).
1013 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1014 name; added for Sendmail compatibility; requires admin caller.
1015 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1016 just ignore it); requires trusted caller.
1017 Also parse but ignore: -Ac -Am -X<logfile>
1020 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1022 TL/02 Add +smtp_confirmation as a default logging option.
1024 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1025 Patch by Magnus Holmgren from 2007-02-20.
1027 TL/04 Bugzilla 1281 - Spec typo.
1028 Bugzilla 1283 - Spec typo.
1029 Bugzilla 1290 - Spec grammar fixes.
1031 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1033 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1035 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1036 mailing list by by Dmitry Isaikin.
1038 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1040 JH/02 Support "G" suffix to numbers in ${if comparisons.
1042 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1044 NM/01 Bugzilla 1197 - Spec typo
1045 Bugzilla 1196 - Spec examples corrections
1047 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1049 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1050 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1053 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1054 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1056 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1057 "acl {{name}{arg}...}", and optional args on acl condition
1060 JH/05 Permit multiple router/transport headers_add/remove lines.
1062 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1064 JH/07 Avoid using a waiting database for a single-message-only transport.
1065 Performance patch from Paul Fisher. Bugzilla 1262.
1067 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1070 JH/09 Add $headers_added variable, with content from use of ACL modifier
1071 add_header (but not yet added to the message). Bugzilla 199.
1073 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1074 Pulled from Bugzilla 817 by Wolfgang Breyha.
1076 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1078 (nb: this is the same fix as in Exim 4.80.1)
1080 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1083 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1086 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1088 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1089 advertise SMTP AUTH mechanism to us, instead of a generic
1090 protocol violation error. Also, make Exim more robust to bad
1091 data from the Dovecot auth socket.
1093 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1095 When a queue runner is handling a message, Exim first routes the
1096 recipient addresses, during which it prunes them based on the retry
1097 hints database. After that it attempts to deliver the message to
1098 any remaining recipients. It then updates the hints database using
1101 So if a recipient address works intermittently, it can get repeatedly
1102 deferred at routing time. The retry hints record remains fresh so the
1103 address never reaches the final cutoff time.
1105 This is a fairly common occurrence when a user is bumping up against
1106 their storage quota. Exim had some logic in its local delivery code
1107 to deal with this. However it did not apply to per-recipient defers
1108 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1110 This change adds a proper retry rule check during routing so that the
1111 final cutoff time is checked against the message's age. We only do
1112 this check if there is an address retry record and there is not a
1113 domain retry record; this implies that previous attempts to handle
1114 the address had the retry_use_local_parts option turned on. We use
1115 this as an approximation for the destination being like a local
1116 delivery, as in LMTP.
1118 I suspect this new check makes the old local delivery cutoff check
1119 redundant, but I have not verified this so I left the code in place.
1121 TF/02 Correct gecos expansion when From: is a prefix of the username.
1123 Test 0254 submits a message to Exim with the header
1127 When I ran the test suite under the user fanf2, Exim expanded
1128 the header to contain my full name, whereas it should have added
1129 a Resent-Sender: header. It erroneously treats any prefix of the
1130 username as equal to the username.
1132 This change corrects that bug.
1134 GF/01 DCC debug and logging tidyup
1135 Error conditions log to paniclog rather than rejectlog.
1136 Debug lines prefixed by "DCC: " to remove any ambiguity.
1138 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1140 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1141 Bug spotted by Jeremy Harris; was flawed since initial commit.
1142 Would have resulted in OCSP responses post-SNI triggering an Exim
1143 NULL dereference and crash.
1145 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1147 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1148 Bug detection, analysis and fix by Samuel Thibault.
1149 Bugzilla 1331, Debian bug #698092.
1151 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1153 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1154 Server implementation by Todd Lyons, client by JH.
1155 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1156 config variable "prdr_enable" controls whether the server
1157 advertises the facility. If the client requests PRDR a new
1158 acl_data_smtp_prdr ACL is called once for each recipient, after
1159 the body content is received and before the acl_smtp_data ACL.
1160 The client is controlled by both of: a hosts_try_prdr option
1161 on the smtp transport, and the server advertisement.
1162 Default client logging of deliveries and rejections involving
1163 PRDR are flagged with the string "PRDR".
1165 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1166 fclose(). Diagnosis by Todd Lyons.
1168 PP/17 Update configure.default to handle IPv6 localhost better.
1169 Patch by Alain Williams (plus minor tweaks).
1172 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1173 This is now consistent with GnuTLS, and is now documented: the
1174 previous undocumented portable approach to treating the option as
1175 unset was to force an expansion failure. That still works, and
1176 an empty string is now equivalent.
1178 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1179 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1180 not performing validation itself.
1182 PP/20 Added force_command boolean option to pipe transport.
1183 Patch from Nick Koston, of cPanel Inc.
1185 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1188 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1190 PP/21 Fix eximon continuous updating with timestamped log-files.
1191 Broken in a format-string cleanup in 4.80, missed when I repaired the
1192 other false fix of the same issue.
1193 Report and fix from Heiko Schlichting.
1196 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1197 Report from Prashanth Katuri.
1199 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1200 It's SecureTransport, so affects any MacOS clients which use the
1201 system-integrated TLS libraries, including email clients.
1203 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1204 using a MIME ACL for non-SMTP local injection.
1205 Report and assistance in diagnosis by Warren Baker.
1207 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1209 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1211 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1212 last id that failed so it may be referenced in subsequent ACL's.
1214 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1217 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1218 ldap library initialization, allowing self-signed CA's to be
1219 used. Also properly sets require_cert option later in code by
1220 using NULL (global ldap config) instead of ldap handle (per
1221 session). Bug diagnosis and testing by alxgomz.
1223 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1224 the src/util/ subdirectory.
1226 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1227 renamed to Transport Post Delivery Action by Jeremy Harris, as
1230 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1231 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1232 redis_servers = needs to be configured which will be used by the redis
1233 lookup. Patch from Warren Baker, of The Packet Hub.
1235 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1237 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1238 hostname or reverse DNS when processing a host list. Used suggestions
1239 from multiple comments on this bug.
1241 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1243 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1244 Missed a few lines, added it to make the runtest require no keyboard
1247 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1248 contains upper case chars. Make router use caseful_local_part.
1250 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1251 support when GnuTLS has been built with p11-kit.
1257 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1259 This, or similar/improved, will also be change PP/11 of 4.82.
1265 PP/01 Handle short writes when writing local log-files.
1266 In practice, only affects FreeBSD (8 onwards).
1267 Bugzilla 1053, with thanks to Dmitry Isaikin.
1269 NM/01 Bugzilla 949 - Documentation tweak
1271 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1274 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1276 PP/02 Implemented gsasl authenticator.
1278 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1280 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1281 `pkg-config foo` for cflags/libs.
1283 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1284 with rest of GSASL and with heimdal_gssapi.
1286 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1287 `pkg-config foo` for cflags/libs for the TLS implementation.
1289 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1290 properties get this fed in as external SSF. A number of robustness
1291 and debugging improvements to the cyrus_sasl authenticator.
1293 PP/08 cyrus_sasl server now expands the server_realm option.
1295 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1296 Patch by Jeremy Harris.
1298 PP/10 Added dbmjz lookup type.
1300 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1302 PP/12 MAIL args handles TAB as well as SP, for better interop with
1303 non-compliant senders.
1304 Analysis and variant patch by Todd Lyons.
1306 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1307 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1308 Patch from Dirk Mueller <dmueller@suse.com>
1310 PP/13 tls_peerdn now print-escaped for spool files.
1311 Observed some $tls_peerdn in wild which contained \n, which resulted
1312 in spool file corruption.
1314 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1315 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1316 or write after TLS renegotiation, which otherwise led to messages
1319 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1320 as a tracking header (ie: a signed header comes before the signature).
1321 Patch from Wolfgang Breyha.
1323 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1324 comma-sep list; embedded commas doubled.
1326 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1328 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1330 Report and patch from Dmitry Banschikov.
1332 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1333 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1334 failures. We appear to now support TLS1.1+ with Exim.
1336 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1337 lets Exim select keys and certificates based upon TLS SNI from client.
1338 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1339 before an outbound SMTP session. New log_selector, +tls_sni.
1341 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1342 NULL dereference. Report and patch from Alun Jones.
1344 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1345 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1346 Not seeing resolver debug output on NetBSD, but suspect this is a
1347 resolver implementation change.
1349 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1350 Left warnings. Added "eximon gdb" invocation mode.
1352 PP/21 Defaulting "accept_8bitmime" to true, not false.
1354 PP/22 Added -bw for inetd wait mode support.
1356 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1357 locate the relevant includes and libraries. Made this the default.
1359 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1360 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1362 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1363 This may cause build issues on older platforms.
1365 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1366 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1367 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1368 Added SNI support via GnuTLS too.
1369 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1371 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1373 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1374 Applied second patch from Janne, implementing suggestion to default
1375 multiple-strings-in-record handling to match SPF spec.
1377 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1379 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1380 read-only, out of scope).
1381 Patch from Wolfgang Breyha, report from Stuart Northfield.
1383 PP/29 Fix three issues highlighted by clang analyser static analysis.
1384 Only crash-plausible issue would require the Cambridge-specific
1385 iplookup router and a misconfiguration.
1386 Report from Marcin Mirosław.
1388 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1390 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1391 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1392 As part of this, removing so much warning spew let me fix some minor
1393 real issues in debug logging.
1395 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1396 assignment on my part. Fixed.
1398 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1399 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1400 Janne Snabb (who went above and beyond: thank you).
1402 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1403 string otherwise requires a connection and a bunch more work and it's
1404 relatively easy to get wrong. Should also expose TLS library linkage
1407 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1408 64-bit ${eval} (JH/03).
1410 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1411 GNU libc to support some of the 64-bit stuff, should not lead to
1412 conflicts. Defined before os.h is pulled in, so if a given platform
1413 needs to override this, it can.
1415 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1416 protection layer was required, which is not implemented.
1417 Bugzilla 1254, patch from Wolfgang Breyha.
1419 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1420 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1421 tls_dhparam take prime identifiers. Also unbreak combination of
1422 OpenSSL+DH_params+TLSSNI.
1424 PP/39 Disable SSLv2 by default in OpenSSL support.
1430 PP/01 Solaris build fix for Oracle's LDAP libraries.
1431 Bugzilla 1109, patch from Stephen Usher.
1433 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1435 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1438 TF/02 Fix a couple more cases where we did not log the error message
1439 when unlink() failed. See also change 4.74-TF/03.
1441 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1442 lock up or crash if it happened to be inside a call to libc when it
1443 got a SIGUSR1 from exiwhat.
1445 The SIGUSR1 handler appends the current process status to the process
1446 log which is later printed by exiwhat. It used to use the general
1447 purpose logging code to do this, but several functions it calls are
1448 not safe for signals.
1450 The new output code in the SIGUSR1 handler is specific to the process
1451 log, and simple enough that it's easy to inspect for signal safety.
1452 Removing some special cases also simplifies the general logging code.
1453 Removing the spurious timestamps from the process log simplifies
1456 TF/04 Improved ratelimit ACL condition.
1458 The /noupdate option has been deprecated in favour of /readonly which
1459 has clearer semantics. The /leaky, /strict, and /readonly update modes
1460 are mutually exclusive. The update mode is no longer included in the
1461 database key; it just determines when the database is updated. (This
1462 means that when you upgrade Exim will forget old rate measurements.)
1464 Exim now checks that the per_* options are used with an update mode that
1465 makes sense for the current ACL. For example, when Exim is processing a
1466 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1467 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1468 must specify per_mail/readonly. If you omit the update mode it defaults to
1469 /leaky where that makes sense (as before) or /readonly where required.
1471 The /noupdate option is now undocumented but still supported for
1472 backwards compatibility. It is equivalent to /readonly except that in
1473 ACLs where /readonly is required you may specify /leaky/noupdate or
1474 /strict/noupdate which are treated the same as /readonly.
1476 A useful new feature is the /count= option. This is a generalization
1477 of the per_byte option, so that you can measure the throughput of other
1478 aggregate values. For example, the per_byte option is now equivalent
1479 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1481 The per_rcpt option has been generalized using the /count= mechanism
1482 (though it's more complicated than the per_byte equivalence). When it is
1483 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1484 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1485 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1486 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1487 Note that using per_rcpt with a non-readonly update mode in more than
1488 one ACL will cause the recipients to be double-counted. (The per_mail
1489 and per_byte options don't have this problem.)
1491 The handling of very low rates has changed slightly. If the computed rate
1492 is less than the event's count (usually one) then this event is the first
1493 after a long gap. In this case the rate is set to the same as this event's
1494 count, so that the first message of a spam run is counted properly.
1496 The major new feature is a mechanism for counting the rate of unique
1497 events. The new per_addr option counts the number of different
1498 recipients that someone has sent messages to in the last time period. It
1499 behaves like per_rcpt if all the recipient addresses are different, but
1500 duplicate recipient addresses do not increase the measured rate. Like
1501 the /count= option this is a general mechanism, so the per_addr option
1502 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1503 example, measure the rate that a client uses different sender addresses
1504 with the options per_mail/unique=$sender_address. There are further
1505 details in the main documentation.
1507 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1509 TF/06 Removed a few PCRE remnants.
1511 TF/07 Automatically extract Exim's version number from tags in the git
1512 repository when doing development or release builds.
1514 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1515 Bugzilla 879. Patch from Paul Fisher.
1517 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1518 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1521 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1523 PP/05 Variable $av_failed, true if the AV scanner deferred.
1524 Bugzilla 1078. Patch from John Horne.
1526 PP/06 Stop make process more reliably on build failure.
1527 Bugzilla 1087. Patch from Heiko Schlittermann.
1529 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1530 Bugzilla 1089. Patch from Heiko Schlittermann.
1532 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1533 Bugzilla 1131. Patch from Holger Weiß.
1535 PP/09 Handle IPv6 addresses with SPF.
1536 Bugzilla 860. Patch from Wolfgang Breyha.
1538 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1540 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1543 PP/11 match_* no longer expand right-hand-side by default.
1544 New compile-time build option, EXPAND_LISTMATCH_RHS.
1545 New expansion conditions, "inlist", "inlisti".
1547 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1549 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1551 PP/14 fix log_write() format string regression from TF/03.
1552 Bugzilla 1152. Patch from Dmitry Isaikin.
1558 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1560 PP/02 Harmonised TLS library version reporting; only show if debugging.
1561 Layout now matches that introduced for other libraries in 4.74 PP/03.
1563 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1565 PP/04 New "dns_use_edns0" global option.
1567 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1570 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1571 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1573 TK/01 Updated PolarSSL code to 0.14.2.
1574 Bugzilla 1097. Patch from Andreas Metzler.
1576 PP/07 Catch divide-by-zero in ${eval:...}.
1577 Fixes bugzilla 1102.
1579 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1582 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1583 format-string attack -- SECURITY: remote arbitrary code execution.
1585 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1586 time unintentionally subject to list matching rules, letting the header
1587 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1588 arbitrary string expansion). This allowed for information disclosure.
1590 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1591 INT_MIN/-1 -- value coerced to INT_MAX.
1597 NM/01 Workaround for PCRE version dependency in version reporting
1600 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1601 This fixes portability to compilers other than gcc, notably
1602 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1604 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1605 makefiles for portability to HP-UX and POSIX correctness.
1607 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1608 Also via indented variable definition in the Makefile.
1609 (Debugging by Oliver Heesakkers).
1611 PP/02 Restore caching of spamd results with expanded spamd_address.
1612 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1614 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1615 Improves build reliability. Fix from: Frank Elsner
1617 NM/02 Fix wide character breakage in the rfc2047 coding
1618 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1620 NM/03 Allow underscore in dnslist lookups
1621 Fixes bug 1026. Patch from Graeme Fowler
1623 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1624 Code patches from Adam Ciarcinski of NetBSD.
1626 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1629 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1630 is logged, to avoid truncation. Patch from John Horne.
1632 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1633 Patch from Jakob Hirsch.
1635 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1636 SQL string expansion failure details.
1637 Patch from Andrey Oktyabrski.
1639 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1640 Patch from Simon Arlott.
1642 PP/09 New lookups functionality failed to compile on old gcc which rejects
1643 extern declarations in function scope.
1644 Patch from Oliver Fleischmann
1646 PP/10 Use sig_atomic_t for flags set from signal handlers.
1647 Check getgroups() return and improve debugging.
1648 Fixed developed for diagnosis in bug 927 (which turned out to be
1651 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1652 Patch from Mark Zealey.
1654 PP/12 Bugzilla 1056: Improved spamd server selection.
1655 Patch from Mark Zealey.
1657 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1658 Based on patch from Heiko Schlittermann.
1660 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1661 Patch from Uwe Doering, sign-off by Michael Haardt.
1663 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1664 variable declaration deep within a block. Bug and patch from
1667 PP/15 lookups-Makefile IRIX compatibility coercion.
1669 PP/16 Make DISABLE_DKIM build knob functional.
1671 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1672 Patch by Simon Arlott
1674 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1675 variable argument macros. Our copy now differs from upstream.
1681 TF/01 Failure to get a lock on a hints database can have serious
1682 consequences so log it to the panic log.
1684 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1685 controlled using the smtp_confirmation log selector.
1687 TF/03 Include the error message when we fail to unlink a spool file.
1689 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1690 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1691 for maintaining out-of-tree patches for some time.
1693 PP/01 Bugzilla 139: Documentation and portability issues.
1694 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1695 Handle per-OS dynamic-module compilation flags.
1697 PP/02 Let /dev/null have normal permissions.
1698 The 4.73 fixes were a little too stringent and complained about the
1699 permissions on /dev/null. Exempt it from some checks.
1700 Reported by Andreas M. Kirchwitz.
1702 PP/03 Report version information for many libraries, including
1703 Exim version information for dynamically loaded libraries. Created
1704 version.h, now support a version extension string for distributors
1705 who patch heavily. Dynamic module ABI change.
1707 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1708 privilege escalation vulnerability whereby the Exim run-time user
1709 can cause root to append content of the attacker's choosing to
1712 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1715 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1716 If dropping privileges for untrusted macros, we disabled normal logging
1717 on the basis that it would fail; for the Exim run-time user, this is not
1718 the case, and it resulted in successful deliveries going unlogged.
1719 Fixed. Reported by Andreas Metzler.
1725 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1726 only prepend for the Resent-* case. Fixes regression introduced in
1727 Exim 4.70 by NM/22 for Bugzilla 607.
1729 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1730 increasing numbers of administrators be bitten by this.
1732 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1734 PP/03 Bugzilla 994: added openssl_options main configuration option.
1736 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1738 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1740 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1742 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1743 without a peer certificate, leading to a segfault because of an
1744 assumption that peers always have certificates. Be a little more
1745 paranoid. Problem reported by Martin Tscholak.
1747 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1748 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1749 NB: ClamAV planning to remove STREAM in "middle of 2010".
1750 CL also introduces -bmalware, various -d+acl logging additions and
1751 more caution in buffer sizes.
1753 PP/09 Implemented reverse_ip expansion operator.
1755 PP/10 Bugzilla 937: provide a "debug" ACL control.
1757 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1759 PP/12 Bugzilla 973: Implement --version.
1761 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1763 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1765 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1767 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1768 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1769 ignore trailing whitespace.
1771 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1773 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1776 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1777 Notification from Dr Andrew Aitchison.
1779 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1780 ExtendedDetectionInfo response format.
1781 Notification from John Horne.
1783 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1786 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1787 XSL and documented dependency on system catalogs, with examples of how
1790 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1793 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1794 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1795 configuration file which is writeable by the Exim user or group.
1797 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1798 of configuration files to cover files specified with the -C option if
1799 they are going to be used with root privileges, not just the default
1802 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1803 option (effectively making it always true).
1805 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1806 files to be used while preserving root privileges.
1808 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1809 that rogue child processes cannot use them.
1811 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1812 run-time user, instead of root.
1814 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1815 Exim run-time user without dropping privileges.
1817 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1818 result string, instead of calling string_vformat() twice with the same
1821 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1822 for other users. Others should always drop root privileges if they use
1823 -C on the command line, even for a whitelisted configure file.
1825 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1827 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1833 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1834 $data_path, and $header_path variables; fixed documentation bugs and
1837 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1838 exipick to access non-standard spools, including the "frozen" queue
1841 NM/01 Bugzilla 965: Support mysql stored procedures.
1842 Patch from Alain Williams
1844 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1846 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1847 Patch from Andreas Metzler
1849 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1850 Patch from Kirill Miazine
1852 NM/05 Bugzilla 671: Added umask to procmail example.
1854 JJ/03 installed exipick 20100323.0, fixing doc bug
1856 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1857 directory. Notification and patch from Dan Rosenberg.
1859 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1861 TK/02 Improve log output when DKIM signing operation fails.
1863 MH/01 Treat the transport option dkim_domain as a colon separated
1864 list, not as a single string, and sign the message with each element,
1865 omitting multiple occurences of the same signer.
1867 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1868 Bugzilla 985, 986. Patch by Simon Arlott
1870 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1871 Patch by Simon Arlott
1873 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1874 MBX locking. Notification from Dan Rosenberg.
1880 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1882 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1884 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1886 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1888 NM/04 Bugzilla 915: Flag broken perl installation during build.
1894 TK/01 Added patch by Johannes Berg that expands the main option
1895 "spamd_address" if it starts with a dollar sign.
1897 TK/02 Write list of recipients to X-Envelope-Sender header when building
1898 the mbox-format spool file for content scanning (suggested by Jakob
1901 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1902 (http://www.dcc-servers.net/) support via dccifd. Activated by
1903 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1905 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1906 by Mark Daniel Reidel <mr@df.eu>.
1908 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1909 When building exim an external PCRE library is now needed -
1910 PCRE is a system library on the majority of modern systems.
1911 See entry on PCRE_LIBS in EDITME file.
1913 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1914 conversation. Added nologin parameter to request.
1915 Patch contributed by Kirill Miazine.
1917 TF/01 Do not log submission mode rewrites if they do not change the address.
1919 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1921 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1922 log files in place. Contributed by Roberto Lima.
1924 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1926 TF/03 Bugzilla 615: When checking the local_parts router precondition
1927 after a local_part_suffix or local_part_prefix option, Exim now
1928 does not use the address's named list lookup cache, since this
1929 contains cached lookups for the whole local part.
1931 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1932 Robert Millan. Documentation is in experimental-spec.txt.
1934 TF/04 Bugzilla 668: Fix parallel build (make -j).
1936 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1938 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1939 Patch provided by Jan Srzednicki.
1941 TF/05 Leading white space used to be stripped from $spam_report which
1942 wrecked the formatting. Now it is preserved.
1944 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1945 that they are available at delivery time.
1947 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1949 TF/08 TLS error reporting now respects the incoming_interface and
1950 incoming_port log selectors.
1952 TF/09 Produce a more useful error message if an SMTP transport's hosts
1953 setting expands to an empty string.
1955 NM/06 Bugzilla 744: EXPN did not work under TLS.
1956 Patch provided by Phil Pennock.
1958 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1959 Patch provided by Richard Godbee.
1961 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1962 acl_smtp_notquit, added index entry.
1964 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1965 Patch provided by Eugene Bujak.
1967 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1968 accept(). Patch provided by Maxim Dounin.
1970 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1971 Patch provided by Phil Pennock.
1973 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1975 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1976 Patch provided by Brad "anomie" Jorsch.
1978 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1979 Patch provided by Dean Brooks.
1981 TK/05 Add native DKIM support (does not depend on external libraries).
1983 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1984 Patch provided by Graeme Fowler.
1986 NM/16 Bugzilla 851: Documentation example syntax fix.
1988 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1990 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1993 NM/19 Bugzilla 745: TLS version reporting.
1994 Patch provided by Phil Pennock.
1996 NM/20 Bugzilla 167: bool: condition support.
1997 Patch provided by Phil Pennock.
1999 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2000 clients. Patch provided by Phil Pennock.
2002 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2003 Patch provided by Brad "anomie" Jorsch.
2005 NM/23 Bugzilla 687: Fix misparses in eximstats.
2006 Patch provided by Heiko Schlittermann.
2008 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2009 Patch provided by Heiko Schlittermann.
2011 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2012 plus update to original patch.
2014 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2016 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2017 Patch provided by David Brownlee.
2019 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2021 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2023 NM/30 Bugzilla 888: TLS documentation bugfixes.
2025 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2027 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2028 Unlike the original bugzilla I have changed all shell scripts in src tree.
2030 NM/33 Bugzilla 898: Transport filter timeout fix.
2031 Patch by Todd Rinaldo.
2033 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2034 Patch by Serge Demonchaux.
2036 NM/35 Bugzilla 39: Base64 decode bug fixes.
2037 Patch by Jakob Hirsch.
2039 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2041 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2043 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2045 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2051 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2052 ALT-N's libdkim that I have put here:
2053 http://duncanthrax.net/exim-experimental/
2055 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2056 were called 'true' and it seems that C99 defines that as a reserved
2057 keyword to be used with 'bool' variable types. That means you could
2058 not include C99-style headers which use bools without triggering
2059 build errors in sieve.c.
2061 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2062 as mailq or other aliases. Changed the --help handling significantly
2063 to do whats expected. exim_usage() emits usage/help information.
2065 SC/01 Added the -bylocaldomain option to eximstats.
2067 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2069 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2071 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2077 PH/01 Another patch from the Sieve maintainer.
2079 PH/02 When an IPv6 address is converted to a string for single-key lookup
2080 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2081 dots are used instead of colons so that keys in lsearch files need not
2082 contain colons. This was done some time before quoting was made available
2083 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2084 (notated using the quote facility) so as to distinguish them from IPv4
2085 keys. This meant that lookups for IP addresses in host lists did not work
2086 for iplsearch lookups.
2088 This has been fixed by arranging for IPv6 addresses to be expressed with
2089 colons if the lookup type is iplsearch. This is not incompatible, because
2090 previously such lookups could never work.
2092 The situation is now rather anomalous, since one *can* have colons in
2093 ordinary lsearch keys. However, making the change in all cases is
2094 incompatible and would probably break a number of configurations.
2096 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2099 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2100 conversion specification without a maximum field width, thereby enabling
2101 a rogue spamd server to cause a buffer overflow. While nobody in their
2102 right mind would setup Exim to query an untrusted spamd server, an
2103 attacker that gains access to a server running spamd could potentially
2104 exploit this vulnerability to run arbitrary code as the Exim user.
2106 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2107 $primary_hostname instead of what libspf2 thinks the hosts name is.
2109 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2110 a directory entry by the name of the lookup key. Previously, if a
2111 symlink pointed to a non-existing file or a file in a directory that
2112 Exim lacked permissions to read, a lookup for a key matching that
2113 symlink would fail. Now it is enough that a matching directory entry
2114 exists, symlink or not. (Bugzilla 503.)
2116 PH/03 The body_linecount and body_zerocount variables are now exported in the
2119 PH/04 Added the $dnslist_matched variable.
2121 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2122 This means they are set thereafter only if the connection becomes
2125 PH/06 Added the client_condition to authenticators so that some can be skipped
2126 by clients under certain conditions.
2128 PH/07 The error message for a badly-placed control=no_multiline_responses left
2129 "_responses" off the end of the name.
2131 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2133 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2134 (without spaces) instead of just copying the configuration text.
2136 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2138 PH/11 Added $max_received_linelength.
2140 PH/12 Added +ignore_defer and +include_defer to host lists.
2142 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2143 way in which PCRE > 7.0 is built.
2145 PH/14 Implemented queue_only_load_latch.
2147 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2148 MAIL command. The effect was to mangle the value on 64-bit systems.
2150 PH/16 Another patch from the Sieve maintainer.
2152 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2154 PH/18 If a system quota error occurred while trying to create the file for
2155 a maildir delivery, the message "Mailbox is full" was not appended to the
2156 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2157 only to a quota excession during the actual writing of the file.
2159 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2160 characters?) which causes problems in log lines. The DN values are now
2161 passed through string_printing() before being added to log lines.
2163 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2164 and InterBase are left for another time.)
2166 PH/21 Added message_body_newlines option.
2168 PH/22 Guard against possible overflow in moan_check_errorcopy().
2170 PH/23 POSIX allows open() to be a macro; guard against that.
2172 PH/24 If the recipient of an error message contained an @ in the local part
2173 (suitably quoted, of course), incorrect values were put in $domain and
2174 $local_part during the evaluation of errors_copy.
2180 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2181 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2184 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2185 issue a MAIL command.
2187 PH/02 In an ACL statement such as
2189 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2191 if a client was not listed at all, or was listed with a value other than
2192 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2193 the condition was not true (as it should be), so access was not denied.
2194 The bug was that the ! inversion was incorrectly passed on to the second
2195 item. This has been fixed.
2197 PH/03 Added additional dnslists conditions == and =& which are different from
2198 = and & when the dns lookup returns more than one IP address.
2200 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2201 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2203 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2204 FSYNC, which compiles an option called disable_fsync that allows for
2205 bypassing fsync(). The documentation is heavily laced with warnings.
2207 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2209 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2210 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2211 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2212 including adding "make clean"; (3) Added -fPIC when compiling the test
2213 dynamically loaded module, to get rid of a warning.
2215 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2216 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2217 = 0s. The bug is otherwise harmless.
2219 PH/07 There was a bug in the dovecot authenticator such that the value of
2220 $auth1 could be overwritten, and so not correctly preserved, after a
2221 successful authentication. This usually meant that the value preserved by
2222 the server_setid option was incorrect.
2224 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2226 PH/09 Installed PCRE release 7.0.
2228 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2229 run for batched SMTP input. It is now run at the start of every message
2230 in the batch. While fixing this I discovered that the process information
2231 (output by running exiwhat) was not always getting set for -bs and -bS
2232 input. This is fixed, and it now also says "batched" for BSMTP.
2234 PH/11 Added control=no_pipelining.
2236 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2237 patch, slightly modified), and move the expansion of helo_data till after
2238 the connection is made in the smtp transport (so it can use these
2241 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2243 PH/14 Added log_selector = +pid.
2245 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2247 PH/16 Add ${if forany and ${if forall.
2249 PH/17 Added dsn_from option to vary the From: line in DSNs.
2251 PH/18 Flush SMTP output before performing a callout, unless control =
2252 no_callout_flush is set.
2254 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2255 was true (the default) a successful delivery failed to delete the retry
2256 item, thus causing premature timeout of the address. The bug is now
2259 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2261 PH/21 Long custom messages for fakedefer and fakereject are now split up
2262 into multiline responses in the same way that messages for "deny" and
2263 other ACL rejections are.
2265 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2266 with slight modification.
2268 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2269 draft, changing the syntax and factoring some duplicate code".
2271 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2272 for deliveries of the second and subsequent messages over the same SMTP
2275 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2276 ${reduce, with only minor "tidies".
2278 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2280 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2281 expansion side effects.
2283 PH/27 When a message times out after an over-quota error from an Exim-imposed
2284 quota, the bounce message says "mailbox is full". This message was not
2285 being given when it was a system quota that was exceeded. It now should
2288 MH/03 Made $recipients available in local_scan(). local_scan() already has
2289 better access to the recipient list through recipients_list[], but
2290 $recipients can be useful in postmaster-provided expansion strings.
2292 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2293 in the case of a MAIL command with additional options following the
2294 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2295 were accidentally chopped off.
2297 PH/29 SMTP synchronization checks are implemented when a command is read -
2298 there is a check that no more input is waiting when there shouldn't be
2299 any. However, for some commands, a delay in an ACL can mean that it is
2300 some time before the response is written. In this time, more input might
2301 arrive, invalidly. So now there are extra checks after an ACL has run for
2302 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2303 pipelining has not been advertised.
2305 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2307 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2308 returned in the SMTP response when smtp_return_error_details was false.
2309 This has been fixed.
2311 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2312 instead of the C I/O that was originally supplied, because problems were
2313 reported on Solaris.
2315 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2316 Exim which did not show up earlier: it was assuming that a call to
2317 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2318 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2319 was a macro that became an assignment, so it seemed to work. This has
2320 changed to a proper function call with a void return, hence the compile
2321 error. Exim's code has been fixed.
2323 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2326 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2328 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2330 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2331 because it turns out to be too big or there is a timeout) there is no
2332 "Completed" line in the log. When some messages of this type were
2333 selected by exigrep, they were listed as "not completed". Others were
2334 picked up by some special patterns. I have improved the selection
2335 criteria to be more general.
2337 PH/38 The host_find_failed option in the manualroute router can now be set
2338 to "ignore", to completely ignore a host whose IP address cannot be
2339 found. If all hosts are ignored, the behaviour is controlled by the new
2340 host_all_ignored option.
2342 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2343 homing or because of multiple MX records with /mx) generated more than
2344 one IP address, and the following item turned out to be the local host,
2345 all the secondary addresses of the first item were incorrectly removed
2346 from the list, along with the local host and any following hosts (which
2347 is what is supposed to happen).
2349 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2350 whoever called Exim into the -H file. In the case of the daemon it was
2351 behaving confusingly. When first started, it used values for whoever
2352 started the daemon, but after a SIGHUP it used the Exim user (because it
2353 calls itself on a restart). I have changed the code so that it now always
2356 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2357 message are rejected with the same error (e.g. no authentication or bad
2358 sender address), and a DATA command is nevertheless sent (as can happen
2359 with PIPELINING or a stupid MUA), the error message that was given to the
2360 RCPT commands is included in the rejection of the DATA command. This is
2361 intended to be helpful for MUAs that show only the final error to their
2364 PH/42 Another patch from the Sieve maintainer.
2366 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2367 Eximstats - Fixed some broken HTML links and added missing column headers
2369 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2370 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2372 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2374 PH/43 Yet another patch from the Sieve maintainer.
2376 PH/44 I found a way to check for a TCP/IP connection going away before sending
2377 the response to the final '.' that terminates a message, but only in the
2378 case where the client has not sent further data following the '.'
2379 (unfortunately, this is allowed). However, in many cases there won't be
2380 any further data because there won't be any more messages to send. A call
2381 to select() can be used: if it shows that the input is "ready", there is
2382 either input waiting, or the socket has been closed. An attempt to read
2383 the next input character can distinguish the two cases. Previously, Exim
2384 would have sent an OK response which the client would never have see.
2385 This could lead to message repetition. This fix should cure that, at
2386 least in a lot of common cases.
2388 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2389 advertised in response to EHLO.
2395 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2396 fixed by 4.65/MH/01 (is this a record?) are fixed:
2398 (i) An empty string was always treated as zero by the numeric comparison
2399 operators. This behaviour has been restored.
2401 (ii) It is documented that the numeric comparison operators always treat
2402 their arguments as decimal numbers. This was broken in that numbers
2403 starting with 0 were being interpreted as octal.
2405 While fixing these problems I realized that there was another issue that
2406 hadn't been noticed. Values of message_size_limit (both the global option
2407 and the transport option) were treated as octal if they started with 0.
2408 The documentation was vague. These values are now always treated as
2409 decimal, and I will make that clear in the documentation.
2415 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2416 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2419 MH/01 Don't check that the operands of numeric comparison operators are
2420 integers when their expansion is in "skipping" mode (fixes bug
2421 introduced by 4.64-PH/07).
2423 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2424 child addresses, Exim now panics and dies. Previously, because the count
2425 is held in a short int, deliveries were likely to be lost. As such a
2426 large number of recipients for a single message is ridiculous
2427 (performance will be very, very poor), I have chosen to impose a limit
2428 rather than extend the field.
2434 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2435 leftover -K file (the existence of which was triggered by #402).
2436 While we were at it, introduced process PID as part of the -K
2437 filename. This should rule out race conditions when creating
2440 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2441 processing considerably. Previous code took too long for large mails,
2442 triggering a timeout which in turn triggers #401.
2444 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2445 in the DK code in transports.c. sendfile() is not really portable,
2446 hence the _LINUX specificness.
2448 TF/01 In the add_headers option to the mail command in an Exim filter,
2449 there was a bug that Exim would claim a syntax error in any
2450 header after the first one which had an odd number of characters
2453 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2454 callout verification, Exim cached a "reject" for the entire domain. This
2455 is correct for most verifications, but it is not correct for a recipient
2456 verification with use_sender or use_postmaster set, because in that case
2457 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2458 case of MAIL FROM:<> rejection from other early rejections (e.g.
2459 rejection of HELO). When verifying a recipient using a non-null MAIL
2460 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2461 Whatever the result of the callout, the value of the domain cache is
2462 left unchanged (for any other kind of callout, getting as far as trying
2463 RCPT means that the domain itself is ok).
2465 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2468 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2469 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2470 session, the daemon ignores ECONNECT errors and logs others; it now
2471 ignores EPIPE as well.
2473 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2474 (quoted-printable decoding).
2476 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2477 later the small subsequent patch to fix an introduced bug.
2479 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2481 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2483 PH/08 An error is now given if message_size_limit is specified negative.
2485 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2486 to be given (somewhat) arbitrary names.
2488 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2491 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2492 miscellaneous code fixes
2494 PH/10 Added the log_reject_target ACL modifier to specify where to log
2497 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2498 hostname. This is wrong, because it relates to the incoming message (and
2499 probably the interface on which it is arriving) and not to the outgoing
2500 callout (which could be using a different interface). This has been
2501 changed to use the value of the helo_data option from the smtp transport
2502 instead - this is what is used when a message is actually being sent. If
2503 there is no remote transport (possible with a router that sets up host
2504 addresses), $smtp_active_hostname is used.
2506 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2507 tweaks were necessary in order to get it to work (see also 21 below):
2508 (a) The code assumed that strncpy() returns a negative number on buffer
2509 overflow, which isn't the case. Replaced with Exim's string_format()
2511 (b) There were several signed/unsigned issues. I just did the minimum
2512 hacking in of casts. There is scope for a larger refactoring.
2513 (c) The code used strcasecmp() which is not a standard C function.
2514 Replaced with Exim's strcmpic() function.
2515 (d) The code set only $1; it now sets $auth1 as well.
2516 (e) A simple test gave the error "authentication client didn't specify
2517 service in request". It would seem that Dovecot has changed its
2518 interface. Fortunately there's a specification; I followed it and
2519 changed what the client sends and it appears to be working now.
2521 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2524 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2525 address A is aliased to B and C, where B exists and C does not. Without
2526 -v the output is "A verified" because verification stops after a
2527 successful redirection if more than one address is generated. However,
2528 with -v the child addresses are also verified. Exim was outputting "A
2529 failed to verify" and then showing the successful verification for C,
2530 with its parentage. It now outputs "B failed to verify", showing B's
2531 parentage before showing the successful verification of C.
2533 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2534 look up a TXT record in a specific list after matching in a combined
2537 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2538 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2539 they consult the DNS. I had assumed they would set it the way they
2540 wanted; and indeed my experiments on Linux seem to show that in some
2541 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2542 To be on the safe side, however, I have now made the interface to
2543 host_find_byname() similar to host_find_bydns(), with an argument
2544 containing the DNS resolver options. The host_find_byname() function now
2545 sets these options at its start, just as host_find_bydns() does. The smtp
2546 transport options dns_qualify_single and dns_search_parents are passed to
2547 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2548 of host_find_byname() use the default settings of RES_DEFNAMES
2549 (qualify_single) but not RES_DNSRCH (search_parents).
2551 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2552 spool_read_header() do less string testing, by means of a preliminary
2553 switch on the second character of optional "-foo" lines. (This is
2554 overdue, caused by the large number of possibilities that now exist.
2555 Originally there were few.) While I was there, I also converted the
2556 str(n)cmp tests so they don't re-test the leading "-" and the first
2557 character, in the hope this might squeeze out yet more improvement.
2559 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2560 flag allowing group syntax was set by the header_syntax check but not
2561 turned off, possible causing trouble later; (2) The flag was not being
2562 set at all for the header_verify test, causing "group"-style headers to
2563 be rejected. I have now set it in this case, and also caused header_
2564 verify to ignore an empty address taken from a group. While doing this, I
2565 came across some other cases where the code for allowing group syntax
2566 while scanning a header line wasn't quite right (mostly, not resetting
2567 the flag correctly in the right place). These bugs could have caused
2568 trouble for malformed header lines. I hope it is now all correct.
2570 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2571 with the "reply" argument non-NULL. The code, however (which originally
2572 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2573 but it didn't always do it. This confused somebody who was copying the
2574 code for some other use. I have removed all the tests.
2576 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2577 feature that was used to support insecure browsers during the U.S. crypto
2578 embargo. It requires special client support, and Exim is probably the
2579 only MTA that supported it -- and would never use it because real RSA is
2580 always available. This code has been removed, because it had the bad
2581 effect of slowing Exim down by computing (never used) parameters for the
2582 RSA_EXPORT functionality.
2584 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2585 authenticator to fail if there's a tab character in the incoming data
2586 (there should never be unless someone is messing about, as it's supposed
2587 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2588 if the connection is using TLS or if the remote IP is the same as the
2589 local IP, and the "valid-client-cert option" if a client certificate has
2592 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2593 authenticators. This can be used for authorization after authentication
2594 succeeds. (In the case of plaintext, it servers for both authentication
2597 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2598 if any retry times were supplied.
2600 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2601 connection, where there is no client IP address to check. In this
2602 situation, the verify now always succeeds.
2604 PH/25 Applied John Jetmore's -Mset patch.
2606 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2608 PH/27 In a string expansion for a processed (not raw) header when multiple
2609 headers of the same name were present, leading whitespace was being
2610 removed from all of them, but trailing whitespace was being removed only
2611 from the last one. Now trailing whitespace is removed from each header
2612 before concatenation. Completely empty headers in a concatenation (as
2613 before) are ignored.
2615 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2616 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2618 PH/29 [Removed. This was a change that I later backed out, and forgot to
2619 correct the ChangeLog entry (that I had efficiently created) before
2620 committing the later change.]
2622 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2623 address errors (4xx response to RCPT) over the same connection as other
2624 messages routed to the same hosts. Such deliveries are always "forced",
2625 so retry times are not inspected. This resulted in far too many retries
2626 for the affected addresses. The effect occurred only when there were more
2627 hosts than the hosts_max_try setting in the smtp transport when it had
2628 the 4xx errors. Those hosts that it had tried were not added to the list
2629 of hosts for which the message was waiting, so if all were tried, there
2630 was no problem. Two fixes have been applied:
2632 (i) If there are any address or message errors in an SMTP delivery, none
2633 of the hosts (tried or untried) are now added to the list of hosts
2634 for which the message is waiting, so the message should not be a
2635 candidate for sending over the same connection that was used for a
2636 successful delivery of some other message. This seems entirely
2637 reasonable: after all the message is NOT "waiting for some host".
2638 This is so "obvious" that I'm not sure why it wasn't done
2639 previously. Hope I haven't missed anything, but it can't do any
2640 harm, as the worst effect is to miss an optimization.
2642 (ii) If, despite (i), such a delivery is accidentally attempted, the
2643 routing retry time is respected, so at least it doesn't keep
2644 hammering the server.
2646 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2647 in ${readsocket because some servers need this prod.
2649 PH/32 Added some extra debug output when updating a wait-xxx database.
2651 PH/33 The hint "could be header name not terminated by colon", which has been
2652 given for certain expansion errors for a long time, was not being given
2653 for the ${if def:h_colon_omitted{... case.
2655 PH/34 The spec says: "With one important exception, whenever a domain list is
2656 being scanned, $domain contains the subject domain." There was at least
2657 one case where this was not true.
2659 PH/35 The error "getsockname() failed: connection reset by peer" was being
2660 written to the panic log as well as the main log, but it isn't really
2661 panic-worthy as it just means the connection died rather early on. I have
2662 removed the panic log writing for the ECONNRESET error when getsockname()
2665 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2666 runs only) independently of the message's sender address. This meant
2667 that, if the 4xx error was in fact related to the sender, a different
2668 message to the same recipient with a different sender could confuse
2669 things. In particular, this can happen when sending to a greylisting
2670 server, but other circumstances could also provoke similar problems.
2671 I have changed the default so that the retry time for these errors is now
2672 based a combination of the sender and recipient addresses. This change
2673 can be overridden by setting address_retry_include_sender=false in the
2676 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2677 remote server are returned as part of bounce messages. This was not
2678 happening for LMTP over a pipe (the lmtp transport), but now it is the
2679 same for both kinds of LMTP.
2681 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2682 in header lines that were in fact CNAMEs. This is no longer the case.
2684 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2685 and queue runs started by the daemon processed all messages. This has
2686 been fixed so that -R and -S can now usefully be given with -q<time>.
2688 PH/40 Import PCRE release 6.7 (fixes some bugs).
2690 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2692 PH/42 Give an error if -q is specified more than once.
2694 PH/43 Renamed the variables $interface_address and $interface_port as
2695 $received_ip_address and $received_port, to make it clear that these
2696 values apply to message reception, and not to the outgoing interface when
2697 a message is delivered. (The old names remain recognized, of course.)
2699 PH/44 There was no timeout on the connect() call when using a Unix domain
2700 socket in the ${readsocket expansion. There now is.
2702 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2703 be meaningful with "accept".
2705 SC/01 Eximstats V1.43
2706 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2708 SC/02 Eximstats V1.44
2709 Use a glob alias rather than an array ref in the generated
2710 parser. This improves both readability and performance.
2712 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2713 Collect SpamAssassin and rejection statistics.
2714 Don't display local sender or destination tables unless
2715 there is data to show.
2716 Added average volumes into the top table text output.
2718 SC/04 Eximstats V1.46
2719 Collect data on the number of addresses (recipients)
2720 as well as the number of messages.
2722 SC/05 Eximstats V1.47
2723 Added 'Message too big' to the list of mail rejection
2724 reasons (thanks to Marco Gaiarin).
2726 SC/06 Eximstats V1.48
2727 Mainlog lines which have GMT offsets and are too short to
2728 have a flag are now skipped.
2730 SC/07 Eximstats V1.49 (Alain Williams)
2731 Added the -emptyok flag.
2733 SC/08 Eximstats V1.50
2734 Fixes for obtaining the IP address from reject messages.
2736 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2737 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2738 whitespace changes from 4.64-PH/27
2740 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2743 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2744 are found, allow negative numbers in numeric criteria)
2746 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2748 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2749 to match changes made in 4.64-PH/43
2751 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2753 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2754 "domain" is actually an IP address, and force a failure. This locks out
2755 those revolvers/nameservers that support "A-for-A" lookups, in
2756 contravention of the specifications.
2758 PH/48 When a host name was looked up from an IP address, and the subsequent
2759 forward lookup of the name timed out, the host name was left in
2760 $sender_host_name, contrary to the specification.
2762 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2763 restricted to single-key lookups, Exim was not diagnosing an error if
2764 * or *@ was used with a query-style lookup.
2766 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2768 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2769 long ago, but noone interested enough thought of it. Let's just say that
2770 the "1.1" means that there are some new functions that weren't there at
2771 some point in the past.
2773 PH/51 Error processing for expansion failure of helo_data from an smtp
2774 transport during callout processing was broken.
2776 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2777 tested/used via the -bh/-bhc/-bs options.
2779 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2780 bug, fixed in subsequent PCRE releases).
2782 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2783 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2785 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2791 SC/01 Use a glob alias rather than an array ref in eximstats generated
2792 parser. This improves both readability and performance.
2794 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2795 Don't display local sender or destination tables in eximstats unless
2796 there is data to show.
2797 Added average volumes into the eximstats top table text output.
2799 SC/03 Collect data on the number of addresses (recipients) as well
2800 as the number of messages in eximstats.
2802 TF/01 Correct an error in the documentation for the redirect router. Exim
2803 does (usually) call initgroups() when daemonizing.
2805 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2806 with consistent privilege compared to when running as a daemon.
2808 TF/03 Note in the spec that $authenticated_id is not set for local
2809 submissions from trusted users.
2811 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2812 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2814 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2815 by adding some example configuration directives to the default
2816 configuration file. A little bit of work is required to uncomment the
2817 directives and define how usernames and passwords are checked, but
2818 there is now a framework to start from.
2820 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2821 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2822 without this. I don't know how relevant this is to other LDAP libraries.
2824 PH/02 Add the verb name to the "unknown ACL verb" error.
2826 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2828 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2830 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2831 directory not expanded when it should be if an expanded home directory
2832 was set for the address (which is overridden by the transport).
2834 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2837 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2838 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2839 because it is too late at that time, and has no effect.
2841 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2842 security issue with \' (bugzilla #107). I could not use the
2843 PQescapeStringConn() function, because it needs a PGconn value as one of
2846 PH/08 When testing addresses using -bt, indicate those final addresses that
2847 are duplicates that would not cause an additional delivery. At least one
2848 person was confused, thinking that -bt output corresponded to deliveries.
2849 (Suppressing duplicates isn't a good idea as you lose the information
2850 about possibly different redirections that led to the duplicates.)
2852 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2853 systems where poll() doesn't work, in particular OS X.
2855 PH/10 Added more information to debugging output for retry time not reached.
2857 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2858 operations in malware.c.
2860 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2863 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2864 syslog_duplication set false, log lines that would normally be written
2865 both the the main log and to the reject log were not written to syslog at
2868 PH/14 In the default configuration, change the use of "message" in ACL warn
2869 statements to "add_header".
2871 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2872 not followed by a command (e.g. "seen endif").
2874 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2875 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2878 PH/17 Added extra conditions to the default value of delay_warning_condition
2882 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2883 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2884 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2887 The Auto-Submitted: and various List- headers are standardised, whereas I
2888 don't think Precedence: ever was.
2890 PH/18 Refactored debugging code in route_finduser() to show more information,
2891 in particular, the error code if getpwnam() issues one.
2893 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2894 This is apparently needed in addition to the PH/07 change above to avoid
2895 any possible encoding problems.
2897 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2898 but not after initializing Perl.
2900 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2901 output them only if debugging. By default they are written stderr,
2902 apparently, which is not desirable.
2904 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2907 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2910 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2912 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2913 authenticated or an ident call has been made. Suppress the default
2914 values for $authenticated_id and $authenticated_sender (but permit -oMai
2915 and -oMas) when testing with -bh.
2917 PH/24 Re-jigged the order of the tests in the default configuration so that the
2918 tests for valid domains and recipients precede the DNS black list and CSA
2919 tests, on the grounds that those ones are more expensive.
2921 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2922 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2923 This bug exists in every version of Exim that I still have, right back to
2926 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2927 However, an attempt to turn on case-sensitivity in a regex key by
2928 including (?-i) didn't work because the subject string was already
2929 lowercased, and the effects were non-intuitive. It turns out that a
2930 one-line patch can be used to allow (?-i) to work as expected.
2936 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2937 other effects) broke the use of negated acl sub-conditions.
2939 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2942 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2943 "Deny" causes Exim to reject the incoming connection with a 554 error.
2944 Unfortunately, if there is a major crisis, such as a disk failure,
2945 tcp-wrappers gives "deny", whereas what one would like would be some
2946 kind of temporary error. A kludge has been added to help with this.
2947 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2948 554 error is used if errno is still zero or contains ENOENT (which occurs
2949 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2952 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2954 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2955 errors. Otherwise a message that provokes a temporary error (when other
2956 messages do not) can cause a whole host to time out.
2958 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2959 addresses were routed directly to files or pipes from a redirect router.
2960 File deliveries just didn't batch; pipe deliveries might have suffered
2963 PH/06 A failure to get a lock for a hints database would erroneously always say
2964 "Failed to get write lock", even when it was really a read lock.
2966 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2967 of 0600. This has been changed to use the value of the lockfile_mode
2968 option (which defaults to 0600).
2970 PH/08 Applied small patch from the Sieve maintainer.
2972 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2973 folder from quota calculations, a direct delivery into this folder messed
2974 up the contents of the maildirsize file. This was because the regex was
2975 used only to exclude .Trash (or whatever) when the size of the mailbox
2976 was calculated. There was no check that a delivery was happening into an
2977 excluded directory. This bug has been fixed by ignoring all quota
2978 processing for deliveries into excluded directories.
2980 PH/10 Added the maildirfolder_create_regex option to appendfile.
2986 PH/01 The code for finding all the local interface addresses on a FreeBSD
2987 system running IPv6 was broken. This may well have applied to all BSD
2988 systems, as well as to others that have similar system calls. The broken
2989 code found IPv4 interfaces correctly, but gave incorrect values for the
2990 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2991 that it would not match correctly against @[] and not recognize the IPv6
2994 PH/02 The ipliteral router was not recognizing addresses of the form user@
2995 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2997 PH/03 Added disable_ipv6.
2999 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3000 decoded form, because it is most often used to construct To: headers
3001 lines in autoreplies, and the decoded form may well be syntactically
3002 invalid. However, $reply_address has leading white space removed, and all
3003 newlines turned into spaces so that the autoreply transport does not
3006 PH/05 If group was specified without a user on a router, and no group or user
3007 was specified on a transport, the group from the router was ignored.
3009 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3010 for visible compile-time settings that can be used to change these
3011 numbers, for those that want even more. Backwards compatibility with old
3012 spool files has been maintained. However, going back to a previous Exim
3013 release will lost any variables that are in spool files.
3015 PH/07 Two small changes when running in the test harness: increase delay when
3016 passing a TCP/IP connection to a new process, in case the original
3017 process has to generate a bounce, and remove special handling of
3018 127.0.0.2 (sic), which is no longer necessary.
3020 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3021 be the same on different OS.
3023 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3026 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3027 whether --show-vars was specified or not
3029 JJ/02 exipick: Added support for new ACL variable spool format introduced
3032 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3033 syntactically invalid From: or Reply-to: line, and a filter used this to
3034 generate an autoreply, and therefore failed to obtain an address for the
3035 autoreply, Exim could try to deliver to a non-existent relative file
3036 name, causing unrelated and misleading errors. What now happens is that
3037 it logs this as a hard delivery error, but does not attempt to create a
3040 PH/11 The exinext utility has a -C option for testing purposes, but although
3041 the given file was scanned by exinext itself; it wasn't being passed on
3042 when Exim was called.
3044 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3045 an end-of-file indication when reading a command response.
3047 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3048 compiled. In many other places in Exim, IPv6 addresses are always
3049 recognized, so I have changed this. It also means that IPv4 domain
3050 literals of the form [IPV4:n.n.n.n] are now always recognized.
3052 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3053 used if the router is not running as root, for example, when verifying at
3054 ACL time, or when using -bh. The debugging output from this situation was
3055 non-existent - all you got was a failure to exec. I have made two
3058 (a) Failures to set uid/gid, the current directory, or a process leader
3059 in a subprocess such as that created by queryprogram now generate
3060 suitable debugging output when -d is set.
3062 (b) The queryprogram router detects when it is not running as root,
3063 outputs suitable debugging information if -d is set, and then runs
3064 the subprocess without attempting to change uid/gid.
3066 PH/15 Minor change to Makefile for building test_host (undocumented testing
3069 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3070 additional section of a DNS packet that returns MX or SRV records.
3071 Instead, it always explicitly searches for A/AAAA records. This avoids
3072 major problems that occur when a DNS server includes only records of one
3073 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3074 fixed another bug: if SRV records were looked up and the corresponding
3075 address records were *not* found in the additional section, the port
3076 values from the SRV records were lost.
3078 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3079 using the correct key (the original address) when searching the retry
3080 rules in order to find which one to use for generating the retry hint.
3082 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3083 adding the default one. Similarly, if it contains a Reply-To: header, the
3084 errors_reply_to option, if set, is not used.
3086 PH/19 When calculating a retry time, Exim used to measure the "time since
3087 failure" by looking at the "first failed" field in the retry record. Now
3088 it does not use this if it is later than than the arrival time of the
3089 message. Instead it uses the arrival time. This makes for better
3090 behaviour in cases where some deliveries succeed, thus re-setting the
3091 "first failed" field. An example is a quota failure for a huge message
3092 when small messages continue to be delivered. Without this change, the
3093 "time since failure" will always be short, possible causing more frequent
3094 delivery attempts for the huge message than are intended.
3095 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3097 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3098 $1, $2, $3) because the numerical variables can be reset during some
3099 expansion items (e.g. "match"), thereby losing the authentication data.
3101 PH/21 Make -bV show the size of off_t variables so that the test suite can
3102 decide whether to run tests for quotas > 2G.
3104 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3105 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3106 filecount value is greater than 2G or if a quota value is greater than 2G
3107 on a system where the size of off_t is not greater than 4, a panic error
3110 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3111 never match. The debug and -bh output now contains an explicit error
3112 message indicating a malformed IPv4 address or mask.
3114 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3115 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3116 PH/23 above applies.
3118 PH/25 Do not write to syslog when running in the test harness. The only
3119 occasion when this arises is a failure to open the main or panic logs
3120 (for which there is an explicit test).
3122 PH/26 Added the /no_tell option to "control=freeze".
3124 PH/27 If a host name lookup failed very early in a connection, for example, if
3125 the IP address matched host_lookup and the reverse lookup yielded a name
3126 that did not have a forward lookup, an error message of the form "no IP
3127 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3128 could be logged. Now it outputs the IP address instead of "NULL".
3130 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3131 allows the sender and the authenticated sender to be set when
3132 submitting a message from within Exim. Since child_open_exim() is
3133 documented for local_scan(), the new function should be too.
3135 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3136 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3137 results in an empty string is now treated as unset.
3139 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3141 PH/31 Added sender_verify_fail logging option.
3143 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3144 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3145 tidied the source and removed it altogether.
3147 PH/33 When a queue run was abandoned because the load average was too high, a
3148 log line was always written; now it is written only if the queue_run log
3149 selector is set. In addition, the log line for abandonment now contains
3150 information about the queue run such as the pid. This is always present
3151 in "start" and "stop" lines but was omitted from the "abandon" line.
3153 PH/34 Omit spaces between a header name and the colon in the error message that
3154 is given when verify = headers_syntax fails (if there are lots of them,
3155 the message gets confusing).
3157 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3158 names, as there are now some PTR records that contain slashes. This check
3159 is only to protect against broken name servers that fall over on strange
3160 characters, so the fact that it applies to all lookups doesn't matter.
3162 PH/36 Now that the new test suite is complete, we can remove some of the
3163 special code in Exim that was needed for the old test suite. For example,
3164 sorting DNS records because real resolvers return them in an arbitrary
3165 order. The new test suite's fake resolver always returns records in the
3168 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3169 bounces) except when queue_only is set, to avoid logging races between
3170 the different processes.
3172 PH/38 Panic-die if .include specifies a non-absolute path.
3174 PH/39 A tweak to the "H" retry rule from its user.
3176 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3177 a label. They prevented compilation on older perls.
3179 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3180 a warning to be raised on newish perls.
3182 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3183 on queue. Changes to match documented behaviour of showing count of
3184 messages matching specified criteria.
3186 PH/40 Changed the default ident timeout from 30s to 5s.
3188 PH/41 Added support for the use of login_cap features, on those BSD systems
3189 that have them, for controlling the resources used by pipe deliveries.
3191 PH/42 The content-scanning code uses fopen() to create files in which to put
3192 message data. Previously it was not paying any attention to the mode of
3193 the files. Exim runs with umask(0) because the rest of the code creates
3194 files with open(), and sets the required mode explicitly. Thus, these
3195 files were ending up world-writeable. This was not a big issue, because,
3196 being within the spool directory, they were not world-accessible. I have
3197 created a function called modefopen, which takes an additional mode
3198 argument. It sets umask(777), creates the file, chmods it to the required
3199 mode, then resets the umask. All the relevant calls to fopen() in the
3200 content scanning code have been changed to use this function.
3202 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3203 to 24 hours. This avoids potential overflow problems when processing G
3204 and H retry rules. I suspect nobody ever tinkers with this value.
3206 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3208 PH/45 When the plaintext authenticator is running as a client, the server's
3209 challenges are checked to ensure they are valid base64 strings. By
3210 default, the authentication attempt is cancelled if an invalid string is
3211 received. Setting client_ignore_invalid_base64 true ignores these errors.
3212 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3213 they are received. Thus, the responses can be made to depend on the
3214 challenges. If an invalid string is ignored, an empty string is placed in
3217 PH/46 Messages that are created by the autoreply transport now contains a
3218 References: header, in accordance with RFCs 2822 and 3834.
3220 PH/47 Added authenticated_sender_force to the smtp transport.
3222 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3224 PH/49 Installed latest patch from the Sieve maintainer.
3226 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3227 was also set, the appendfile transport was unnecessarily scanning a
3228 directory of message files (e.g. for maildir delivery) to find the count
3229 of files (along with the size), even though it did not need this
3230 information. It now does the scan only if it needs to find either the
3231 size of the count of files.
3233 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3235 PH/52 Two bugs concerned with error handling when the smtp transport is
3238 (i) Exim was not creating retry information for temporary errors given
3239 for individual recipients after the DATA command when the smtp transport
3240 was used in LMTP mode. This meant that they could be retried too
3241 frequently, and not timed out correctly.
3243 (ii) Exim was setting the flag that allows error details to be returned
3244 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3245 recipients that were returned after the DATA command.
3247 PH/53 This is related to PH/52, but is more general: for any failing address,
3248 when detailed error information was permitted to be returned to the
3249 sender, but the error was temporary, then after the final timeout, only
3250 "retry timeout exceeded" was returned. Now it returns the full error as
3251 well as "retry timeout exceeded".
3253 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3254 do this, and (what is worse) MTAs that accept it.
3256 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3257 will now be deprecated.
3259 PH/56 New os.c-cygwin from the Cygwin maintainer.
3261 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3262 formats (previously only available in exim formats via -bpr, -bpru,
3263 and -bpra. Now also available in native and exiqgrep formats)
3265 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3266 with very large, slow to parse queues
3268 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3270 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3272 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3273 responses to authentication challenges, though it was showing the
3274 challenges; (ii) I've removed the CR characters from the debug output for
3277 PH/58 Allow for the insertion of a newline as well as a space when a string
3278 is turned into more than one encoded-word during RFC 2047 encoding. The
3279 Sieve code now uses this.
3281 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3282 data_4xx, lost_connection, tls_required.
3284 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3285 message was being sent as an SMTP response.
3287 PH/61 Add -l and -k options to exicyclog.
3289 PH/62 When verifying, if an address was redirected to one new address, so that
3290 verification continued, and the new address failed or deferred after
3291 having set something in $address_data, the value of $address_data was not
3292 passed back to the ACL. This was different to the case when no
3293 redirection occurred. The value is now passed back in both cases.
3295 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3296 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3297 use login_cap.h, so on its own it isn't the distinguishing feature. The
3298 new name refers directly to the setclassresources() function.
3300 PH/65 Added configuration files for NetBSD3.
3302 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3304 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3305 is preferred over IPv4.
3307 PH/68 The bounce_return_message and bounce_return_body options were not being
3308 honoured for bounces generated during the reception of non-SMTP messages.
3309 In particular, this applied to messages rejected by the ACL. This bug has
3310 been fixed. However, if bounce_return_message is true and bounce_return_
3311 body is false, the headers that are returned for a non-SMTP message
3312 include only those that have been read before the error was detected.
3313 (In the case of an ACL rejection, they have all been read.)
3315 PH/69 The HTML version of the specification is now built in a directory called
3316 spec_html instead of spec.html, because the latter looks like a path with
3317 a MIME-type, and this confuses some software.
3319 PH/70 Catch two compiler warnings in sieve.c.
3321 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3322 function verify_get_ident() calls ip_connect() to connect a socket, but
3323 if the "connect()" function timed out, ip_connect() used to close the
3324 socket. However, verify_get_ident() also closes the socket later, and in
3325 between Exim writes to the log, which may get opened at this point. When
3326 the socket was closed in ip_connect(), the log could get the same file
3327 descriptor number as the socket. This naturally causes chaos. The fix is
3328 not to close the socket in ip_connect(); the socket should be closed by
3329 the function that creates it. There was only one place in the code where
3330 this was missing, in the iplookup router, which I don't think anybody now
3331 uses, but I've fixed it anyway.
3333 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3334 well as to direct DNS lookups. Otherwise the handling of names in host
3335 lists is inconsistent and therefore confusing.
3341 PH/01 Two changes to the default runtime configuration:
3343 (1) Move the checks for relay_from_hosts and authenticated clients from
3344 after to before the (commented out) DNS black list checks.
3346 (2) Add control=submission to the relay_from_hosts and authenticated
3347 clients checks, on the grounds that messages accepted by these
3348 statements are most likely to be submissions.
3350 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3352 (1) Generate an error if the third argument for the ${prvs expansion is
3355 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3358 (3) Reset the variables that are obtained from the first argument of
3359 ${prvscheck and used in the second argument before leaving the code,
3360 because their memory is reclaimed, so using them afterwards may do
3363 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3364 one (it's much easier than Tom thought :-).
3366 (5) Because of (4), we can now allow for the use of $prvscheck_result
3367 inside the third argument.
3369 PH/03 For some reason, the default setting of PATH when running a command from
3370 a pipe transport was just "/usr/bin". I have changed it to
3373 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3374 anything to be listed in the output from -bV.
3376 PH/05 When a filter generated an autoreply, the entire To: header line was
3377 quoted in the delivery log line, like this:
3379 => >A.N.Other <ano@some.domain> <original@ddress> ...
3381 This has been changed so that it extracts the operative address. There
3382 may be more than one such address. If so, they are comma-separated, like
3385 => >ano@some.domain,ona@other.domain <original@ddress> ...
3387 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3388 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3389 not being looked up in the rDNS to get a host name, Exim was showing the
3390 IP address twice in Received: lines, even though the IP addresses were
3391 identical. For example:
3393 Received: from [1.2.3.4] (helo=[1.2.3.4])
3395 However, if the real host name was known, it was omitting the HELO data
3396 if it matched the actual IP address. This has been tidied up so that it
3397 doesn't show the same IP address twice.
3399 PH/07 When both +timestamp and +memory debugging was on, the value given by
3400 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3401 called by the debug printing, thereby overwriting the timestamp buffer.
3402 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3404 PH/08 When the original message was included in an autoreply transport, it
3405 always said "this is a copy of the message, including all the headers",
3406 even if body_only or headers_only was set. It now gives an appropriate
3409 PH/09 Applied a patch from the Sieve maintainer which:
3411 o fixes some comments
3412 o adds the (disabled) notify extension core
3413 o adds some debug output for the result of if/elsif tests
3414 o points to the current vacation draft in the documentation
3415 and documents the missing references header update
3419 o fixes a bug in processing the envelope test (when testing
3420 multiple envelope elements, the last element determined the
3423 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3424 Electronic Mail") by including:
3426 Auto-submitted: auto-generated
3428 in the messages that it generates (bounce messages and others, such as
3429 warnings). In the case of bounce messages for non-SMTP messages, there was
3430 also a typo: it was using "Auto_submitted" (underscore instead of
3431 hyphen). Since every message generated by Exim is necessarily in response
3432 to another message, thes have all been changed to:
3434 Auto-Submitted: auto-replied
3436 in accordance with these statements in the RFC:
3438 The auto-replied keyword:
3440 - SHOULD be used on messages sent in direct response to another
3441 message by an automatic process,
3443 - MUST NOT be used on manually-generated messages,
3445 - MAY be used on Delivery Status Notifications (DSNs) and Message
3446 Disposition Notifications (MDNs),
3448 - MUST NOT be used on messages generated by automatic or periodic
3449 processes, except for messages which are automatic responses to
3452 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3453 to the default Received: header definition.
3455 PH/12 Added log selector acl_warn_skipped (default on).
3457 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3458 variables because (a) they are in the wrong storage pool and (b) even if
3459 they were copied, it wouldn't work properly because of the caching.
3461 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3462 checking when decoding. Apparently there are clients that generate
3463 overlong encoded strings. Why am I not surprised?
3465 PH/15 If the first argument of "${if match_address" was not empty, but did not
3466 contain an "@" character, Exim crashed. Now it writes a panic log message
3467 and treats the condition as false.
3469 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3471 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3472 extension "envelope-auth". The code is finished and in agreement with
3473 other implementations, but there is no documentation so far and in fact,
3474 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3475 not changing the active code.
3477 Print executed "if" and "elsif" statements when debugging is used. This
3478 helps a great deal to understand what a filter does.
3480 Document more things not specified clearly in RFC3028. I had all this
3481 sorted out, when out of a sudden new issues came to my mind. Oops."
3483 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3486 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3487 canonical form (as documented). However, after a host name lookup from
3488 the IP address, check_host() was doing a simple string comparison with
3489 addresses acquired from the DNS when checking that the found name did
3490 have the original IP as one of its addresses. Since any found IPv6
3491 addresses are likely to be in abbreviated form, the comparison could
3492 fail. Luckily, there already exists a function for doing the comparison
3493 by converting both addresses to binary, so now that is used instead of
3494 the text comparison.
3496 PH/20 There was another similar case to PH/19, when a complete host name was
3497 given in a host list; looking up its IP address could give an abbreviated
3498 form, whereas the current host's name might or might not be abbreviated.
3499 The same fix has been applied.
3505 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3506 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3509 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3510 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3512 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3514 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3515 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3516 Also, for "Auto-Submitted", treat anything other than "no" as
3517 non-personal, in accordance with RFC 3834. (Previously it treated
3518 anything starting "auto-" as non-personal.)
3520 TF/01 The control=submission/name=... option had a problem with syntax
3521 errors if the name included a slash character. The /name= option
3522 now slurps the rest of the string, so it can include any characters
3523 but it must come last in the list of options (after /sender_retain
3526 PH/05 Some modifications to the interface to the fake nameserver for the new
3534 TK/01 Added the "success_on_redirect" address verification option. See
3535 NewStuff for rationale and an example.
3537 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3539 PH/02 Patch to exigrep to allow it to work on syslog lines.
3541 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3542 fread() to skip over the body file's header line, because in Cygwin the
3543 header line is locked and is inaccessible.
3545 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3546 co-exist for some time) to make it clear that it is the Exim ID that is
3547 referenced, not the Message-ID: header line.
3549 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3550 string_format() function, because snprintf() does not exist on all
3553 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3554 ${stat: expansion item.
3556 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3557 protocol synchronization error", to keep the pedants happy.
3559 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3560 well as for IRIX systems, when gcc is being used. See the host.c source
3563 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3565 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3568 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3569 transport and to the smtp transport in LMTP mode.
3571 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3573 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3574 run a filter in a subprocess. This could lead to confusion in subsequent
3575 lookups in the parent process. There should also be a search_tidyup() at
3576 the end of the subprocess.
3578 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3579 only if the host matched helo_try_verify_hosts, which caused the
3580 verification to occur when the EHLO/HELO command was issued. The ACL just
3581 tested the remembered result. Now, if a previous verification attempt has
3582 not happened, "verify = helo" does it there and then.
3584 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3586 TK/03 Fix log output including CR from clamd.
3588 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3589 exist provoked a memory error which could cause a segfault.
3591 PH/15 Installed PCRE 6.2
3593 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3595 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3596 of the problem. Specifically, suggested +O2 rather than +O1 for the
3599 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3601 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3602 record, the port was not correctly logged when the outgoing_port log
3603 selector was set (it logged the transort's default port).
3605 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3606 fallback_hosts, and "hosts" in the smtp transport.
3608 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3609 host errors such as "Connection refused".
3611 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3612 authentication with radiusclient 0.4.9:
3614 - Error returned from rc_read_config was caught wrongly
3615 - Username/password not passed on to radius server due to wrong length.
3617 The presumption is that some radiusclient API changes for 4.51/PH/17
3618 were not taken care of correctly. The code is still untested by me (my
3619 Linux distribution still has 0.3.2 of radiusclient), but it was
3620 contributed by a Radius user.
3622 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3623 expanding the "port" option of the smtp transport.
3625 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3626 while reading a MIME header. Thanks to Tom Hughes for a patch.
3628 PH/24 Include config.h inside local_scan.h so that configuration settings are
3631 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3632 that in an ACL for RCPT (for example), you can examine exactly what was
3635 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3636 commands, but it was not correctly comparing the address with the actual
3637 client host address. Thus, it would show the EHLO address in Received:
3638 header lines when this was not necessary.
3640 PH/27 Added the % operator to ${eval:}.
3642 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3643 it should be ignoring failures (because with -C, for example, it has lost
3644 privilege). It wasn't ignoring creation failures other than "already
3647 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3650 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3651 because an input line was too long, either on its own, or by virtue of
3652 too many continuations, the temporary file was not being removed, and the
3653 return code was incorrect.
3655 PH/31 Missing "BOOL" in function definition in filtertest.c.
3657 PH/32 Applied Sieve patches from the maintainer.
3659 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3661 PH/33 Added "verify = not_blind".
3663 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3664 Local/Makefile (with some defaults set). These are used in built scripts
3665 such as exicyclog, but they have never been used in the exim_install
3666 script (though there are many overriding facilities there). I have
3667 arranged that the exim_install script now takes note of these two
3670 PH/35 Installed configuration files for Dragonfly.
3672 PH/36 When a locally submitted message by a trusted user did not contain a
3673 From: header, and the sender address was obtained from -f or from an SMTP
3674 MAIL command, and the trusted user did not use -F to supply a sender
3675 name, $originator_name was incorrectly used when constructing a From:
3676 header. Furthermore, $originator_name was used for submission mode
3677 messages from external hosts without From: headers in a similar way,
3678 which is clearly wrong.
3680 PH/37 Added control=suppress_local_fixups.
3682 PH/38 When log_selector = +received_sender was set, and the addition of the
3683 sender made the log line's construction buffer exactly full, or one byte
3684 less than full, an overflow happened when the terminating "\n" was
3687 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3688 when the result of a list match is failure because a DNS lookup failed.
3690 PH/40 RM_COMMAND is now used in the building process.
3692 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3693 the "build-* directories that it finds.
3695 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3696 address in a domain literal was a prefix of an interface address.
3698 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3699 when verifying a sender address, unless rewrite_headers is false.
3701 PH/44 Wrote a long comment about why errors_to addresses are verified as
3702 recipients, not senders.
3704 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3705 the ratelimit ACL was added.
3707 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3709 PH/46 Added extra information about PostgreSQL errors to the error string.
3711 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3712 avoiding the need to install special zones in a real server. This is
3713 backwards compatible; if it can't find the fake resolver, it drops back.
3714 Thus, both old and new test suites can be run.
3716 TF/02 Added util/ratelimit.pl
3718 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3719 clock is set back in time.
3721 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3722 Candler <B.Candler@pobox.com>.
3724 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3725 OK for addresses that are the result of redirections.
3727 PH/48 A number of further additions for the benefit of the new test suite,
3728 including a fake gethostbyname() that interfaces to the fake DNS resolver
3731 TF/06 The fix for widen_domains has also been applied to qualify_single and
3732 search_parents which are the other dnslookup options that can cause
3735 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3738 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3740 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3741 service" if it encountered an MX record with an empty target hostname.
3742 The message is now "an MX or SRV record indicated no SMTP service".
3744 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3745 if the DNS of the sending site is misconfigured. This is quite a
3746 common situation. This change restores the behaviour of treating a
3747 helo verification defer as a failure.
3749 PH/51 If self=fail was set on a router, the bounce message did not include the
3750 actual error message.
3756 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3758 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3759 command itself ended in error, the underlying message about the transport
3760 filter timeout was being overwritten with the pipe command error. Now the
3761 underlying error message should be appended to the second error message.
3763 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3765 PH/02 Reduce the amount of output that "make" produces by default. Full output
3766 can still be requested.
3768 PH/03 The warning log line about a condition test deferring for a "warn" verb
3769 was being output only once per connection, rather than after each
3770 occurrence (because it was using the same function as for successful
3771 "warn" verbs). This seems wrong, so I have changed it.
3773 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3774 it should not have, which might have caused a crash in the right
3775 circumstances, but probably never did.
3777 PH/04 Installed a modified version of Tony Finch's patch to make submission
3778 mode fix the return path as well as the Sender: header line, and to
3779 add a /name= option so that you can make the user's friendly name appear
3782 TF/03 Added the control = fakedefer ACL modifier.
3784 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3785 Mark Lowes for thorough testing.
3787 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3789 TK/03 Merged latest SRS patch from Miles Wilton.
3791 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3792 with the definition in sysexits.h (which is #included earlier).
3793 Fortunately, Exim does not actually use EX_OK. The code used to try to
3794 preserve the sysexits.h value, by assuming that macro definitions were
3795 scanned for macro replacements. I have been disabused of this notion,
3796 so now the code just undefines EX_OK before #including unistd.h.
3798 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3799 in the smtp transport. When a block could not be written in a single
3800 write() function, the timeout was being re-applied to each part-write.
3801 This seems wrong - if the receiver was accepting one byte at a time it
3802 would take for ever. The timeout is now adjusted when this happens. It
3803 doesn't have to be particularly precise.
3805 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3806 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3808 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3809 without a domain if the check to <postmaster@domain> fails.
3811 SC/01 Eximstats: added -xls and the ability to specify output files
3812 (patch written by Frank Heydlauf).
3814 SC/02 Eximstats: use FileHandles for outputting results.
3816 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3818 SC/04 Eximstats: fixed display of large numbers with -nvr option
3820 SC/05 Eximstats: fixed merging of reports with empty tables.
3822 SC/06 Eximstats: added the -include_original_destination flag
3824 SC/07 Eximstats: removed tabs and trailing whitespace.
3826 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3828 TK/06 MBOX spool code: Add real "From " MBOX separator line
3829 so the .eml file is really in mbox format (even though
3830 most programs do not really care). Patch from Alex Miller.
3832 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3833 The latter is generated from $received_to and is only set if the
3834 message has one envelope recipient. SA can use these headers,
3835 obviously out-of-the-box. Patch from Alex Miller.
3837 PH/08 The ${def test on a variable was returning false if the variable's
3838 value was "0", contrary to what the specification has always said!
3839 The result should be true unless the variable is empty.
3841 PH/09 The syntax error of a character other than { following "${if
3842 def:variable_name" (after optional whitespace) was not being diagnosed.
3843 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3844 accidental colon was present, for example, could give incorrect results.
3846 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3847 result is used (not including appendfile, where other changes are about
3850 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3851 This involved changing a lot of size variables from int to off_t. It
3852 should work with maildirs and everything.
3854 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3855 spamd dying while we are connected to it.
3857 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3860 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3861 function for BDB 4.3.
3863 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3865 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3866 the sources for PCRE have been re-arranged and more files are now
3869 PH/15 The code I had for printing potentially long long variables in PH/11
3870 above was not the best (it lost precision). The length of off_t variables
3871 is now inspected at build time, and an appropriate printing format (%ld
3872 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3873 to be "long long int" or "long int". This is needed for the internal
3874 formatting function string_vformat().
3876 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3877 the configuration file to be ":syslog", then the script "guesses" where
3878 the logs files are, rather than using the compiled in default. In our
3879 case the guess is not the same as the compiled default, so the script
3880 suddenly stopped working when I started to use syslog. The patch checks
3881 to see if log_file_path is "". If so, it attempts to read it from exim
3882 with no configuration file to get the compiled in version, before it
3883 falls back to the previous guessing code."
3885 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3886 implementing BATV in an Exim configuration. See NewStuff for the gory
3889 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3890 Makefile that are specific to HP-UX.
3892 PH/18 If the "use_postmaster" option was set for a recipient callout together
3893 with the "random" option, the postmaster address was used as the MAIL
3894 FROM address for the random test, but not for the subsequent recipient
3895 test. It is now used for both.
3897 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3898 patch removes a few documentation additions to RFC 3028, because the
3899 latest draft now contains them. It adds the new en;ascii-case comparator
3900 and a new error check for 8bit text in MIME parts. Comparator and
3901 require names are now matched exactly. I enabled the subaddress
3902 extension, but it is not well tested yet (read: it works for me)."
3904 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3905 rework some of the code of TK/09 above to avoid the hardwired use of
3906 "%lld" and "long long". Replaced the call to snprintf() with a call to
3909 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3910 records point to non-existent hosts", "retry timeout exceeded", and
3911 "retry time not reached for any host after a long failure period".
3913 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3914 experimental DomainKeys support:
3916 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3917 (2) On an error such as an illegally used "control", the wrong name for
3918 the control was given.
3920 These problems did NOT occur unless DomainKeys support was compiled.
3922 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3924 PH/24 Added ${if match_ip condition.
3926 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3927 databases so that it will be absolutely obvious if a crash occurs in the
3928 DB library. This is a regular occurrence (often caused by mis-matched
3931 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3932 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3933 picked up on a user's system that detects such things. There doesn't seem
3934 to be a gcc warning option for this - only an attribute that has to be
3935 put on the function's prototype. It seems that in Fedora Core 4 they have
3936 set this on a number of new functions. No doubt there will be more in due
3939 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3940 specify a transport. However, if an address that was verified by such a
3941 router was the subject of a callout, Exim crashed because it tried to
3942 read the rcpt_include_affixes from the non-existent transport. Now it
3943 just assumes that the setting of that option is false. This bug was
3944 introduced by 4.51/PH/31.
3946 PH/28 Changed -d+all to exclude +memory, because that information is very
3947 rarely of interest, but it makes the output a lot bigger. People tend to
3948 do -d+all out of habit.
3950 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3951 code in os-type was giving problems when libc.so lives in lib64, like on
3954 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3955 aren't the modern standard, and it seems that some systems' include files
3956 don't always have them. Exim was already checking for some of the newer
3957 ones like T_AAAA, and defining it itself. I've added checks for all the
3958 record types that Exim uses.
3960 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3961 not automatically generating a new one, as it is supposed to. This
3962 prevented TLS from working. If the file did exist, but contained invalid
3963 data, a new version was generated, as expected. It was only the case of a
3964 non-existent file that was broken.
3966 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3967 with a change in libdomainkeys > 0.64.
3969 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3970 from DNS. If the selector record carries the flag, it now has
3971 precedence over the domain-wide flag.
3973 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3975 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3976 the use of a port name that isn't defined in /etc/services) occurred, the
3977 message was deferred as in a normal delivery, and thus remained on the
3978 spool, instead of being failed because of the mua_wrapper setting. This
3979 is now fixed, and I tidied up some of the mua_wrapper messages at the
3982 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3983 the messages in a hash of arrays rather than using individual hashes.
3984 This is a bit cleaner and results in dramatic memory savings, albeit
3985 at a slight CPU cost.
3987 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3988 as requested by Marc Sherman.
3990 SC/10 Eximstats: added histograms for user specified patterns as requested
3993 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3995 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3996 fopen() in the content-scanning modules that did not already have it.
4002 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4003 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4005 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4007 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4009 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4010 to test Sieve filters that use "vacation".
4012 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4013 that changes the way the GnuTLS parameters are stored in the cache file.
4014 The new format can be generated externally. For backward compatibility,
4015 if the data in the cache doesn't make sense, Exim assumes it has read an
4016 old-format file, and it generates new data and writes a new file. This
4017 means that you can't go back to an older release without removing the
4020 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4021 work if there are any delivery delays because "one_time" forces the
4022 parent to be marked "delivered", so its unseen clone is never tried
4023 again. For this reason, Exim now forbids the simultaneous setting of
4026 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4027 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4028 entry if you want to know the details. The fix, however, neglected to
4029 consider the case when local delivery batching is involved. The test for
4030 "previously delivered" was not happening when checking to see if an
4031 address could be batched with a previous (undelivered) one; under
4032 certain circumstances this could lead to multiple deliveries to the same
4035 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4036 in its include files, and this causes problems building Exim.
4038 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4039 header_syntax) but Exim was just ignoring anything given after a slash.
4040 In particular, this caused confusion with an attempt to use "verify =
4041 reverse_host_lookup/defer_ok". An error is now given when options are
4042 supplied for verify items that do not have them. (Maybe reverse_host_
4043 lookup should have a defer_ok option, but that's a different point.)
4045 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4046 defined by RFC 821) to 2048, because there were problems with some AUTH
4047 commands, and RFC 1869 says the size should be increased for extended
4048 SMTP commands that take arguments.
4050 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4053 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4054 "unknown" error; now it says that the functionality isn't in the binary.
4056 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4057 an address' error message when a string expansion fails (syntax or
4058 whatever). Otherwise the password may appear in the log. Following change
4059 PH/42 below, there is no longer a chance of it appearing in a bounce
4062 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4064 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4065 of its addresses was ever tried. (Bugzilla bug #2.)
4067 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4068 the result incorrectly in the debug output. (It correctly added a newline
4069 to what was transported.)
4071 TF/01 Added $received_time.
4073 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4074 commented out examples of how to interface to a virus scanner and to
4075 SpamAssassin. Also added commented examples of av_scanner and
4076 spamd_address settings.
4078 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4079 and controls are allowed in which ACLs. There were a couple of minor
4080 errors. Some of the entries in the conditions table (which is a table of
4081 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4082 negation of where the condition IS allowed.
4084 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4086 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4087 header file does not have a version number, so I've had to invent a new
4088 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4089 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4090 radiusclient), but it was contributed by a Radius user.
4092 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4093 files or directories, for OpenSSL.
4095 PH/19 When an Exim process that is running as root has to create an Exim log
4096 file, it does so in a subprocess that runs as exim:exim so as to get the
4097 ownership right at creation (otherwise, other Exim processes might see
4098 the file with the wrong ownership). There was no test for failure of this
4099 fork() call, which would lead to the process getting stuck as it waited
4100 for a non-existent subprocess. Forks do occasionally fail when resources
4101 run out. I reviewed all the other calls to fork(); they all seem to check
4104 PH/20 When checking for unexpected SMTP input at connect time (before writing
4105 the banner), Exim was not dealing correctly with a non-positive return
4106 from the read() function. If the client had disconnected by this time,
4107 the result was a log entry for a synchronization error with an empty
4108 string after "input=" when read() returned zero. If read() returned -1
4109 (an event I could not check), uninitialized data bytes were printed.
4110 There were reports of junk text (parts of files, etc) appearing after
4113 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4115 PH/22 Added support for macro redefinition, and (re)definition in between
4116 driver and ACL definitions.
4118 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4119 forgetting to use the resulting value; it was using the unexpanded value.
4121 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4122 hadn't been configured. The fix is from Juergen Kreileder, who
4123 understands it better than I do:
4125 "Here's what I see happening with three configured cyrus_sasl
4126 authenticators configured (plain, login, cram-md5):
4128 On startup auth_cyrus_sasl_init() gets called for each of these.
4129 This means three calls to sasl_listmech() without a specified mech_list.
4130 => SASL tests which mechs of all available mechs actually work
4131 => three warnings about OTP not working
4132 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4134 With the patch, sasl_listmech() also gets called three times. But now
4135 SASL's mech_list option is set to the server_mech specified in the the
4136 authenticator. Or in other words, the answer from sasl_listmech()
4137 gets limited to just the mech you're testing for (which is different
4139 => the return list contains just 'plain' or 'login', 'cram-md5' or
4140 nothing depending on the value of ob->server_mech.
4142 I've just tested the patch: Authentication still works fine,
4143 unavailable mechs specified in the exim configuration are still
4144 caught, and the auth.log warnings about OTP are gone."
4146 PH/25 When debugging is enabled, the contents of the command line are added
4147 to the debugging output, even when log_selector=+arguments is not
4150 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4151 answer is "GNU", and only if the return is "GNU/something" is the answer
4154 PH/27 $acl_verify_message is now set immediately after the failure of a
4155 verification in an ACL, and so is available in subsequent modifiers. In
4156 particular, the message can be preserved by coding like this:
4158 warn !verify = sender
4159 set acl_m0 = $acl_verify_message
4161 Previously, $acl_verify_message was set only while expanding "message"
4162 and "log_message" when a very denied access.
4164 PH/28 Modified OS/os.c-Linux with
4166 -#ifndef OS_LOAD_AVERAGE
4167 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4169 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4170 nomenclature these days.)
4172 PH/29 Installed patch from the Sieve maintainer that adds the options
4173 sieve_useraddress and sieve_subaddress to the redirect router.
4175 PH/30 In these circumstances:
4176 . Two addresses routed to the same list of hosts;
4177 . First host does not offer TLS;
4178 . First host accepts first address;
4179 . First host gives temporary error to second address;
4180 . Second host offers TLS and a TLS session is established;
4181 . Second host accepts second address.
4182 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4183 and peerdn, if requested) that were in fact used only for the second
4186 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4187 attention to any local part prefix or suffix that was matched by the
4188 router that accepted the address. It now behaves in the same way as it
4189 does for delivery: the affixes are removed from the local part unless
4190 rcpt_include_affixes is set on the transport.
4192 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4193 timeout during the DATA phase of an incoming message.
4195 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4196 applied a patch sanctioned by the Sieve maintainer.
4198 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4199 the uid or gid is negative. A case of a negative gid caused this to be
4200 noticed. The fix allows for either to be negative.
4202 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4203 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4206 PH/36 The change PH/12 above was broken. Fixed it.
4208 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4209 the grounds that routing the same address twice would always produce the
4210 same answer. This might have been true once, but it is certainly no
4211 longer true now. Routing a child address may depend on the previous
4212 routing that produced that child. Some complicated redirection strategies
4213 went wrong when messages had multiple recipients, and made Exim's
4214 behaviour dependent on the order in which the addresses were given.
4216 I have moved the duplicate checking until after the routing is complete.
4217 Exim scans the addresses that are assigned to local and remote
4218 transports, and removes any duplicates. This means that more work will be
4219 done, as duplicates will always all be routed, but duplicates are
4220 presumably rare, so I don't expect this is of any significance.
4222 For deliveries to pipes, files, and autoreplies, the duplicate checking
4223 still happens during the routing process, since they are not going to be
4226 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4227 It corrects a timeout issue with spamd. This is Ian's comment: "The
4228 background is that sometimes spamd either never reads data from a
4229 connection it has accepted, or it never writes response data. The exiscan
4230 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4231 blindly assumes that writes won't block so it may never time out."
4233 PH/39 Allow G after quota size as well as K and M.
4235 PH/40 The value set for $authenticated_id in an authenticator may not contain
4236 binary zeroes or newlines because the value is written to log lines and
4237 to spool files. There was no check on this. Now the value is run through
4238 the string_printing() function so that such characters are converted to
4239 printable escape sequences.
4241 PH/41 $message_linecount is a new variable that contains the total number of
4242 lines in the message. Compare $body_linecount, which is the count for the
4245 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4246 bounce and delay warning messages, except in certain special cases, which
4249 (a) An SMTP error message from a remote host;
4250 (b) A message specified in a :fail: redirection;
4251 (c) A message specified in a "fail" command in a system filter;
4252 (d) A message specified in a FAIL return from the queryprogram router;
4253 (e) A message specified by the cannot_route_message router option.
4255 In these cases only, Exim does include the error details in bounce and
4256 warning messages. There are also a few cases where bland messages such
4257 as "unrouteable address" or "local delivery error" are given.
4259 PH/43 $value is now also set for the "else" part of a ${run expansion.
4261 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4262 being worked on, but at least Exim now implements the latest version to
4265 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4266 process to complete was treated as a delivery failure, a timeout while
4267 writing the message to the pipe was logged, but erroneously treated as a
4268 successful delivery. Such timeouts include transport filter timeouts. For
4269 consistency with the overall process timeout, these timeouts are now
4270 treated as errors, giving rise to delivery failures by default. However,
4271 there is now a new Boolean option for the pipe transport called
4272 timeout_defer, which, if set TRUE, converts the failures into defers for
4273 both kinds of timeout. A transport filter timeout is now identified in
4276 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4277 systems where "make" and "gmake" are different, calling "gmake" at top
4278 level broke things. I've arranged for the value of $(MAKE) to be passed
4279 from the Makefile to this script so that it can call the same version of
4283 A note about Exim versions 4.44 and 4.50
4284 ----------------------------------------
4286 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4287 changes of various kinds. As a consequence, a big documentation update was
4288 needed. This delayed the release for rather longer than seemed good, especially
4289 in the light of a couple of (minor) security issues. Therefore, the changes
4290 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4291 release. So 4.44 and 4.50 are in effect two different branches that both start
4294 I have left the 4.50 change log unchanged; it contains all the changes since
4295 4.43. The change log for 4.44 is below; many of its items are identical to
4296 those for 4.50. This seems to be the most sensible way to preserve the
4297 historical information.
4303 1. Minor wording change to the doc/README.SIEVE file.
4305 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4306 computation of the current number of files was incorrect.
4308 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4309 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4312 4. Give more explanation in the error message when the command for a transport
4313 filter fails to execute.
4315 5. There are several places where Exim runs a non-Exim command in a
4316 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4317 was being done only for the command run by the queryprogram router. It is
4318 now done for all such subprocesses. The other cases are: ${run, transport
4319 filters, and the commands run by the lmtp and pipe transports.
4321 6. Added CONFIGURE_GROUP build-time option.
4323 7. Some older OS have a limit of 256 on the maximum number of file
4324 descriptors. Exim was using setrlimit() to set 1000 as a large value
4325 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4326 systems. I've change it so that if it can't get 1000, it tries for 256.
4328 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4329 was an oversight, and furthermore, ever since the addition of extra
4330 controls (e.g. 4.43/32), the checks on when to allow different forms of
4331 "control" were broken. There should now be diagnostics for all cases when a
4332 control that does not make sense is encountered.
4334 9. Added the /retain_sender option to "control=submission".
4336 10. $recipients is now available in the predata ACL (oversight).
4338 11. Tidy the search cache before the fork to do a delivery from a message
4339 received from the command line. Otherwise the child will trigger a lookup
4340 failure and thereby defer the delivery if it tries to use (for example) a
4341 cached ldap connection that the parent has called unbind on.
4343 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4344 of $address_data from the recipient verification was clobbered by the
4345 sender verification.
4347 13. The value of address_data from a sender verification is now available in
4348 $sender_address_data in subsequent conditions in the ACL statement.
4350 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4352 15. Added a new option "connect=<time>" to callout options, to set a different
4355 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4356 was its contents. (It was OK if the option was not defined at all.)
4358 17. A "Completed" log line is now written for messages that are removed from
4359 the spool by the -Mrm option.
4361 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4362 information about exactly what failed.
4364 19. Added -dd to debug only the daemon process.
4366 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4367 handles timeouts, both on the server side and network timeouts. Renamed the
4368 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4370 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4371 It is now set to "smtps".
4373 22. $host_address is now set to the target address during the checking of
4374 ignore_target_hosts.
4376 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4377 being passed; this would have caused $sender_host_name to have been used if
4378 matching the list had actually called for a host name (not very likely,
4379 since this list is usually IP addresses). A host name is now passed as
4382 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4383 code that specifies a non-restarting handler (typically sigaction() in
4384 modern systems) in an attempt to fix a rare and obscure crash bug.
4386 25. Narrowed the window for a race in the daemon that could cause it to ignore
4387 SIGCHLD signals. This is not a major problem, because they are used only to
4388 wake it up if nothing else does.
4390 26. A malformed maildirsize file could cause Exim to calculate negative values
4391 for the mailbox size or file count. Odd effects could occur as a result.
4392 The maildirsize information is now recalculated if the size or filecount
4395 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4396 support for a long time. Removed HAVE_SYS_VFS_H.
4398 28. Installed the latest version of exipick from John Jetmore.
4400 29. In an address list, if the pattern was not a regular expression, an empty
4401 subject address (from a bounce message) matched only if the pattern was an
4402 empty string. Non-empty patterns were not even tested. This was the wrong
4403 because it is perfectly reasonable to use an empty address as part of a
4404 database query. An empty address is now tested by patterns that are
4405 lookups. However, all the other forms of pattern expect the subject to
4406 contain a local part and a domain, and therefore, for them, an empty
4407 address still always fails if the pattern is not itself empty.
4409 30. Exim went into a mad DNS loop when attempting to do a callout where the
4410 host was specified on an smtp transport, and looking it up yielded more
4411 than one IP address.
4413 31. Re-factored the code for checking spool and log partition space into a
4414 function that finds that data and another that does the check. The former
4415 is then used to implement four new variables: $spool_space, $log_space,
4416 $spool_inodes, and $log_inodes.
4418 32. The RFC2047 encoding function was originally intended for short strings
4419 such as real names; it was not keeping to the 75-character limit for
4420 encoded words that the RFC imposes. It now respects the limit, and
4421 generates multiple encoded words if necessary. To be on the safe side, I
4422 have increased the buffer size for the ${rfc2047: expansion operator from
4425 33. It is now permitted to omit both strings after an "if" condition; if the
4426 condition is true, the result is "true". As before, when the second string
4427 is omitted, a false condition yields an empty string. This makes it less
4428 cumbersome to write custom ACL and router conditions.
4430 34. Failure to deliver a bounce message always caused it to be frozen, even if
4431 there was an errors_to setting on the router. The errors_to setting is now
4434 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4435 canonical form (fully expanded) before being placed in
4436 $sender_host_address.
4438 36. The table in the code that translates DNS record types into text (T_A to
4439 "A" for instance) was missing entries for NS and CNAME. It is just possible
4440 that this could have caused confusion if both these types were looked up
4441 for the same domain, because the text type is used as part of Exim's
4442 per-process caching. But the chance of anyone hitting this buglet seems
4445 37. The dnsdb lookup has been extended in a number of ways.
4447 (1) There is a new type, "zns", which walks up the domain tree until it
4448 finds some nameserver records. It should be used with care.
4450 (2) There is a new type, "mxh", which is like "mx" except that it returns
4451 just the host names, not the priorities.
4453 (3) It is now possible to give a list of domains (or IP addresses) to be
4454 looked up. The behaviour when one of the lookups defers can be
4455 controlled by a keyword.
4457 (4) It is now possible to specify the separator character for use when
4458 multiple records are returned.
4460 38. The dnslists ACL condition has been extended: it is now possible to supply
4461 a list of IP addresses and/or domains to be looked up in a particular DNS
4464 39. Added log_selector=+queue_time_overall.
4466 40. When running the queue in the test harness, wait just a tad after forking a
4467 delivery process, to get repeatability of debugging output.
4469 41. Include certificate and key file names in error message when GnuTLS fails
4470 to set them up, because the GnuTLS error message doesn't include the name
4471 of the failing file when there is a problem reading it.
4473 42. Allow both -bf and -bF in the same test run.
4475 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4477 44. The "Exiscan patch" is now merged into the mainline Exim source.
4479 45. Sometimes the final signoff response after QUIT could fail to get
4480 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4481 < 0 before doing a fflush(). This bug looks as though it goes back to the
4482 introduction of TLS in release 3.20, but "sometimes" must have been rare
4483 because the tests only now provoked it.
4485 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4486 (this can affect the format of dates).
4488 47. exim_tidydb, when checking for the continued existence of a message for
4489 which it has found a message-specific retry record, was not finding
4490 messages that were in split spool directories. Consequently, it was
4491 deleting retry records that should have stayed in existence.
4493 48. Steve fixed some bugs in eximstats.
4495 49. The SPA authentication driver was not abandoning authentication and moving
4496 on to the next authenticator when an expansion was forced to fail,
4497 contradicting the general specification for all authenticators. Instead it
4498 was generating a temporary error. It now behaves as specified.
4500 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4501 (the order specifies the preference for clients). The order is now AES256,
4502 AES128, 3DES, ARCFOUR128.
4504 51. Small patch to Sieve code - explicitly set From: when generating an
4507 52. Exim crashed if a remote delivery caused a very long error message to be
4508 recorded - for instance if somebody sent an entire SpamAssassin report back
4509 as a large number of 550 error lines. This bug was coincidentally fixed by
4510 increasing the size of one of Exim's internal buffers (big_buffer) that
4511 happened as part of the Exiscan merge. However, to be on the safe side, I
4512 have made the code more robust (and fixed the comments that describe what
4515 53. Now that there can be additional text after "Completed" in log lines (if
4516 the queue_time_overall log selector is set), a one-byte patch to exigrep
4517 was needed to allow it to recognize "Completed" as not the last thing in
4520 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4521 patch that reportedly fixes this has been added. I am not expert enough to
4522 create a test for it. This is what the patch creator wrote:
4524 "I found a little strange behaviour of ldap code when working with
4525 Windows 2003 AD Domain, where users was placed in more than one
4526 Organization Units. When I tried to give exim partial DN, the exit code
4527 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4528 But simultaneously result of request was absolutely normal ldap result,
4529 so I produce this patch..."
4531 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4532 REFERENCE, so I have modified the code to exclude the patch when that macro
4535 55. Some experimental protocols are using DNS PTR records for new purposes. The
4536 keys for these records are domain names, not reversed IP addresses. The
4537 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4538 leaves it alone. Component reversal etc. now happens only for IP addresses.
4541 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4543 57. Double the size of the debug message buffer (to 2048) so that more of very
4544 long debug lines gets shown.
4546 58. The exicyclog utility now does better if the number of log files to keep
4547 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4549 59. Two changes related to the smtp_active_hostname option:
4551 (1) $smtp_active_hostname is now available as a variable.
4552 (2) The default for smtp_banner uses $smtp_active_hostname instead
4553 of $primary_hostname.
4555 60. The host_aton() function is supposed to be passed a string that is known
4556 to be a valid IP address. However, in the case of IPv6 addresses, it was
4557 not checking this. This is a hostage to fortune. Exim now panics and dies
4558 if the condition is not met. A case was found where this could be provoked
4559 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4560 components; fortuitously, this particular loophole had already been fixed
4561 by change 4.50/55 above.
4563 If there are any other similar loopholes, the new check in host_aton()
4564 itself should stop them being exploited. The report I received stated that
4565 data on the command line could provoke the exploit when Exim was running as
4566 exim, but did not say which command line option was involved. All I could
4567 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4568 running as the user.
4571 61. There was a buffer overflow vulnerability in the SPA authentication code
4572 (which came originally from the Samba project). I have added a test to the
4573 spa_base64_to_bits() function which I hope fixes it.
4576 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4577 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4579 63. The daemon start-up calls getloadavg() while still root for those OS that
4580 need the first call to be done as root, but it missed one case: when
4581 deliver_queue_load_max is set with deliver_drop_privilege. This is
4582 necessary for the benefit of the queue runner, because there is no re-exec
4583 when deliver_drop_privilege is set.
4585 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4586 This has been fixed.
4588 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4589 was in use, was not putting the data itself into the right store pool;
4590 consequently, it could be overwritten for a subsequent message in the same
4591 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4594 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4596 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4597 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4598 the function were treating the return as a boolean value, which happened to
4599 work because 0=false and not-0=true, but is not correct code.
4601 68. The host_aton() function was not handling scoped IPv6 addresses (those
4602 with, for example, "%eth0" on the end) correctly.
4604 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4605 negated items (that is, ~something) in unsigned ints. Some compilers
4606 apparently mutter when there is no cast.
4608 70. If an address verification called from an ACL failed, and did not produce a
4609 user-specific message (i.e. there was only a "system" message), nothing was
4610 put in $acl_verify_message. In this situation, it now puts the system
4613 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4614 session; change 4.31/43 added the unwanted input to the log line - except
4615 that it did not do this in the start of session case. It now does.
4617 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4618 This is wrong and can cause the other end to generate a synchronization
4619 error if it is another Exim or anything else that does the synchronization
4620 check. A QUIT command is no longer sent after a timeout.
4622 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4623 during host lookups.
4625 74. The defer_ok option of callout verification was not working if it was used
4626 when verifying addresses in header lines, that is, for this case:
4628 verify = header_sender/callout=defer_ok
4630 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4631 those file descriptors could be used for SMTP connections. If anything
4632 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4633 could be sent to the SMTP client, causing chaos. The daemon now opens
4634 stdin, stdout, and stderr to /dev/null when it puts itself into the
4637 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4638 log by default. The user can override this with suitable Perl magic.
4640 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4641 the log message when discard triggers, was not working for the DATA ACL or
4642 for the non-SMTP ACL.
4644 78. Error message wording change in sieve.c.
4646 79. If smtp_accept_max_per_host was set, the number of connections could be
4647 restricted to fewer than expected, because the daemon was trying to set up
4648 a new connection before checking whether the processes handling previous
4649 connections had finished. The check for completed processes is now done
4650 earlier. On busy systems, this bug wouldn't be noticed because something
4651 else would have woken the daemon, and it would have reaped the completed
4654 80. If a message was submitted locally by a user whose login name contained one
4655 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4656 It caused a spool format error. I have fixed the spool reading code. A
4657 related problem was that the "from" clause in the Received: line became
4658 illegal because of the space(s). It is now covered by ${quote_local_part.
4660 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4663 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4665 83. Patch from Sieve maintainer to fix unterminated string problem in
4666 "vacation" handling.
4668 84. Some minor changes to the Linux configuration files to help with other
4669 OS variants using glibc.
4671 85. One more patch for Sieve to update vacation handling to latest spec.
4674 ----------------------------------------------------
4675 See the note above about the 4.44 and 4.50 releases.
4676 ----------------------------------------------------
4682 1. Change 4.43/35 introduced a bug that caused file counts to be
4683 incorrectly computed when quota_filecount was set in an appendfile
4686 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4687 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4690 3. Give more explanation in the error message when the command for a transport
4691 filter fails to execute.
4693 4. There are several places where Exim runs a non-Exim command in a
4694 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4695 was being done only for the command run by the queryprogram router. It is
4696 now done for all such subprocesses. The other cases are: ${run, transport
4697 filters, and the commands run by the lmtp and pipe transports.
4699 5. Some older OS have a limit of 256 on the maximum number of file
4700 descriptors. Exim was using setrlimit() to set 1000 as a large value
4701 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4702 systems. I've change it so that if it can't get 1000, it tries for 256.
4704 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4705 was an oversight, and furthermore, ever since the addition of extra
4706 controls (e.g. 4.43/32), the checks on when to allow different forms of
4707 "control" were broken. There should now be diagnostics for all cases when a
4708 control that does not make sense is encountered.
4710 7. $recipients is now available in the predata ACL (oversight).
4712 8. Tidy the search cache before the fork to do a delivery from a message
4713 received from the command line. Otherwise the child will trigger a lookup
4714 failure and thereby defer the delivery if it tries to use (for example) a
4715 cached ldap connection that the parent has called unbind on.
4717 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4718 of $address_data from the recipient verification was clobbered by the
4719 sender verification.
4721 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4722 was its contents. (It was OK if the option was not defined at all.)
4724 11. A "Completed" log line is now written for messages that are removed from
4725 the spool by the -Mrm option.
4727 12. $host_address is now set to the target address during the checking of
4728 ignore_target_hosts.
4730 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4731 being passed; this would have caused $sender_host_name to have been used if
4732 matching the list had actually called for a host name (not very likely,
4733 since this list is usually IP addresses). A host name is now passed as
4736 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4737 code that specifies a non-restarting handler (typically sigaction() in
4738 modern systems) in an attempt to fix a rare and obscure crash bug.
4740 15. Narrowed the window for a race in the daemon that could cause it to ignore
4741 SIGCHLD signals. This is not a major problem, because they are used only to
4742 wake it up if nothing else does.
4744 16. A malformed maildirsize file could cause Exim to calculate negative values
4745 for the mailbox size or file count. Odd effects could occur as a result.
4746 The maildirsize information is now recalculated if the size or filecount
4749 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4750 support for a long time. Removed HAVE_SYS_VFS_H.
4752 18. Updated exipick to current release from John Jetmore.
4754 19. Allow an empty sender to be matched against a lookup in an address list.
4755 Previously the only cases considered were a regular expression, or an
4758 20. Exim went into a mad DNS lookup loop when doing a callout where the
4759 host was specified on the transport, if the DNS lookup yielded more than
4762 21. The RFC2047 encoding function was originally intended for short strings
4763 such as real names; it was not keeping to the 75-character limit for
4764 encoded words that the RFC imposes. It now respects the limit, and
4765 generates multiple encoded words if necessary. To be on the safe side, I
4766 have increased the buffer size for the ${rfc2047: expansion operator from
4769 22. Failure to deliver a bounce message always caused it to be frozen, even if
4770 there was an errors_to setting on the router. The errors_to setting is now
4773 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4774 canonical form (fully expanded) before being placed in
4775 $sender_host_address.
4777 24. Updated eximstats to version 1.33
4779 25. Include certificate and key file names in error message when GnuTLS fails
4780 to set them up, because the GnuTLS error message doesn't include the name
4781 of the failing file when there is a problem reading it.
4783 26. Expand error message when OpenSSL has problems setting up cert/key files.
4786 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4787 (this can affect the format of dates).
4789 28. exim_tidydb, when checking for the continued existence of a message for
4790 which it has found a message-specific retry record, was not finding
4791 messages that were in split spool directories. Consequently, it was
4792 deleting retry records that should have stayed in existence.
4794 29. eximstats updated to version 1.35
4795 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4796 1.35 - bugfix such that pie charts by volume are generated correctly
4798 30. The SPA authentication driver was not abandoning authentication and moving
4799 on to the next authenticator when an expansion was forced to fail,
4800 contradicting the general specification for all authenticators. Instead it
4801 was generating a temporary error. It now behaves as specified.
4803 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4804 (the order specifies the preference for clients). The order is now AES256,
4805 AES128, 3DES, ARCFOUR128.
4807 31. Small patch to Sieve code - explicitly set From: when generating an
4810 32. Exim crashed if a remote delivery caused a very long error message to be
4811 recorded - for instance if somebody sent an entire SpamAssassin report back
4812 as a large number of 550 error lines. This bug was coincidentally fixed by
4813 increasing the size of one of Exim's internal buffers (big_buffer) that
4814 happened as part of the Exiscan merge. However, to be on the safe side, I
4815 have made the code more robust (and fixed the comments that describe what
4818 33. Some experimental protocols are using DNS PTR records for new purposes. The
4819 keys for these records are domain names, not reversed IP addresses. The
4820 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4821 leaves it alone. Component reversal etc. now happens only for IP addresses.
4824 34. The host_aton() function is supposed to be passed a string that is known
4825 to be a valid IP address. However, in the case of IPv6 addresses, it was
4826 not checking this. This is a hostage to fortune. Exim now panics and dies
4827 if the condition is not met. A case was found where this could be provoked
4828 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4829 components; fortuitously, this particular loophole had already been fixed
4830 by change 4.50/55 or 4.44/33 above.
4832 If there are any other similar loopholes, the new check in host_aton()
4833 itself should stop them being exploited. The report I received stated that
4834 data on the command line could provoke the exploit when Exim was running as
4835 exim, but did not say which command line option was involved. All I could
4836 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4837 running as the user.
4840 35. There was a buffer overflow vulnerability in the SPA authentication code
4841 (which came originally from the Samba project). I have added a test to the
4842 spa_base64_to_bits() function which I hope fixes it.
4845 36. The daemon start-up calls getloadavg() while still root for those OS that
4846 need the first call to be done as root, but it missed one case: when
4847 deliver_queue_load_max is set with deliver_drop_privilege. This is
4848 necessary for the benefit of the queue runner, because there is no re-exec
4849 when deliver_drop_privilege is set.
4851 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4852 was in use, was not putting the data itself into the right store pool;
4853 consequently, it could be overwritten for a subsequent message in the same
4854 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4857 38. Sometimes the final signoff response after QUIT could fail to get
4858 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4859 < 0 before doing a fflush(). This bug looks as though it goes back to the
4860 introduction of TLS in release 3.20, but "sometimes" must have been rare
4861 because the tests only now provoked it.
4867 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4868 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4869 Now it returns an integer. A number of calls were still expecting a T/F
4870 return. Fortuitously, in all cases, the tests worked in OK situations,
4871 which is the norm. However, things would have gone wrong on any write
4872 failures on the smtp file descriptor. This function is used when sending
4873 messages over SMTP and also when doing verify callouts.
4875 2. When Exim is called to do synchronous delivery of a locally submitted
4876 message (the -odf or -odi options), it no longer closes stderr before doing
4879 3. Implemented the mua_wrapper option.
4881 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4883 5. Implemented the functions header_remove(), header_testname(),
4884 header_add_at_position(), and receive_remove_recipient(), and exported them
4887 6. If an ACL "warn" statement specified the addition of headers, Exim already
4888 inserted X-ACL-Warn: at the start if there was no header name. However, it
4889 was not making this test for the second and subsequent header lines if
4890 there were newlines in the string. This meant that an invalid header could
4891 be inserted if Exim was badly configured.
4893 7. Allow an ACL "warn" statement to add header lines at the start or after all
4894 the Received: headers, as well as at the end.
4896 8. Added the rcpt_4xx retry error code.
4898 9. Added postmaster_mailfrom=xxx to callout verification option.
4900 10. Added mailfrom=xxxx to the callout verification option, for verify=
4903 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4904 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4906 12. Inserted some casts to stop certain compilers warning when using pointer
4907 differences as field lengths or precisions in printf-type calls (mostly
4908 affecting debugging statements).
4910 13. Added optional readline() support for -be (dynamically loaded).
4912 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4913 same clock tick as a message's arrival, so that its received time was the
4914 same as the "first fail" time on the retry record, and that message
4915 remained on the queue past the ultimate address timeout, every queue runner
4916 would try a delivery (because it was past the ultimate address timeout) but
4917 after another failure, the ultimate address timeout, which should have then
4918 bounced the address, did not kick in. This was a "< instead of <=" error;
4919 in most cases the first failure would have been in the next clock tick
4920 after the received time, and all would be well.
4922 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4923 being recognized when the domain list was tested by the match_domain
4924 condition in an expansion string.
4926 16. Added the ${str2b64: operator.
4928 17. Exim was always calling setrlimit() to set a large limit for the number of
4929 processes, without checking whether the existing limit was already
4930 adequate. (It did check for the limit on file descriptors.) Furthermore,
4931 errors from getrlimit() and setrlimit() were being ignored. Now they are
4932 logged to the main and panic logs, but Exim does carry on, to try to do its
4933 job under whatever limits there are.
4935 18. Imported PCRE 5.0.
4937 19. Trivial typo in log message " temporarily refused connection" (the leading
4940 20. If the log selector return_path_on_delivery was set and an address was
4941 redirected to /dev/null, the delivery process crashed because it assumed
4942 that a return path would always be set for a "successful" delivery. In this
4943 case, the whole delivery is bypassed as an optimization, and therefore no
4946 21. Internal re-arrangement: the function for sending a challenge and reading
4947 a response while authentication was assuming a zero-terminated challenge
4948 string. It's now changed to take a pointer and a length, to allow for
4949 binary data in such strings.
4951 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4953 23. Exim was not respecting finduser_retries when seeking the login of the
4954 uid under which it was called; it was always trying 10 times. (The default
4955 setting of finduser_retries is zero.) Also, it was sleeping after the final
4956 failure, which is pointless.
4958 24. Implemented tls_on_connect_ports.
4960 25. Implemented acl_smtp_predata.
4962 26. If the domain in control=submission is set empty, Exim assumes that the
4963 authenticated id is a complete email address when it generates From: or
4964 Sender: header lines.
4966 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4967 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4968 chown and chgrp in /bin and hostname in /usr/bin.
4970 28. Exim was keeping the "process log" file open after each use, just as it
4971 does for the main log. This opens the possibility of it remaining open for
4972 long periods when the USR1 signal hits a daemon. Occasional processlog
4973 errors were reported, that could have been caused by this. Anyway, it seems
4974 much more sensible not to leave this file open at all, so that is what now
4977 29. The long-running daemon process does not normally write to the log once it
4978 has entered its main loop, and it closes the log before doing so. This is
4979 so that log files can straightforwardly be renamed and moved. However,
4980 there are a couple of unusual error situations where the daemon does write
4981 log entries, and I had neglected to close the log afterwards.
4983 30. The text of an SMTP error response that was received during a remote
4984 delivery was being truncated at 512 bytes. This is too short for some of
4985 the long messages that one sometimes sees. I've increased the limit to
4988 31. It is now possible to make retry rules that apply only when a message has a
4989 specific sender, in particular, an empty sender.
4991 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4992 it possible to be selective about when SMTP synchronization is enforced.
4994 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4996 32. Implemented hosts_connection_nolog.
4998 33. Added an ACL for QUIT.
5000 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5003 35. Added mailbox_size and mailbox_filecount to appendfile.
5005 36. Added control = no_multiline_responses to ACLs.
5007 37. There was a bug in the logic of the code that waits for the clock to tick
5008 in the case where the clock went backwards by a substantial amount such
5009 that the microsecond fraction of "now" was more than the microsecond
5010 fraction of "then" (but the whole seconds number was less).
5012 38. Added support for the libradius Radius client library this is found on
5013 FreeBSD (previously only the radiusclient library was supported).
5019 1. When certain lookups returned multiple values in the form name=value, the
5020 quoting of the values was not always being done properly. Specifically:
5021 (a) If the value started with a double quote, but contained no whitespace,
5023 (b) If the value contained whitespace other than a space character (i.e.
5024 tabs or newlines or carriage returns) it was not quoted.
5025 This fix has been applied to the mysql and pgsql lookups by writing a
5026 separate quoting function and calling it from the lookup code. The fix
5027 should probably also be applied to nisplus, ibase and oracle lookups, but
5028 since I cannot test any of those, I have not disturbed their existing code.
5030 2. A hit in the callout cache for a specific address caused a log line with no
5031 reason for rejecting RCPT. Now it says "Previous (cached) callout
5032 verification failure".
5034 3. There was an off-by-one bug in the queryprogram router. An over-long
5035 return line was truncated at 256 instead of 255 characters, thereby
5036 overflowing its buffer with the terminating zero. As well as fixing this, I
5037 have increased the buffer size to 1024 (and made a note to document this).
5039 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5040 when Exim is waiting for an SMTP response from a remote server, Exim
5041 restarts its select() call on the socket, thereby resetting its timeout.
5042 This is not a problem when such interrupts are rare. Somebody set up a cron
5043 job to run exiwhat every 2 minutes, which is less than the normal select()
5044 timeout (5 or 10 minutes). This meant that the select() timeout never
5045 kicked in because it was always reset. I have fixed this by comparing the
5046 time when an interrupt arrives with the time at the start of the first call
5047 to select(). If more time than the timeout has elapsed, the interrupt is
5048 treated as a timeout.
5050 5. Some internal re-factoring in preparation for the addition of Sieve
5051 extensions (by MH). In particular, the "personal" test is moved to a
5052 separate function, and given an option for scanning Cc: and Bcc: (which is
5053 not set for Exim filters).
5055 6. When Exim created an email address using the login of the caller as the
5056 local part (e.g. when creating a From: or Sender: header line), it was not
5057 quoting the local part when it contained special characters such as @.
5059 7. Installed new OpenBSD configuration files.
5061 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5062 try to make them clearer.
5064 9. Callout options, other than the timeout value, were being ignored when
5065 verifying sender addresses in header lines. For example, when using
5067 verify = header_sender/callout=no_cache
5069 the cache was (incorrectly) being used.
5071 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5072 only the Cygwin environment.
5074 11. When return_path_on_delivery was set as a log selector, if different remote
5075 addresses in the same message used different return paths and parallel
5076 remote delivery occurred, the wrong values would sometimes be logged.
5077 (Whenever a remote delivery process finished, the return path value from
5078 the most recently started remote delivery process was logged.)
5080 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5081 lines when AUTH and/or TLS are in use. This is the "received protocol"
5082 field. Exim used to use "asmtp" for authenticated SMTP, without any
5083 indication (in the protocol name) for TLS use. Now it follows the RFC and
5084 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5085 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5086 names appear in log lines as well as in Received: header lines.
5088 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5089 extensions, and comparison tests, and to fix some bugs.
5091 14. Changes to the "personal" filter test:
5093 (1) The test was buggy in that it was just doing the equivalent of
5094 "contains" tests on header lines. For example, if a user's address was
5095 anne@some.where, the "personal" test would incorrectly be true for
5097 To: susanne@some.where
5099 This test is now done by extracting each address from the header in turn,
5100 and checking the entire address. Other tests that are part of "personal"
5101 are now done using regular expressions (for example, to check local parts
5102 of addresses in From: header lines).
5104 (2) The list of non-personal local parts in From: addresses has been
5105 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5106 taken from the Sieve specification recommendations.
5108 (3) If the message contains any header line starting with "List-" it is
5109 treated as non-personal.
5111 (4) The test for "circular" in the Subject: header line has been removed
5112 because it now seems ill-conceived.
5114 15. Minor typos in src/EDITME comments corrected.
5116 16. Installed latest exipick from John Jetmore.
5118 17. If headers_add on a router specified a text string that was too long for
5119 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5120 of string_sprintf() is now avoided.
5122 18. $message_body_size was not set (it was always zero) when running the DATA
5123 ACL and the local_scan() function.
5125 19. For the "mail" command in an Exim filter, no default was being set for
5126 the once_repeat time, causing a random time value to be used if "once" was
5127 specified. (If the value happened to be <= 0, no repeat happened.) The
5128 default is now 0s, meaning "never repeat". The "vacation" command was OK
5129 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5130 (I found it when inspecting the code).
5132 20. There is now an overall timeout for performing a callout verification. It
5133 defaults to 4 times the callout timeout, which applies to individual SMTP
5134 commands during the callout. The overall timeout applies when there is more
5135 than one host that can be tried. The timeout is checked before trying the
5136 next host. This prevents very long delays if there are a large number of
5137 hosts and all are timing out (e.g. when the network connections are timing
5138 out). The value of the overall timeout can be changed by specifying an
5139 additional sub-option for "callout", called "maxwait". For example:
5141 verify = sender/callout=5s,maxwait=20s
5143 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5144 to the end before writing, but this should make it even safer).
5146 22. Exim was forgetting that it had advertised PIPELINING for the second and
5147 subsequent messages on an SMTP connection. It was also not resetting its
5148 memory on STARTTLS and an internal HELO.
5150 23. When Exim logs an SMTP synchronization error within a session, it now
5151 records whether PIPELINING has been advertised or not.
5153 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5154 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5155 rather than long int.
5157 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5159 26. Added the never_mail option to autoreply.
5165 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5166 crash if the getsockname() call failed; this can happen if a connection is
5167 closed very soon after it is established. The problem was simply in the
5168 order in which certain operations were done, causing Exim to try to write
5169 to the SMTP stream before it had set up the file descriptor. The bug has
5170 been fixed by making things happen in the correct order.
5176 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5177 before the connection was closed, thus losing the rejection response.
5179 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5180 some early Solaris releases, but causes trouble in current releases where
5181 socklen_t is defined.
5183 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5186 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5189 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5190 that releasing the top of it at the end releases what was used for sub-
5191 expansions (unless the block got too big). However, discard this block if
5192 the first thing is a variable or header, so that we can use its block when
5193 it is dynamic (useful for very large $message_headers, for example).
5195 6. Lookups now cache *every* query, not just the most recent. A new, separate
5196 store pool is used for this. It can be recovered when all lookup caches are
5197 flushed. Lookups now release memory at the end of their result strings.
5198 This has involved some general refactoring of the lookup sources.
5200 7. Some code has been added to the store_xxx() functions to reduce the amount
5201 of flapping under certain conditions.
5203 8. log_incoming_interface used to affect only the <= reception log lines. Now
5204 it causes the local interface and port to be added to several more SMTP log
5205 lines, for example "SMTP connection from", and rejection lines.
5207 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5209 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5211 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5212 could be overwritten at the end of the current message (or the start of a
5213 new message if it was set in a HELO ACL). The value is now preserved for
5214 the duration of the SMTP connection.
5216 12. If a transport had a headers_rewrite setting, and a matching header line
5217 contained an unqualified address, that address was qualified, even if it
5218 did not match any rewriting rules. The underlying bug was that the values
5219 of the flags that permit the existence of unqualified sender and recipient
5220 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5221 non-local messages, and by -bnq for local messages) were not being
5222 preserved with the message after it was received.
5224 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5225 "next input=" as part of the text comprising the host identity instead of
5226 the correct text. The code was using the same buffer for two different
5227 strings. However, depending on which order the printing function evaluated
5228 its arguments, the bug did not always show up. Under Linux, for example, my
5229 test suite worked just fine.
5231 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5232 doesn't work with some older versions of Perl. It has been changed to "my",
5233 which in any case is probably the better facility to use.
5235 15. A really picky compiler found some instances of statements for creating
5236 error messages that either had too many or two few arguments for the format
5239 16. The size of the buffer for calls to the DNS resolver has been increased
5240 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5241 for addresses that have a lot of PTR records. This alleviates a problem; it
5242 does not fully solve it.
5244 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5245 buffer now truncates the list and logs the incident, which is the same
5246 action as happens when Exim is looking up a host name and its aliases.
5247 Previously in this situation something unpredictable would happen;
5248 sometimes it was "internal error: store_reset failed".
5250 18. If a server dropped the connection unexpectedly when an Exim client was
5251 using GnuTLS and trying to read a response, the client delivery process
5252 crashed while trying to generate an error log message.
5254 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5255 string, for example:
5257 warn message = H1: something\nH2: something
5259 the text was added as a single header line from Exim's point of view
5260 though it ended up OK in the delivered message. However, searching for the
5261 second and subsequent header lines using $h_h2: did not work. This has been
5262 fixed. Similarly, if a system filter added multiple headers in this way,
5263 the routers could not see them.
5265 20. Expanded the error message when iplsearch is called with an invalid key to
5266 suggest using net-iplsearch in a host list.
5268 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5269 ACLs are no longer actually imposed (and a message to that effect is
5272 22. If a "gecos" field in a passwd entry contained escaped characters, in
5273 particular, if it contained a \" sequence, Exim got it wrong when building
5274 a From: or a Sender: header from that name. A second bug also caused
5275 incorrect handling when an unquoted " was present following a character
5276 that needed quoting.
5278 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5279 was not being matched caselessly.
5281 24. Arranged for all hyphens in the exim.8 source to be escaped with
5284 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5285 properly. Recipient callout cache records were still being keyed to include
5286 the sender, even when use_sender was set false. This led to far more
5287 callouts that were necessary. The sender is no longer included in the key
5288 when use_sender is false.
5290 26. Added "control = submission" modifier to ACLs.
5292 27. Added the ${base62d: operator to decode base 62 numbers.
5294 28. dnsdb lookups can now access SRV records.
5296 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5297 the configuration file.
5299 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5300 (-v) mode. This makes the output for a verbose queue run more intelligible.
5302 31. Added a use_postmaster feature to recipient callouts.
5304 32. Added the $body_zerocount variable, containing the number of binary zero
5305 bytes in the message body.
5307 33. The time of last modification of the "new" subdirectory is now used as the
5308 "mailbox time last read" when there is a quota error for a maildir
5311 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5313 35. Added +ignore_unknown as a special item in host lists.
5315 36. Code for decoding IPv6 addresses in host lists is now included, even if
5316 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5317 address was recognized as an IP address, but was then not correctly decoded
5318 into binary, causing unexpected and incorrect effects when compared with
5325 1. Very minor rewording of debugging text in manualroute to say "list of
5326 hosts" instead of "hostlist".
5328 2. If verify=header_syntax was set, and a header line with an unqualified
5329 address (no domain) and a large number of spaces between the end of the
5330 name and the colon was received, the reception process suffered a buffer
5331 overflow, and (when I tested it) crashed. This was caused by some obsolete
5332 code that should have been removed. The fix is to remove it!
5334 3. When running in the test harness, delay a bit after writing a bounce
5335 message to get a bit more predictability in the log output.
5337 4. Added a call to search_tidyup() just before forking a reception process. In
5338 theory, someone could use a lookup in the expansion of smtp_accept_max_
5339 per_host which, without the tidyup, could leave open a database connection.
5341 5. Added the variables $recipient_data and $sender_data which get set from a
5342 lookup success in an ACL "recipients" or "senders" condition, or a router
5343 "senders" option, similar to $domain_data and $local_part_data.
5345 6. Moved the writing of debug_print from before to after the "senders" test
5348 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5349 problems for message scanning, either using a data ACL, or using
5350 local_scan() because the Received: header was not generated till after they
5351 were called (in order to set the time as the time of reception completion).
5352 I have revised the way this works. The header is now generated after the
5353 body is received, but before the ACL or local_scan() are called. After they
5354 are run, the timestamp in the header is updated.
5360 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5361 before starting a queue runner without re-exec. This happened only when
5362 deliver_drop_privilege was set or when the Exim user was set to root. The
5363 effect of the bug was that timeouts during subsequent deliveries caused
5364 crashes instead of being properly handled. The handler is now left at its
5365 default (and expected) setting.
5367 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5368 message, again when deliver_drop_privilege is set or Exim is run as root.
5369 The bug described in (1) was not present in this case, but the tidying up
5370 of the other signals was missing. I have made the two cases consistent.
5372 3. The ignore_target_hosts setting on a manualroute router was being ignored
5373 for hosts that were looked up using the /MX notation.
5375 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5378 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5379 operated on the sender address. After changing the $sender_address to <>
5380 for the sender address verify, Exim was re-instated it as the original
5381 (before rewriting) address, but remembering that it had rewritten it, so it
5382 wasn't rewriting it again. This bug also had the effect of breaking the
5383 sender address verification caching when the sender address was rewritten.
5385 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5386 This has been changed so that if the ip literal address matches
5387 ignore_target_hosts, the router declines.
5389 7. Added expansion conditions match_domain, match_address, and match_local_
5390 part (NOT match_host).
5392 8. The placeholder for the Received: header didn't have a length field set.
5394 9. Added code to Exim itself and to exim_lock to test for a specific race
5395 condition that could lead to file corruption when using MBX delivery. The
5396 issue is with the lockfile that is created in /tmp. If this file is removed
5397 after a process has opened it but before that process has acquired a lock,
5398 there is the potential for a second process to recreate the file and also
5399 acquire a lock. This could lead to two Exim processes writing to the file
5400 at the same time. The added code performs the same test as UW imapd; it
5401 checks after acquiring the lock that its file descriptor still refers to
5402 the same named file.
5404 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5405 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5408 11. Added the smtp_active_hostname option. If used, this will typically be made
5409 to depend on the incoming interface address. Because $interface_address is
5410 not set up until the daemon has forked a reception process, error responses
5411 that can happen earlier (such as "too many connections") no longer contain
5414 12. If an expansion in a condition on a "warn" statement fails because a lookup
5415 defers, the "warn" statement is abandoned, and the next ACL statement is
5416 processed. Previously this caused the whole ACL to be aborted.
5418 13. Added the iplsearch lookup type.
5420 14. Added ident_timeout as a log selector.
5422 15. Added tls_certificate_verified as a log selector.
5424 16. Added a global option tls_require_ciphers (compare the smtp transport
5425 option of the same name). This controls incoming TLS connections.
5427 17. I finally figured out how to make tls_require_ciphers do a similar thing
5428 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5429 before starting the TLS session.
5431 18. Tabs are now shown as \t in -bP output.
5433 19. If the log selector return_path_on_delivery was set, Exim crashed when
5434 bouncing a message because it had too many Received: header lines.
5436 20. If two routers both had headers_remove settings, and the first one included
5437 a superfluous trailing colon, the final name in the first list and the
5438 first name in the second list were incorrectly joined into one item (with a
5439 colon in the middle).
5445 1. Added -C and -D options to the exinext utility, mainly to make it easier
5446 to include in the automated testing, but these could be helpful when
5447 multiple configurations are in use.
5449 2. The exinext utility was not formatting the output nicely when there was
5450 an alternate port involved in the retry record key, nor when there was a
5451 message id as well (for retries that were specific to a specific message
5452 and a specific host). It was also confused by IPv6 addresses, because of
5453 the additional colons they contain. I have fixed the IPv4 problem, and
5454 patched it up to do a reasonable job for IPv6.
5456 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5457 delivery, the log line now contains "pipelined" if PIPELINING was used.
5459 4. An SMTP transport process used to panic and die if the bind() call to set
5460 an explicit outgoing interface failed. This has been changed; it is now
5461 treated in the same way as a connect() failure.
5463 5. A reference to $sender_host_name in the part of a conditional expansion
5464 that was being skipped was still causing a DNS lookup. This no longer
5467 6. The def: expansion condition was not recognizing references to header lines
5468 that used bh_ and bheader_.
5470 7. Added the _cache feature to named lists.
5472 8. The code for checking quota_filecount in the appendfile transport was
5473 allowing one more file than it should have been.
5475 9. For compatibility with Sendmail, the command line option
5483 and sets the incoming protocol and host name (for trusted callers). The
5484 host name and its colon can be omitted when only the protocol is to be set.
5485 Note the Exim already has two private options, -pd and -ps, that refer to
5486 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5487 "s", but I don't think that's a major issue.
5489 10. A number of refactoring changes to the code, none of which should affect
5492 (a) The number of logging options was getting close to filling up the
5493 32-bit word that was used as a bit map. I have split them into two classes:
5494 those that are passed in the argument to log_write(), and those that are
5495 only ever tested independently outside of that function. These are now in
5496 separate 32-bit words, so there is plenty of room for expansion again.
5497 There is no change in the user interface or the logging behaviour.
5499 (b) When building, for example, log lines, the code previously used a
5500 macro that called string_cat() twice, in order to add two strings. This is
5501 not really sufficiently general. Furthermore, there was one instance where
5502 it was actually wrong because one of the argument was used twice, and in
5503 one call a function was used. (As it happened, calling the function twice
5504 did not affect the overall behaviour.) The macro has been replaced by a
5505 function that can join an arbitrary number of extra strings onto a growing
5508 (c) The code for expansion conditions now uses a table and a binary chop
5509 instead of a serial search (which was left over from when there were very
5510 few conditions). Also, it now recognizes conditions like "pam" even when
5511 the relevant support is not compiled in: a suitably worded error message is
5512 given if an attempt is made to use such a condition.
5514 11. Added ${time_interval:xxxxx}.
5516 12. A bug was causing one of the ddress fields not to be passed back correctly
5517 from remote delivery subprocesses. The field in question was not being
5518 subsequently used, so this caused to problems in practice.
5520 13. Added new log selectors queue_time and deliver_time.
5522 14. Might have fixed a bug in maildirsizefile handling that threw up
5523 "unexpected character" debug warnings, and recalculated the data
5524 unnecessarily. In any case, I expanded the warning message to give more
5527 15. Added the message "Restricted characters in address" to the statements in
5528 the default ACL that block characters like @ and % in local parts.
5530 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5531 Three changes have been made:
5533 (a) There was a serious bug; a negative response to MAIL caused the whole
5534 recipient domain to be cached as invalid, thereby blocking all messages
5535 to all local parts at the same domain, from all senders. This bug has
5536 been fixed. The domain is no longer cached after a negative response to
5537 MAIL if the sender used is not empty.
5539 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5542 (c) A new callout option, "use_sender" has been added for people who want
5543 the modified behaviour.
5549 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5552 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5553 indeed breaks things for older releases.
5555 3. Added additional logging to the case where there is a problem reading data
5556 from a filter that is running in a subprocess using a pipe, in order to
5557 try to track down a specific problem.
5559 4. Testing facility fudge: when running in the test harness and attempting
5560 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5561 getting "No route to host". Convert this to a timeout.
5563 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5566 6. Some OS don't have socklen_t but use size_t instead. This affects the
5567 fifth argument of getsockopt() amongst other things. This is now
5568 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5569 can be set for individual OS. I have set it for SunOS5, OSF1, and
5570 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5571 some earlier ones do not.
5573 7. Change 4.30/15 was not doing the test caselessly.
5575 8. The standard form for an IPv6 address literal was being rejected by address
5576 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5577 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5578 this, as well as the form without the "IPv6" on the front (but only when
5579 address literals are enabled, of course).
5581 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5583 10. Exim crashed if a message with an empty sender address specified by -f
5584 encountered a router with an errors_to setting. This could be provoked only
5585 by a command such as
5589 where an empty string was supplied; "<>" did not hit this bug.
5591 11. Installed PCRE release 4.5.
5593 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5594 remained set. It is now erased.
5596 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5597 times from message ids (which are base 36 rather than the normal 62).
5599 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5600 were being counted as actual protocol errors, and logged if the log
5601 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5602 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5603 and DATA following a set of rejected RCPTs do not count as protocol errors.
5604 In other words, Exim assumes they were pipelined, though this may not
5605 actually be the case. Of course, in all cases the client gets an
5606 appropriate error code.
5608 15. If a lookup fails in an ACL condition, a message about the failure may
5609 be available; it is used if testing the ACL cannot continue, because most
5610 such messages specify what the cause of the deferral is. However, some
5611 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5612 that caused an old message to be retained and used if a later statement
5613 caused a defer, replacing the real cause of the deferral.
5615 16. If an IP address had so many PTR records that the DNS lookup buffer
5616 was not large enough to hold them, Exim could crash while trying to process
5617 the truncated data. It now detects and logs this case.
5619 17. Further to 4.21/58, another change has been made: if (and only if) the
5620 first line of a message (the first header line) ends with CRLF, a bare LF
5621 in a subsequent header line has a space inserted after it, so as not to
5622 terminate the header.
5624 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5625 unnecessarily, used atoi() instead of strtol(), and didn't check the
5626 termination when getting file sizes from file names by regex.
5628 19. Completely re-implemented the support for maildirsize files, in the light
5629 of a number of problems with the previous contributed implementation
5630 (4.30/29). In particular:
5632 . If the quota is zero, the maildirsize file is maintained, but no quota is
5635 . If the maildir directory does not exist, it is created before any attempt
5636 to write a maildirsize file.
5638 . The quota value in the file is just a cache; if the quota is changed in
5639 the transport, the new value overrides.
5641 . A regular expression is available for excluding directories from the
5644 20. The autoreply transport checks the characters in options that define the
5645 message's headers; it allows continued headers, but it was checking with
5646 isspace() after an embedded newline instead of explicitly looking for a
5649 21. If all the "regular" hosts to which an address was routed had passed their
5650 expiry times, and had not reached their retry times, the address was
5651 bounced, even if fallback hosts were defined. Now Exim should go on to try
5654 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5655 equivalent code in the SMTP transport. Some hosts send humungous responses
5656 to HELO/EHLO, more than 1024 it seems.
5658 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5659 gives compiler warnings in some environments. I've now done it "properly",
5662 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5663 (because of problems with the built-in one) was declared to return uschar *
5664 instead of char *, causing compiler failure.
5666 25. Fixed a file descriptor leak when processing alias/forward files.
5668 26. Fixed a minor format string issue in dbfn.c.
5670 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5672 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5673 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5676 29. When a sender address is verified, it is cached, to save repeating the test
5677 when there is more than one recipient in a message. However, when the
5678 verification involves a callout, it is possible for different callout
5679 options to be set for different recipients. It is too complicated to keep
5680 track of this in the cache, so now Exim always runs a verification when a
5681 callout is required, relying on the callout cache for the optimization.
5682 The overhead is duplication of the address routing, but this should not be
5685 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5686 to be verified with callout=postmaster, and the main callout worked but the
5687 postmaster check failed, the verification correctly failed. However, if a
5688 subsequent RCPT command asked for sender verification *without* the
5689 postmaster check, incorrect caching caused this verification also to fail,
5692 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5693 it was not caching the DNS options (qualify_single, search_parents) that
5694 were used when the lookup failed. A subsequent lookup with different
5695 options therefore always gave the same answer, though there were cases
5696 where it should not have. (Example: a "domains = !$mx_any" option on a
5697 dnslookup router: the "domains" option is always processed without any
5698 widening, but the router might have qualify_single set.) Now Exim uses the
5699 cached value only when the same options are set.
5701 32. Added John Jetmore's "exipick" utility to the distribution.
5703 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5704 than a timeout (e.g. a certificate is required, and is not provided), an
5705 Exim server now closes the connection immediately. Previously it waited for
5706 the client to close - but if the client is SSL, it seems that they each
5707 wait for each other, leading to a delay before one of them times out.
5709 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5710 maintained 0.8.x compatibility because I don't think many are using it, and
5711 it is clearly obsolete.
5713 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5716 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5717 client certificate was expired. A simple patch fixes this, though I don't
5718 understand the full logic of why the verify callback is called multiple
5721 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5722 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5723 which causes problems with some clients (such as the Certicom SSL Plus
5724 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5725 disables the coutermeasure allowing Eudora to connect."
5727 38. Exim was not checking that a write() to a log file succeeded. This could
5728 lead to Bad Things if a log got too big, in particular if it hit a file
5729 size limit. Exim now panics and dies if it cannot write to a log file, just
5730 as it does if it cannot open a log file.
5732 39. Modified OS/Makefile-Linux so that it now contains
5734 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5736 The two -D definitions ensure that Exim is compiled with large file
5737 support, which makes it possible to handle log files that are bigger than
5740 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5741 instance) a domain was checked against a named list that involved a lookup,
5742 causing $domain_data to be set, then another domain was checked against the
5743 same list, then the first domain was re-checked, the value of $domain_data
5744 after the final check could be wrong. In particular, if the second check
5745 failed, it could be set empty. This bug probably also applied to
5748 41. The strip_trailing_dot option was not being applied to the address given
5749 with the -f command-line option.
5751 42. The code for reading a message's header from the spool was incrementing
5752 $received_count, but never initializing it. This meant that the value was
5753 incorrect (doubled) while delivering a message in the same process in which
5754 it was received. In the most common configuration of Exim, this never
5755 happens - a fresh exec is done - but it can happen when
5756 deliver_drop_privilege is set.
5758 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5759 it now includes up to 150 characters of the unexpected data in the log
5762 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5763 and building data strings. The size of both of these buffers was 10 000
5764 bytes - far larger than anybody would *ever* want, thought I. Needless to
5765 say, somebody hit the limit. I have increased the maximum line length to
5766 20 000 and the maximum data length of concatenated lines to 100 000. I have
5767 also fixed two bugs, because there was no checking on these buffers. Tsk,
5768 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5769 buffer is too small.
5771 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5772 lsearch lookups. Now it does.
5774 46. When parsing a route_list item in a manualroute router, a fixed-length
5775 buffer was used for the list of hosts. I made this 1024 bytes long,
5776 thinking that nobody would ever have a list of hosts that long. Wrong.
5777 Somebody had a whole pile of complicated expansion conditions, and the
5778 string was silently truncated, leading to an expansion error. It turns out
5779 that it is easier to change to an unlimited length (owing to other changes
5780 that have happened since this code was originally written) than to build
5781 structure for giving a limitation error. The length of the item that
5782 expands into the list of hosts is now unlimited.
5784 47. The lsearch lookup could not handle data where the length of text line was
5785 more than 4095 characters. Such lines were truncated, leading to shortened
5786 data being returned. It should now handle lines of any length.
5788 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5789 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5792 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5793 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5794 no longer happens in comment lines. A list of replacements is now placed
5795 at the head of all of the source files, except those whose only change is
5796 to replace PERL_COMMAND in the very first #! line.
5798 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5799 messages on the queue, with a bottom-up merge sort, using code contributed
5800 by Michael Haardt. This should make operations like -bp somewhat faster on
5801 large queues. It won't affect queue runners, except when queue_run_in_order
5804 51. Installed eximstats 1.31 in the distribution.
5806 52. Added support for SRV lookups to the dnslookup router.
5808 53. If an ACL referred to $message_body or $message_body_end, the value was not
5809 reset for any messages that followed in the same SMTP session.
5811 54. The store-handling optimization for building very long strings was not
5812 differentiating between the different store pools. I don't think this
5813 actually made any difference in practice, but I've tidied it.
5815 55. While running the routers to verify a sender address, $sender_address
5816 was still set to the sender address. This is wrong, because when routing to
5817 send a bounce to the sender, it would be empty. Therefore, I have changed
5818 it so that, while verifying a sender address, $sender_address is set to <>.
5819 (There is no change to what happens when verifying a recipient address.)
5821 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5822 target A or AAAA records (if not already returned) without resetting the
5823 qualify_single or search_parents options of the DNS resolver. These are
5824 inappropriate in this case because the targets of MX and SRV records must
5825 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5826 target that, when qualified, matched something in the local domain. These
5827 two options are now turned off when doing these lookups.
5829 57. It seems that at least some releases of Reiserfs (which does not have the
5830 concept of a fixed number of inodes) returns zero and not -1 for the
5831 number of available inodes. This interacted badly with check_spool_inodes,
5832 which assumed that -1 was the "no such thing" setting. What I have done is
5833 to check that the total number of inodes is greater than zero before doing
5834 the test of how many are available.
5836 58. When a "warn" ACL statement has a log_message modifier, the message is
5837 remembered, and not repeated. This is to avoid a lot of repetition when a
5838 message has many recipients that cause the same warning to be written.
5839 However, Exim was preserving the list of already written lines for an
5840 entire SMTP session, which doesn't seem right. The memory is now reset if a
5841 new message is started.
5843 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5844 debugging output unless log_rewrite was also set.
5846 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5847 of (int)(handle) when we know that handle contains (void *)(-1).
5849 61. The Exim daemon panic-logs an error return when it closes the incoming
5850 connection. However "connection reset by peer" seems to be common, and
5851 isn't really an error worthy of noting specially, so that particular error
5854 62. When Exim is trying to find all the local interfaces, it used to panic and
5855 die if the ioctl to get the interface flags failed. However, it seems that
5856 on at least one OS (Solaris 9) it is possible to have an interface that is
5857 included in the list of interfaces, but for which you get a failure error
5858 for this call. This happens when the interface is not "plumbed" into a
5859 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5860 failure of the "get flags" call assumes that the interface is down.
5862 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5863 makes life easier for people who are doing arithmetic on fields extracted
5864 from dates, where you often get leading zeros that should not be
5865 interpreted as octal.
5867 64. Added qualify_domain to the redirect router, to override the global
5870 65. If a pathologically long header line contained very many addresses (the
5871 report of this problem mentioned 10 000) and each of them was rewritten,
5872 Exim could use up a very large amount of memory. (It kept on making new
5873 copies of the header line as it rewrote, and never released the old ones.)
5874 At the expense of a bit more processing, the header rewriting function has
5875 been changed so that it no longer eats memory in this way.
5877 66. The generation of the Received: header has been moved from the time that a
5878 message starts to be received, to the time that it finishes. The timestamp
5879 in the Received: header should now be very close to that of the <= log
5880 line. There are two side-effects of this change:
5882 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5883 logged header lines no longer include the local Received: line, because
5884 it has not yet been created. The same applies to a copy of the message
5885 that is returned to a non-SMTP sender when a message is rejected.
5887 (b) When a filter file is tested using -bf, no additional Received: header
5888 is added to the test message. After some thought, I decided that this
5891 This change does not affect the value of $received_for. It is still set
5892 after address rewriting, but before local_scan() is called.
5894 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5896 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5897 gave an unhelpful panic error message, and a defer error. I have managed to
5898 change this behaviour so that it now rejects any supplied certificate,
5899 which seems right, as the list of acceptable certificates is empty.
5901 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5902 gave an unhelpful defer error. I have not managed to make this reject any
5903 supplied certificates, but the error message it gives is "no certificate
5904 supplied", which is not helpful.
5906 70. exigrep's output now also includes lines that are not associated with any
5907 message, but which match the given pattern. Implemented by a patch from
5908 Martin Sluka, which also tidied up the Perl a bit.
5910 71. Recipient callout verification, like sender verification, was using <> in
5911 the MAIL FROM command. This isn't really the right thing, since the actual
5912 sender may affect whether the remote host accepts the recipient or not. I
5913 have changed it to use the actual sender in the callout; this means that
5914 the cache record is now keyed on a recipient/sender pair, not just the
5915 recipient address. There doesn't seem to be a real danger of callout loops,
5916 since a callout by the remote host to check the sender would use <>.
5917 [SEE ABOVE: changed after hitting problems.]
5919 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5920 temporary errors. However, in the case of such a code being given after
5921 the end of a data transmission (i.e. after ".") Exim was failing to write
5922 a retry record for the message. (Yes, there was some broken host that was
5923 actually sending 8xx at this point.)
5925 73. An unknown lookup type in a host list could cause Exim to panic-die when
5926 the list was checked. (An example that provoked this was putting <; in the
5927 middle of a list instead of at the start.) If this happened during a DATA
5928 ACL check, a -D file could be left lying around. This kind of configuration
5929 error no longer causes Exim to die; instead it causes a defer error. The
5930 incident is still logged to the main and panic logs.
5932 74. Buglet left over from Exim 3 conversion. The message "too many messages
5933 in one connection" was written to the rejectlog but not the mainlog, except
5934 when address rewriting (yes!) was being logged.
5936 75. Added write_rejectlog option.
5938 76. When a system filter was run not as root (that is, when system_filter_user
5939 was set), the values of the $n variables were not being returned to the
5940 main process; thus, they were not subsequently available in the $sn
5943 77. Added +return_path_on_delivery log selector.
5945 78. A connection timeout was being treated differently from recipients deferred
5946 when testing hosts_max_try with a message that was older than the host's
5947 retry timeout. (The host should not be counted, thus allowing all hosts to
5948 be tried at least once before bouncing.) This may have been the cause of an
5949 occasionally reported bug whereby a message would remain on the queue
5950 longer than the retry timeout, but would be bounced if a delivery was
5951 forced. I say "may" because I never totally pinned down the problem;
5952 setting up timeout/retry tests is difficult. See also the next item.
5954 79. The ultimate address timeout was not being applied to errors that involved
5955 a combination of host plus message (for example, a timeout on a MAIL
5956 command). When an address resolved to a number of possible hosts, and they
5957 were not all tried for each delivery (e.g. because of hosts_max_try), a
5958 message could remain on the queue longer than the retry timeout.
5960 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5963 81. Fixed an obscure SMTP outgoing bug which required at least the following
5964 conditions: (a) there was another message waiting for the same server;
5965 (b) the server returned 5xx to all RCPT commands in the first message so
5966 that the message was not completed; (c) the server dropped the connection
5967 or gave a negative response to the RSET that Exim sends to abort the
5968 transaction. The observed case was a dropped connection after DATA that had
5969 been sent in pipelining mode. That is, the server had advertised PIPELINING
5970 but was not implementing it correctly. The effect of the bug was incorrect
5971 behaviour, such as trying another host, and this could lead to a crash.
5977 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5978 and daemon.c were passed as pointers to ints; they should have been
5979 pointers to socklen_t variables (which are typically unsigned ints).
5981 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5984 3. Fixed a really odd bug that affected only the testing scheme; patching a
5985 certain fixed string in the binary changed the value of another string that
5986 happened to be identical to the end of the original first string.
5988 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5989 name", it returns that address as the IP address. On some operating
5990 systems (e.g. Solaris), it also passes back the IP address string as the
5991 "host name". However, on others (e.g. Linux), it passes back an empty
5992 string. Exim wasn't checking for this, and was changing the host name to an
5993 empty string, assuming it had been canonicalized.
5995 5. Although rare, it is permitted to have more than one PTR record for a given
5996 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5997 all the names associated with an address, because they do in Solaris.
5998 However, it seems that they do not in Linux for data that comes from the
5999 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6000 I found this out when I moved to a new Linux workstation and tried to run
6001 the Exim test suite.
6003 To get round this problem I have changed the code so that it now does its
6004 own call to the DNS to look up PTR records when searching for a host name.
6005 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6006 addresses that are only in /etc/hosts are still found.
6008 This behaviour is, however, controlled by an option called host_lookup_
6009 order, which defaults to "bydns:byaddr". If people want to use the other
6010 order, or indeed, just use one or the other means of lookup, they can
6011 specify it in this variable.
6013 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6014 some operating systems, this comes back from gethostbyaddr() as an empty
6015 string, and this is what Exim used to test for. However, it seems that in
6016 other systems, "." is yielded. Exim now tests for this case too.
6018 7. The values of check_spool_space and check_log_space are now held internally
6019 as a number of kilobytes instead of an absolute number of bytes. If a
6020 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6021 kilobyte. This means that much larger values can be stored.
6023 8. Exim monitor: an attempt to get the action menu when not actually pointing
6024 at a message produces an empty menu entitled "No message selected". This
6025 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6026 no entries in it ("Shell widget menu has zero width and/or height"). So I
6027 have added a single, blank menu entry in this case.
6029 9. Added ${quote_local_part.
6031 10. MIME decoding is now applied to the contents of Subject: header lines when
6034 11. Now that a reference to $sender_host_address automatically causes a reverse
6035 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6036 host lookup before query-style lookups in lists that might use this
6037 variable. This has therefore been abolished, and the "net-" prefix is no
6038 longer necessary for query-style lookups.
6040 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6041 have been a typo for LFLAGS, so it has been changed.
6043 13. The install script calls Exim with "-C /dev/null" in order to find the
6044 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6045 to be output. However, since Exim outputs its version number before the
6046 error, it didn't break the script. It just looked ugly. I fixed this by
6047 always allowing "-C /dev/null" if the caller is root.
6049 14. Ignore overlarge ACL variable number when reading spool file - insurance
6050 against a later release with more variables having written the file.
6052 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6053 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6054 this, as well as the form without the "IPv6" on the front.
6056 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6057 OS/Makefile-Darwin file.
6059 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6060 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6062 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6063 variable. However, some OS use ints for inodes. I've added cast to long int
6064 to get rid of the compiler warning.
6066 19. I had forgotten to lock out "/../" in configuration file names when
6067 ALT_CONFIG_PREFIX was set.
6069 20. Routers used for verification do not need to specify transports. However,
6070 if such a router generated a host list, and callout was configured, Exim
6071 crashed, because it could not find a port number from the (non-existent)
6072 transport. It now assumes port 25 in this circumstance.
6074 21. Added the -t option to exigrep.
6076 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6077 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6078 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6079 them is set, LOOKUP_LSEARCH is forced.
6081 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6083 24. Added sender and host information to the "rejected by local_scan()" log
6084 line; previously there was no indication of these.
6086 25. Added .include_if_exists.
6088 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6089 Linux. It turns out that not all file systems support this. Apparently some
6090 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6091 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6092 not supported on the file descriptor, is now ignored when Exim is trying to
6093 sync a directory. This applies only to Linux.
6095 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6097 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6098 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6101 29. Added support for maildirsize files from supplied patch (modified a bit).
6103 30. The use of :fail: followed by an empty string could lead Exim to respond to
6104 sender verification failures with (e.g.):
6106 550 Verification failed for <xxx>
6107 550 Sender verify failed
6109 where the first response line was missing the '-' that indicates it is not
6110 the final line of the response.
6112 31. The loop for finding the name of the user that called Exim had a hardwired
6113 limit of 10; it now uses the value of finduser_retries, which is used for
6114 all other user lookups.
6116 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6119 33. Exim was neglecting to zero errno before one call of strtol() when
6120 expanding a string and expecting an integer value. On some systems this
6121 resulted in spurious "integer overflow" errors. Also, it was casting the
6122 result into an int without checking.
6124 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6125 did not work. The code looks as if it has *never* worked, though it appears
6126 to have been documented since at least release 1.62. I have made it work.
6128 35. The "timeout_DNS" error in retry rules, also documented since at least
6129 1.62, also never worked. As it isn't clear exactly what this means, and
6130 clearly it isn't a major issue, I have abolished the feature by treating it
6131 as "timeout", and writing a warning to the main and panic logs.
6133 36. The display of retry rules for -brt wasn't always showing the error code
6136 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6137 timeout_connect_A, timeout_connect_MX.
6139 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6140 to the empty sender.
6142 39. The daemon was not analysing the content of -oX till after it had closed
6143 stderr and disconnected from the controlling terminal. This meant that any
6144 syntax errors were only noted on the panic log, and the return code from
6145 the command was 0. By re-arranging the code a little, I've made the
6146 decoding happen first, so such errors now appear on stderr, and the return
6147 code is 1. However, the actual setting up of the sockets still happens in
6148 the disconnected process, so errors there are still only recorded on the
6151 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6152 connections (as happens on some IP stacks) was logged at start up time as
6153 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6154 it from "IPv6 and IPv4", which means that two separate sockets are being
6157 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6158 says whether AF_INET or AF_INET6 was passed as an argument.
6160 42. Exiwhat output was messed up when time zones were included in log
6163 43. Exiwhat now gives more information about the daemon's listening ports,
6164 and whether -tls-on-connect was used.
6166 44. The "port" option of the smtp transport is now expanded.
6168 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6169 silently ignored. Now an error message is written to the main and panic
6172 46. There's a new ACL modifier called "logwrite" which writes to a log file
6173 as soon as it is encountered.
6175 47. Added $local_user_uid and $local_user_gid at routing time.
6177 48. Exim crashed when trying to verify a sender address that was being
6180 49. Exim was recognizing only a space character after ".include". It now also
6181 recognizes a tab character.
6183 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6184 extracting the relevant information from the specification. The man page no
6185 longer contains scrambled data for the -d option, and I've added a section
6186 at the front about calling Exim under different names.
6188 51. Added "extra_headers" argument to the "mail" command in filter files.
6190 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6193 53. Installed eximstats 1.29.
6195 54. Added transport_filter_timeout as a generic transport option.
6197 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6198 Cc: header lines. This was required by RFC 822, but it not required by RFC
6201 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6202 incoming messages that did not have them. Now it does so only if the
6203 message originates locally, that is, if there is no associated remote host
6204 address. When Resent- header lines are present, this applies to the Resent-
6205 lines rather than the non-Resent- lines.
6207 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6208 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6210 58. Messages for configuration errors now include the name of the main
6211 configuration file - useful now that there may be more than one file in a
6212 list (.included file names were always shown).
6214 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6215 for those rare installations that do not start the daemon as root or run it
6216 setuid root. I've cut out the call to initgroups() if the daemon is not
6219 60. The Exim user and group can now be bound into the binary as text strings
6220 that are looked up at the start of Exim's processing.
6222 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6224 62. Added $mailstore_basename variable.
6226 63. Installed patch to sieve.c from Michael Haardt.
6228 64. When Exim failed to open the panic log after failing to open the main log,
6229 the original message it was trying to log was written to stderr and debug
6230 output, but if they were not available (the usual case in production), it
6231 was lost. Now it is written to syslog before the two lines that record the
6232 failures to open the logs.
6234 65. Users' Exim filters run in subprocesses under the user's uid. It is
6235 possible for a "deliver" command or an alias in a "personal" command to
6236 provoke an address rewrite. If logging of address rewriting is configured,
6237 this fails because the process is not running as root or exim. There may be
6238 a better way of dealing with this, but for the moment (because 4.30 needs
6239 to be released), I have disabled address rewrite logging when running a
6240 filter in a non-root, non-exim process.
6246 1. The buildconfig auxiliary program wasn't quoting the value set for
6247 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6248 not defined. This bug was masked in 4.22 by the effect that was fixed in
6251 2. Some messages that were rejected after a message id was allocated were
6252 shown as "incomplete" by exigrep. It no longer does this for messages that
6253 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6255 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6256 have allow_domain_literals set, the ID did not get logged in the <= line.
6257 Domain literals are now always recognized in Message-ID: header lines.
6259 4. The first argument for a ${extract expansion item is the key name or field
6260 number. Leading and trailing spaces in this item were not being ignored,
6261 causing some misleading effects.
6263 5. When deliver_drop_privilege was set, single queue runner processes started
6264 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6265 same command in the process it spins off) were not dropping privilege.
6267 6. When the daemon running as "exim" started a queue runner, it always
6268 re-executed Exim in the spun-off process. This is a waste of effort when
6269 deliver_drop_privilege is set. The new process now just calls the
6270 queue-runner function directly.
6276 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6279 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6280 ignored. Though the use of -oP was forcing the writing of a pid file, it
6281 was always written to the default place.
6283 3. If the message "no IP address found for host xxxx" is generated during
6284 incoming verification, it is now followed by identification of the incoming
6285 connection (so you can more easily find what provoked it).
6287 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6289 5. Added some features to "harden" Exim a bit more against certain attacks:
6291 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6292 be put in Local/Makefile. This is like the never_users runtime option,
6293 but it cannot be overridden. The default setting is "root".
6295 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6296 prefix string with which any file named in a -C command line option
6299 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6300 is retained for -C and -D only if the caller of Exim is root. Without
6301 it, the exim user may also use -C and -D and retain privilege.
6303 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6304 command line option is disabled.
6306 6. Macro names set by the -D option must start with an upper case letter, just
6307 like macro names defined in the configuration file.
6309 7. Added "dereference=" facility to LDAP.
6311 8. Two instances of the typo "uknown" in the source files are fixed.
6313 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6314 the Configure-Makefile script screwed up while processing it.
6316 10. Incorporated PCRE 4.4.
6318 11. The SMTP synchronization check was not operating right at the start of an
6319 SMTP session. For example, it could not catch a HELO sent before the client
6320 waited for the greeting. There is now a check for outstanding input at the
6321 point when the greeting is written. Because of the duplex, asynchronous
6322 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6323 way, but not yet received, when the check is performed.
6325 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6326 on TCP/IP sockets, because this apparently causes some broken clients to
6329 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6330 unchanged) from the Cygwin maintainer.
6332 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6333 was supported instead of testing for mbx. Effectively a typo.
6335 15. The spa authenticator server code was not checking that the input it
6336 received was valid base64.
6338 16. The debug output line for the "set" modifier in ACLs was not showing the
6339 name of the variable that was being set.
6341 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6343 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6344 lookup on its own. Something else was needed to trigger the lookup. For
6345 example, a match in host_lookup or the need for a host name in a host list.
6346 Now, if $sender_host_name is referenced and the host name has not yet been
6347 looked up, a lookup is performed. If the lookup fails, the variable remains
6348 empty, and $host_lookup_failed is set to "1".
6350 19. Added "eqi" as a case-independent comparison operator.
6352 20. The saslauthd authentication condition could segfault if neither service
6353 nor realm was specified.
6355 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6356 error message that was logged was misleading, and incoming SMTP
6357 connections were dropped. The message is now more accurate, and temporary
6358 errors are given to SMTP connections.
6360 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6361 (except RSET) with a 421 error, until QUIT is received. However, it was
6362 failing to send a response to QUIT.
6364 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6365 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6366 ACL is now run after the helo_try_verify_hosts code.
6368 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6369 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6370 case-independent, so other case variants are also recognized). Apparently
6371 some systems use these upper case variants.
6373 25. If more than two messages were waiting for the same host, and a transport
6374 filter was specified for the transport, Exim sent two messages over the
6375 same TCP/IP connection, and then failed with "socket operation on non-
6376 socket" when it tried to send the third.
6378 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6380 27. The extern definition of crypt16() in expand.c was not being excluded when
6381 the OS had its own crypt16() function.
6383 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6384 as a preferred synonym for return_size_limit, both as an option and as an
6387 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6389 30. Changed the default configuration ACL to relax the local part checking rule
6390 for addresses that are not in any local domains. For these addresses,
6391 slashes and pipe symbols are allowed within local parts, but the sequence
6392 /../ is explicitly forbidden.
6394 31. SPA server authentication was not clearing the challenge buffer before
6397 32. log_message in a "warn" ACL statement was writing to the reject log as
6398 well as to the main log, which contradicts the documentation and doesn't
6399 seem right (because no rejection is happening). So I have stopped it.
6401 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6402 However, I am unable to do any testing of this.
6404 34. Fixed an infelicity in the appendfile transport. When checking directories
6405 for a mailbox, to see if any needed to be created, it was accidentally
6406 using path names with one or more superfluous leading slashes; tracing
6407 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6409 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6410 contents are added to the log line that is written for every discarded
6411 recipient. (Previously a log_message setting was ignored.)
6413 36. The ${quote: operator now quotes the string if it is empty.
6415 37. The install script runs exim in order to find its version number. If for
6416 some reason other than non-existence or emptiness, which it checks, it
6417 could not run './exim', it was installing it with an empty version number,
6418 i.e. as "exim-". This error state is now caught, and the installation is
6421 38. An argument was missing from the function that creates an error message
6422 when Exim fails to connect to the socket for saslauthd authentication.
6423 This could cause Exim to crash, or give a corrupted message.
6425 39. Added isip, isip4, and isip6 to ${if conditions.
6427 40. The ACL variables $acl_xx are now saved with the message, and can be
6428 accessed later in routers, transports, and filters.
6430 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6431 strings in the file are not string-expanded.
6433 42. If a MAIL command specified a SIZE value that was too large to fit into an
6434 int variable, the check against message_size_limit failed. Such values are
6435 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6436 day this will have to be increased, but I don't think I want to be around
6437 when emails are that large.
6444 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6445 iconv() is not standard in FreeBSD.
6447 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6448 IPv6 enabled. The observed symptom was a segmentation fault on return from
6449 the function os_common_find_running_interfaces() in src/os.c.
6451 3. In the check_special_case() function in daemon.c I had used "errno" as an
6452 argument name, which causes warnings on some systems. This was basically a
6453 typo, since it was named "eno" in the comments!
6455 4. The code that waits for the clock to tick (at a resolution of some fraction
6456 of a second) so as to ensure message-id uniqueness was always waiting for
6457 at least one whole tick, when it could have waited for less. [This is
6458 almost certainly not relevant at current processor speeds, where it is
6459 unlikely to ever wait at all. But we try to future-proof.]
6461 5. The function that sleeps for a time interval that includes fractions of a
6462 second contained a race. It did not block SIGALRM between setting the
6463 timer, and suspending (a couple of lines later). If the interval was short
6464 and the sigsuspend() was delayed until after it had expired, the suspension
6465 never ended. On busy systems this could lead to processes getting stuck for
6468 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6469 process, before it forks any delivery processes. The open lookup caching
6470 mechanism meant that the open file or database connection was passed into
6471 the delivery process. The problem was that delivery processes always tidy
6472 up cached lookup data. This could cause a problem for the next delivery
6473 process started by the queue runner, because the external queue runner
6474 process does not know about the closure. So the next delivery process
6475 still has data in the lookup cache. In the case of a file lookup, there was
6476 no problem because closing a file descriptor in a subprocess doesn't affect
6477 the parent. However, if the lookup was caching a connection to a database,
6478 the connection was closed, and the second delivery process was likely to
6479 see errors such as "PGSQL: query failed: server closed the connection
6480 unexpectedly". The problem has been fixed by closing all cached lookups
6481 in a queue runner before running a delivery process.
6483 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6484 seem to have the "const" qualifier which it has on other OS. I've
6487 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6488 *of the same type* (client or server) with the same public name that an
6489 error should be diagnosed.
6491 9. When Exim looked up a host name for an IP address, but failed to find the
6492 original IP address when looking up the host name (a safety check), it
6493 output the message "<ip address> does not match any IP for NULL", which was
6494 confusing, to say the least. The bug was that the host name should have
6495 appeared instead of "NULL".
6497 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6498 user that is built into the binary, and the -C or -D options is used, root
6499 privilege is dropped before the configuration file is read. In addition,
6500 logging is switched to stderr instead of the normal log files. If the
6501 configuration then re-defines the Exim user, the unprivileged environment
6502 is probably not what is expected, so Exim logs a panic warning message (but
6505 However, if deliver_drop_privilege is set, the unprivileged state may well
6506 be exactly what is intended, so the warning has been cut out in that case,
6507 and Exim is allowed to try to write to its normal log files.
6513 1. smtp_return_error_details was not giving details for temporary sender
6514 or receiver verification errors.
6516 2. Diagnose a configuration error if two authenticators have the same public
6519 3. Exim used not to create the message log file for a message until the first
6520 delivery attempt. This could be confusing when incoming messages were held
6521 for policy or load reasons. The message log file is now created at the time
6522 the message is received, and an initial "Received" line is written to it.
6524 4. The automatically generated man page for command line options had a minor
6525 bug that caused no ill effects; however, a more serious problem was that
6526 the procedure for building the man page automatically didn't always
6527 operate. Consequently, release 4.20 contains an out-of-date version. This
6528 shouldn't happen again.
6530 5. When building Exim with embedded Perl support, the script that builds the
6531 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6532 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6534 6. The freeze_tell option was not being used for messages that were frozen on
6535 arrival, either by an ACL or by local_scan().
6537 7. Added the smtp_incomplete_transaction log selector.
6539 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6540 was accepting AUTH without a new EHLO.
6542 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6543 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6546 10. Logging of TCP/IP connections (when configured) now happens in the main
6547 daemon process instead of the child process, so that the TCP/IP connection
6548 count is more accurate (but it can never be perfect).
6550 11. The use of "drop" in a nested ACL was not being handled correctly in the
6551 outer ACL. Now, if condition failure induced by the nested "drop" causes
6552 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6553 or "require"), the connection is dropped.
6555 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6556 that yield "discard" can now be used with an "accept" or a "discard" verb,
6557 but an error is generated for any others (because I can't see a useful way
6558 to define what should happen).
6560 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6561 are now processed in the same way as lines in the Exim configuration file.
6562 In particular, continuation lines are supported.
6564 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6566 15. Added -ti meaning -t -i.
6568 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6569 domains, and warn by logging if others are found.
6571 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6572 returned from GIFCONF when Exim is trying to find the list of interfaces on
6573 a host. The code in os.c has been modified to copy each ifreq to an aligned
6574 structure in all cases.
6576 Also, in some cases, the returned ifreq's were being copied to a 'struct
6577 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6578 means the last couple of bytes of an IPv6 address could be chopped if the
6579 ifreq contained only a normal sockaddr (14 bytes storage).
6581 18. Named domain lists were not supported in the hosts_treat_as_local option.
6582 An entry such as +xxxx was not recognized, and was treated as a literal
6585 19. Ensure that header lines added by a DATA ACL are included in the reject log
6586 if the ACL subsequently rejects the message.
6588 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6589 MD5 (which is deprecated).
6591 21. When testing a filter file using -bf, Exim was writing a message when it
6592 took the sender from a "From " line in the message, but it was not doing so
6593 when it took $return_path from a Return-Path: header line. It now does.
6595 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6596 with a valid header line field name (a series of printing characters
6597 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6599 23. Changed "disc" in the source to "disk" to conform to the documentation and
6600 the book and for uniformity.
6602 24. Ignore Sendmail's -Ooption=value command line item.
6604 25. When execve() failed while trying to run a command in a pipe transport,
6605 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6606 could be confused with a return value of 69 from the command itself. This
6607 has been changed to 127, the value the shell returns if it is asked to run
6608 a non-existent command. The wording for the related log line suggests a
6609 non-existent command as the problem.
6611 26. If received_header_text expands to an empty string, do not add a Received:
6612 header line to the message. (Well, it adds a token one on the spool, but
6613 marks it "old" so that it doesn't get used or transmitted.)
6615 27. Installed eximstats 1.28 (addition of -nt option).
6617 28. There was no check for failure on the call to getsockname() in the daemon
6618 code. This can fail if there is a shortage of resources on the system, with
6619 ENOMEM, for example. A temporary error is now given on failure.
6621 29. Contrary to the C standard, it seems that in some environments, the
6622 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6623 program. Exim now does this explicitly; it affects the formatting of
6624 timestamps using strftime().
6626 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6627 complaints. I've now initialized all the variables, to avoid this.
6629 32. Header lines added by a system filter were not being "seen" during
6630 transport-time rewrites.
6632 33. The info_callback() function passed to OpenSSL is set up with type void
6633 (*)(SSL *, int, int), as described somewhere. However, when calling the
6634 function (actually a macro) that sets it up, the type void(*)() is
6635 expected. I've put in a cast to prevent warnings from picky compilers.
6637 34. If a DNS black list lookup found a CNAME record, but there were no A
6638 records associated with the domain it pointed at, Exim crashed.
6640 35. If a DNS black list lookup returned more than one A record, Exim ignored
6641 all but the first. It now scans all returned addresses if a particular IP
6642 value is being sought. In this situation, the contents of the
6643 $dnslist_value variable are a list of all the addresses, separated by a
6646 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6647 to accept a host name and all its aliases if the forward lookup for any of
6648 them yielded the IP address of the incoming connection. Now it accepts only
6649 those names whose forward lookup yields the correct IP address. Any other
6650 names are discarded. This closes a loophole whereby a rogue DNS
6651 administrator could create reverse DNS records to break through a
6652 wildcarded host restriction in an ACL.
6654 37. If a user filter or a system filter that ran in a subprocess used any of
6655 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6656 the wrong values were passed to the pipe command ($thisaddress had the
6657 value of $0, $0 had the value of $1, etc). This bug was introduced by
6658 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6660 38. Improved the line breaking for long SMTP error messages from ACLs.
6661 Previously, if there was no break point between 40 and 75 characters, Exim
6662 left the rest of the message alone. Two changes have been made: (a) I've
6663 reduced the minimum length to 35 characters; (b) if it can't find a break
6664 point between 35 and 75 characters, it looks ahead and uses the first one
6665 that it finds. This may give the occasional overlong line, but at least the
6666 remaining text gets split now.
6668 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6669 file descriptors might be low, and that setting 1000 would always raise it.
6670 It turns out that in some environments, the limit is already over 1000 and
6671 that lowering it causes trouble. So now Exim takes care not to decrease it.
6673 40. When delivering a message, the value of $return_path is set to $sender_
6674 address at the start of routing (routers may change the value). By an
6675 oversight, this default was not being set up when an address was tested by
6676 -bt or -bv, which affected the outcome if any router or filter referred to
6679 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6680 log only when "message" or "log_message" are set. However, if one of the
6681 conditions was an address verification, or a call to a nested ACL, the
6682 messages generated by the underlying test were being passed through. This
6683 no longer happens. The underlying message is available in $acl_verify_
6684 message for both "message" and "log_message" expansions, so it can be
6685 passed through if needed.
6687 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6688 new expansion $bh_ to give the encoded byte string without charset
6689 translation. Translation happens only if iconv() is available; HAVE_ICONV
6690 indicates this at build time. HEADERS_CHARSET gives the charset to
6691 translate to; headers_charset can change it in the configuration, and
6692 "headers charset" can change it in an individual filter file.
6694 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6695 that creates RFC 2047 encoded "words" labels them as that charset instead
6696 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6697 expansion operator; (ii) when Exim creates a From: line for a local
6698 message; (iii) when a header line is rewritten to include a "phrase" part.
6700 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6701 buggy, causing it to skip the first lines of messages whose message ID
6702 ended in 'D'. This would not have bitten before Exim release 4.14, because
6703 message IDs were unlikely to end in 'D' before then. The effect was to have
6704 incorrect size information for certain domains.
6706 45. #include "config.h" was missing at the start of the crypt16.c module. This
6707 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6710 46. If there was a timeout during a "random" callout check, Exim treated it as
6711 a failure of the random address, and carried on sending RSET and the real
6712 address. If the delay was just some slowness somewhere, the response to the
6713 original RCPT would be taken as a response to RSET and so on, causing
6714 mayhem of various kinds.
6716 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6717 when I implemented it. It didn't allow for the fact that some option values
6718 may legitimately be negative (e.g. size_addition), and it didn't even do
6719 the right test for positive values.
6721 48. Domain names in DNS records are case-independent. Exim always looks them up
6722 in lower case. Some resolvers return domain names in exactly the case they
6723 appear in the zone file, that is, they may contain uppercase letters. Not
6724 all resolvers do this - some return always lower case. Exim was treating a
6725 change of case by a resolver as a change of domain, similar to a widening
6726 of a domain abbreviation. This triggered its re-routing code and so it was
6727 trying to route what was effectively the same domain again. This normally
6728 caused routing to fail (because the router wouldn't handle the domain
6729 twice). Now Exim checks for this case specially, and just changes the
6730 casing of the domain that it ultimately uses when it transmits the message
6733 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6736 50. If a filter generated a file delivery with a non-absolute name (possible if
6737 no home directory exists for the router), the forbid_file option was not
6740 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6741 the existing equality matching.
6743 52. Exim was using ints instead of ino_t variables in some places where it was
6744 dealing with inode numbers.
6746 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6747 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6748 called TMPDIR, and if it finds it is different, it changes its value.
6750 54. The smtp_printf() function is now made available to local_scan() so
6751 additional output lines can be written before returning. There is also an
6752 smtp_fflush() function to enable the detection of a dropped connection.
6753 The variables smtp_input and smtp_batched_input are exported to
6756 55. Changed the default runtime configuration: the message "Unknown user"
6757 has been removed from the ACL, and instead placed on the localuser router,
6758 using the cannot_route_message feature. This means that any verification
6759 failures that generate their own messages won't get overridden. Similarly,
6760 the "Unrouteable address" message that was in the ACL for unverifiable
6761 relay addresses has also been removed.
6763 56. Added hosts_avoid_esmtp to the smtp transport.
6765 57. The exicyclog script was not checking for the esoteric option
6766 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6767 will work only if exicyclog is run under the appropriate euid.
6769 58. Following a discussion on the list, the rules by which Exim recognises line
6770 endings on incoming messages have been changed. The -dropcr and drop_cr
6771 options are now no-ops, retained only for backwards compatibility. The
6772 following line terminators are recognized: LF CRLF CR. However, special
6773 processing applies to CR:
6775 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6776 nor a local message in the state where . is a terminator.
6778 (ii) If a bare CR is encountered in a header line, an extra space is added
6779 after the line terminator so as not to end the header. The reasoning
6780 behind this is that bare CRs in header lines are most likely either
6781 to be mistakes, or people trying to play silly games.
6783 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6784 was being incorrectly given as 18 bytes larger than it should have been.
6785 This is a VOB (very old bug).
6787 60. This may never have affected anything current, but just in case it has:
6788 When the local host is found other than at the start of a list of hosts,
6789 the local host, those with the same MX, and any that follow, are discarded.
6790 When the list in question was part of a longer list of hosts, the following
6791 hosts (not currently being processed) were also being discarded. This no
6792 longer happens. I'm not sure if this situation could ever has previously
6795 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6798 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6799 header. This is something that is recommended in a new Internet Draft, and
6800 is something that is documented as being done by Sendmail. There are two
6801 possible values. For messages generated by the autoreply transport, Exim
6804 Auto-Submitted: auto-replied
6806 whereas for all other generated messages (e.g. bounces) it adds
6808 Auto-Submitted: auto-generated
6810 63. The "personal" condition in filters now includes a test for the
6811 Auto-Submitted: header. If it contains the string "auto-" the message it
6812 not considered personal.
6814 64. Added rcpt_include_affixes as a generic transport option.
6816 65. Added queue_only_override (default true).
6818 66. Added the syslog_duplication option.
6820 67. If what should have been the first header line of a message consisted of
6821 a space followed by a colon, Exim was mis-interpreting it as a header line.
6822 It isn't of course - it is syntactically invalid and should therefore be
6823 treated as the start of the message body. The misbehaviour could have
6824 caused a number of strange effects, including loss of data in subsequent
6825 header lines, and spool format errors.
6827 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6828 client host had authenticated. This control can now be exercised by an ACL
6829 for more flexibility.
6831 69. By default, callouts do not happen when testing with -bh. There is now a
6832 variant, -bhc, which does actually run the callout code, including
6833 consulting and updating the callout cache.
6835 70. Added support for saslauthd authentication, courtesy of Alexander
6838 71. If statvfs() failed on the spool or log directories while checking their
6839 size for availability, Exim confusingly gave the error "space shortage".
6840 Furthermore, in debugging mode it crashed with a floating point exception.
6841 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6842 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6843 serious problem, Exim now writes to the main and panic logs when this
6844 happens, with details of the failure. It then refuses to accept the
6845 incoming message, giving the message "spool directory problem" or "log
6846 directory problem" with a 421 code for SMTP messages.
6848 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6849 0, 1, and 2 exist, because some OS complain for execs without them (see
6850 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6851 descriptors. However, the code omitted to check that the open succeeded,
6852 causing mysterious errors if for some reason the permissions on /dev/null
6853 got screwed. Now Exim writes a message to the main and panic logs, and
6854 bombs out if it can't open /dev/null.
6856 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6857 interact so that it is all more flexible. It is supposed to remain
6858 backwards compatible. Also added extra_local_interfaces.
6860 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6861 to bomb out with an assertion failure - to the client this appears as a
6862 connection drop. This problem occurs in the part of the code that was taken
6863 from the Samba project. Fortunately, the assertion is in a very simple
6864 function, so I have fixed this by reproducing the function inline in the
6865 one place where it is called, and arranging for authentication to fail
6866 instead of killing the process with assert().
6868 75. The SPA client code was not working when the server requested OEM rather
6869 than Unicode encoding.
6871 76. Added code to make require_files with a specific uid setting more usable in
6872 the case where statting the file as root fails - usually a non-root-mounted
6873 NFS file system. When this happens and the failure is EACCES, Exim now
6874 forks a subprocess and does the per-uid checking as the relevant uid.
6876 77. Added process_log_path.
6878 78. If log_file_path was not explicitly set, a setting of check_log_space or
6879 check_log_inodes was ignored.
6881 79. If a space check for the spool or log partitions fails, the incident is now
6882 logged. Of course, in the latter case the data may get lost...
6884 80. Added the %p formatting code to string_format() so that it can be used to
6885 print addresses in debug_print(). Adjusted all the address printing in the
6886 debugging in store.c to use %p rather than %d.
6888 81. There was a concern that a line of code in smtp_in.c could overflow a
6889 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6890 initially expressed, the concern was not well-founded, because trailing
6891 spaces are removed early. However, if the trailing spaces were followed by
6892 a NULL, they did not get removed, so the overflow was possible. Two fixes
6895 (a) I re-wrote the offending code in a cleaner fashion.
6896 (b) If an incoming SMTP command contains a NULL character, it is rejected
6899 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6900 runs initgroups(), so that if the Exim user is in any additional groups,
6901 they will be used during message reception.
6907 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob