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
151 PP/07 GitHub PR 56: Apply MariaDB build fix.
152 Patch provided by Jaroslav Škarvada.
154 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
155 during Coverity cleanups [4.87 JH/47]
156 Diagnosis and fix provided by Michael Fischer v. Mollard.
162 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
163 than -2003 did; needs libidn2 in addition to libidn.
165 JH/02 The path option on a pipe transport is now expanded before use.
167 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
168 Patch provided by "Björn", documentation fix added too.
170 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
171 missing a wire-to-host endian conversion.
173 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
174 close after a BDAT command line could be taken as a following command,
175 giving a synch failure. Fix by only checking for synch immediately
176 before acknowledging the chunk.
178 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
179 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
180 macro. Patches provided by Josh Soref.
182 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
183 Previously we did not; the RFC seems ambiguous and VRFY is not listed
184 by IANA as a service extension. However, John Klensin suggests that we
187 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
188 the dkim code may be unix-mode line endings rather than smtp wire-format
189 CRLF, so prepend a CR to any bare LF.
191 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
192 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
194 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
195 the first were themselves being wrongly included in the feed into dkim
196 processing; with most chunk sizes in use this resulted in an incorrect
197 body hash calculated value.
199 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
200 DKIM signature block, for verification. Although advised against by
201 standards it is specifically not ruled illegal.
203 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
205 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
206 missing a body hash (the bh= tag).
208 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
209 It seems that HAProxy sends the Proxy Protocol information in clear and
210 only then does a TLS startup, so do the same.
212 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
213 TCP connections (such as for Spamd) unless the daemon successfully set
214 Fast Open mode on its listening sockets. This fixes breakage seen on
215 too-old kernels or those not configured for Fast Open, at the cost of
216 requiring both directions being enabled for TFO, and TFO never being used
217 by non-daemon-related Exim processes.
219 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
220 endings, at least on the first header line. Try to canonify any that get
221 past that check, despite the cost.
223 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
224 now limited to an arbitrary five deep, while parsing addresses with the
225 strip_excess_angle_brackets option enabled.
227 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
228 instead leave the unprompted TLS handshake in socket buffer for the
229 TLS library to consume.
231 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
233 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
235 JH/16 Drop variables when they go out of scope. Memory management drops a whole
236 region in one operation, for speed, and this leaves assigned pointers
237 dangling. Add checks run only under the testsuite which checks all
238 variables at a store-reset and panics on a dangling pointer; add code
239 explicitly nulling out all the variables discovered. Fixes one known
240 bug: a transport crash, where a dangling pointer for $sending_ip_address
241 originally assigned in a verify callout, is re-used.
243 PP/06 Drop '.' from @INC in various Perl scripts.
245 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
247 PP/08 Reduce a number of compilation warnings under clang; building with
248 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
249 should be warning-free.
251 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
253 HS/01 Fix portability problems introduced by PP/08 for platforms where
254 realloc(NULL) is not equivalent to malloc() [SunOS et al].
256 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
257 chunk. This allows us to accept broken chunked messages. We need a more
258 general solution here.
260 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
261 already-broken messages in the queue.
263 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
265 JH/19 Fix reference counting bug in routing-generated-address tracking.
271 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
272 supports it and a size is available (ie. the sending peer gave us one).
274 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
275 years of being deprecated). The replacements are the ACLs
276 acl_smtp_mime and acl_not_smtp_mime.
278 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
279 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
280 or even an in-clear connection were permitted. Now, if the host lookup
281 was dnssec and dane was requested then the host is only used if the
282 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
283 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
284 if one fails this test.
285 This means that a poorly-configured remote DNS will make it incommunicado;
286 but it protects against a DNS-interception attack on it.
288 JH/04 Bug 1810: make continued-use of an open smtp transport connection
289 non-noisy when a race steals the message being considered.
291 JH/05 If main configuration option tls_certificate is unset, generate a
292 self-signed certificate for inbound TLS connections.
294 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
295 in rewrites and routers.
297 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
298 and logged a warning sing 4.83; now they are a configuration file error.
300 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
301 (lacking @domain). Apply the same qualification processing as RCPT.
303 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
305 JH/10 Support ${sha256:} applied to a string (as well as the previous
308 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
309 a cutthrough deliver is pending, as we always want to make a connection.
310 This also avoids re-routing the message when later placing the cutthrough
311 connection after a verify cache hit.
312 Do not update it with the verify result either.
314 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
315 when routing results in more than one destination address.
317 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
318 signing (which inhibits the cutthrough capability). Previously only
319 the presence of an option was tested; now an expansion evaluating as
320 empty is permissible (obviously it should depend only on data available
321 when the cutthrough connection is made).
323 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
324 the relevant preceding SMTP command did not note the pipelining mode.
326 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
327 Previously they were not counted.
329 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
330 as one having no matching records. Previously we deferred the message
331 that needed the lookup.
333 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
334 distinguished as "(=".
336 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
337 for missing MX records. Previously it only worked for missing A records.
339 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
341 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
342 after the data-go-ahead and data-ack. Patch from Jason Betts.
344 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
345 even for a "none" policy. Patch from Tony Meyer.
347 JH/22 Fix continued use of a connection for further deliveries. If a port was
348 specified by a router, it must also match for the delivery to be
351 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
352 When one of the recipients of a message was unsuitable for the connection
353 (has no matching addresses), we lost track of needing to mark it
354 deferred. As a result mail would be lost.
356 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
358 JH/25 Decoding ACL controls is now done using a binary search; the source code
359 takes up less space and should be simpler to maintain. Merge the ACL
360 condition decode tables also, with similar effect.
362 JH/26 Fix problem with one_time used on a redirect router which returned the
363 parent address unchanged. A retry would see the parent address marked as
364 delivered, so not attempt the (identical) child. As a result mail would
367 JH/27 Fix a possible security hole, wherein a process operating with the Exim
368 UID can gain a root shell. Credit to http://www.halfdog.net/ for
369 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
372 JH/28 Enable {spool,log} filesystem space and inode checks as default.
373 Main config options check_{log,spool}_{inodes,space} are now
374 100 inodes, 10MB unless set otherwise in the configuration.
376 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
377 Previously it only applied to the main-section connection policy
380 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
382 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
383 by me. Added RFC7919 DH primes as an alternative.
385 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
386 are not in the system include path.
388 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
389 GnuTLS, when a session startup failed (eg because the client disconnected)
390 Exim did stdio operations after fclose. This was exposed by a recent
391 change which nulled out the file handle after the fclose.
393 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
394 signed directly by the cert-signing cert, rather than an intermediate
395 OCSP-signing cert. This is the model used by LetsEncrypt.
397 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
399 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
400 an incoming connection.
402 HS/02 Bug 1802: Do not half-close the connection after sending a request
405 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
406 fallback to "prime256v1".
408 JH/34 SECURITY: Use proper copy of DATA command in error message.
409 Could leak key material. Remotely exploitable. CVE-2016-9963.
415 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
416 and 3.4.4 - once the server is enabled to respond to an OCSP request
417 it does even when not requested, resulting in a stapling non-aware
418 client dropping the TLS connection.
420 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
421 support variable-length bit vectors. No functional change.
423 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
424 The I= interface field on outgoing lines is now after the H= remote
425 host field, same as incoming lines. There is a separate
426 outgoing_interface log selector which allows you to disable the
429 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
430 If not running log_selector +smtp_connection the mainlog would be held
431 open indefinitely after a "too many connections" event, including to a
432 deleted file after a log rotate. Leave the per net connection logging
433 leaving it open for efficiency as that will be quickly detected by the
434 check on the next write.
436 HS/01 Bug 1671: Fix post transport crash.
437 Processing the wait-<transport> messages could crash the delivery
438 process if the message IDs didn't exist for some reason. When
439 using 'split_spool_directory=yes' the construction of the spool
440 file name failed already, exposing the same netto behaviour.
442 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
443 mime_regex ACL conditions.
445 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
446 to DSN fail messages (bounces): remote IP, remote greeting, remote response
447 to HELO, local diagnostic string.
449 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
450 log line to debug. Even when configured with a tls authenticator many
451 client connections are expected to not authenticate in this way, so
452 an authenticate fail is not an error.
454 HS/02 Add the Exim version string to the process info. This way exiwhat
455 gives some more detail about the running daemon.
457 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
458 matter for fast-change records such as DNSBLs.
460 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
461 retry record, even if constant. There may be multiple transports with
462 different interface settings and the retry behaviour needs to be kept
465 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
467 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
469 JH/10 Bug 840: fix log_defer_output option of pipe transport
471 JH/11 Bug 830: use same host for all RCPTS of a message, even under
472 hosts_randomize. This matters a lot when combined with mua_wrapper.
474 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
475 ${quote_pgsql:<string>} operator.
477 JH/13 Bug 1708: avoid misaligned access in cached lookup.
479 JH/14 Change header file name for freeradius-client. Relevant if compiling
480 with Radius support; from the Gentoo tree and checked under Fedora.
482 JH/15 Bug 1712: Introduce $prdr_requested flag variable
484 JH/16 Bug 1714: Permit an empty string as expansion result for transport
485 option transport_filter, meaning no filtering.
487 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
489 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
490 defaults to "*" (all hosts). The variable is now available when not built
491 with TLS, default unset, mainly to enable keeping the testsuite sane.
492 If a server certificate is not supplied (via tls_certificate) an error is
493 logged, and clients will find TLS connections fail on startup. Presumably
494 they will retry in-clear.
495 Packagers of Exim are strongly encouraged to create a server certificate
496 at installation time.
498 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
499 with the $config_file variable.
501 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
502 in transport context, after the attempt, and per-recipient. The latter type
503 is per host attempted. The event data is the error message, and the errno
504 information encodes the lookup type (A vs. MX) used for the (first) host,
505 and the trailing two digits of the smtp 4xx response.
507 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
508 to write to mainlog (or rejectlog, paniclog) in the window between file
509 creation and permissions/ownership being changed. Particularly affects
510 installations where exicyclog is run as root, rather than exim user;
511 result is that the running daemon panics and dies.
513 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
515 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
516 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
517 "pri" and "weight". Note that the previous implicit priority given by the
518 list order is no longer honoured.
520 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
523 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
524 by defining SUPPORT_SOCKS.
526 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
527 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
528 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
529 variables are renamed to proxy_{local,external}_{address,port}.
531 JH/27 Move Internationalisation support from Experimental to mainline, enabled
532 for a build by defining SUPPORT_I18N
534 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
535 of the query string, and make ${quote_redis:} do that quoting.
537 JH/29 Move Events support from Experimental to mainline, enabled by default
538 and removable for a build by defining DISABLE_EVENT.
540 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
542 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
543 cached by the daemon.
545 JH/32 Move Redis support from Experimental to mainline, enabled for a build
546 by defining LOOKUP_REDIS. The libhiredis library is required.
548 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
549 keys are given for lookup.
551 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
552 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
553 only supported when built with TLS support. The PolarSSL SHA routines
554 are still used when the TLS library is too old for convenient support.
556 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
557 openssl_options), for security. OpenSSL forces this from version 1.1.0
558 server-side so match that on older versions.
560 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
561 allocation for $value could be released as the expansion processing
562 concluded, but leaving the global pointer active for it.
564 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
565 and to use the domains and local_parts ACL conditions.
567 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
568 incorrectly not doubled on cutthrough transmission, hence seen as a
569 body-termination at the receiving system - resulting in truncated mails.
570 Commonly the sender saw a TCP-level error, and retransmitted the message
571 via the normal store-and-forward channel. This could result in duplicates
572 received - but deduplicating mailstores were liable to retain only the
573 initial truncated version.
575 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
577 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
579 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
580 we're in there, support oversigning also; bug 1309.
582 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
584 HS/04 Add support for keep_environment and add_environment options.
586 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
587 either intentional arithmetic overflow during PRNG, or testing config-
590 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
591 delivery resulted in actual delivery. Cancel cutthrough before DATA
594 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
595 rejects a recipient: pass the reject to the originator.
597 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
598 Many were false-positives and ignorable, but it's worth fixing the
601 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
602 for the new environment-manipulation done at startup. Move the routines
603 from being local to tls.c to being global via the os.c file.
605 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
606 an extract embedded as result-arg for a map, the first arg for extract
607 is unavailable so we cannot tell if this is a numbered or keyed
608 extraction. Accept either.
614 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
617 JH/02 The smtp transport option "multi_domain" is now expanded.
619 JH/03 The smtp transport now requests PRDR by default, if the server offers
622 JH/04 Certificate name checking on server certificates, when exim is a client,
623 is now done by default. The transport option tls_verify_cert_hostnames
624 can be used to disable this per-host. The build option
625 EXPERIMENTAL_CERTNAMES is withdrawn.
627 JH/05 The value of the tls_verify_certificates smtp transport and main options
628 default to the word "system" to access the system default CA bundle.
629 For GnuTLS, only version 3.0.20 or later.
631 JH/06 Verification of the server certificate for a TLS connection is now tried
632 (but not required) by default. The verification status is now logged by
633 default, for both outbound TLS and client-certificate supplying inbound
636 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
639 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
640 Status Notification (bounce) messages are now MIME format per RFC 3464.
641 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
642 under the control of the dsn_advertise_hosts option, and routers may
643 have a dsn_lasthop option.
645 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
646 default, modifiable by a malware= option. The list separator for
647 the options can now be changed in the usual way. Bug 68.
649 JH/10 The smtp_receive_timeout main option is now expanded before use.
651 JH/11 The incoming_interface log option now also enables logging of the
652 local interface on delivery outgoing connections.
654 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
655 if the interface and destination host and port all match.
657 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
660 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
661 Patch from Andrew Lewis.
663 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
664 now supports optional time-restrictions, weighting, and priority
665 modifiers per server. Patch originally by <rommer@active.by>.
667 JH/16 The spamd_address main option now supports a mixed list of local
668 and remote servers. Remote servers can be IPv6 addresses, and
669 specify a port-range.
671 JH/17 Bug 68: The spamd_address main option now supports an optional
672 timeout value per server.
674 JH/18 Bug 1581: Router and transport options headers_add/remove can
675 now have the list separator specified.
677 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
680 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
683 JH/21 Support for the A6 type of dns record is withdrawn.
685 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
686 rather than the verbs used.
688 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
689 from 255 to 1024 chars.
691 JH/24 Verification callouts now attempt to use TLS by default.
693 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
694 are generic router options now. The defaults didn't change.
696 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
697 Original patch from Alexander Shikoff, worked over by JH.
699 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
700 files if ZCAT_COMMAND is not executable.
702 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
704 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
706 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
707 Normally benign, it bites when the pair was led to by a CNAME;
708 modern usage is to not canonicalize the domain to a CNAME target
709 (and we were inconsistent anyway for A-only vs AAAA+A).
711 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
713 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
714 when evaluating $sender_host_dnssec.
716 JH/31 Check the HELO verification lookup for DNSSEC, adding new
717 $sender_helo_dnssec variable.
719 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
721 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
723 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
725 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
726 documented as working, but never had. Support all but $spam_report.
728 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
729 added for tls authenticator.
731 HS/03 Add perl_taintmode main config option
737 TL/01 When running the test suite, the README says that variables such as
738 no_msglog_check are global and can be placed anywhere in a specific
739 test's script, however it was observed that placement needed to be near
740 the beginning for it to behave that way. Changed the runtest perl
741 script to read through the entire script once to detect and set these
742 variables, reset to the beginning of the script, and then run through
743 the script parsing/test process like normal.
745 TL/02 The BSD's have an arc4random API. One of the functions to induce
746 adding randomness was arc4random_stir(), but it has been removed in
747 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
748 function when detected.
750 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
751 cause callback expansion.
753 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
754 syntax errors in an expansion can be treated as a string instead of
755 logging or causing an error, due to the internal use of bool_lax
756 instead of bool when processing it.
758 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
759 server certificates when making smtp deliveries.
761 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
763 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
765 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
767 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
768 Merged patch from Sebastian Wiedenroth.
770 JH/05 Fix results-pipe from transport process. Several recipients, combined
771 with certificate use, exposed issues where response data items split
772 over buffer boundaries were not parsed properly. This eventually
773 resulted in duplicates being sent. This issue only became common enough
774 to notice due to the introduction of connection certificate information,
775 the item size being so much larger. Found and fixed by Wolfgang Breyha.
777 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
778 size buffer was used, resulting in syntax errors when an expansion
781 JH/07 Add support for directories of certificates when compiled with a GnuTLS
782 version 3.3.6 or later.
784 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
785 is EXPERIMENTAL_EVENT, the main-configuration and transport options
786 both become "event_action", the variables become $event_name, $event_data
787 and $event_defer_errno. There is a new variable $verify_mode, usable in
788 routers, transports and related events. The tls:cert event is now also
789 raised for inbound connections, if the main configuration event_action
792 TL/06 In test suite, disable OCSP for old versions of openssl which contained
793 early OCSP support, but no stapling (appears to be less than 1.0.0).
795 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
796 server certificate names available under the smtp transport option
797 "tls_verify_cert_hostname" now do not permit multi-component wildcard
800 JH/10 Time-related extraction expansions from certificates now use the main
801 option "timezone" setting for output formatting, and are consistent
802 between OpenSSL and GnuTLS compilations. Bug 1541.
804 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
805 encoded parameter in the incoming message. Bug 1558.
807 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
808 include certificate info, eximon was claiming there were spoolfile
811 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
813 JH/14 Log delivery-related information more consistently, using the sequence
814 "H=<name> [<ip>]" wherever possible.
816 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
817 are problematic for Debian distribution, omit them from the release
820 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
822 JH/16 Fix string representation of time values on 64bit time_t architectures.
825 JH/17 Fix a null-indirection in certextract expansions when a nondefault
826 output list separator was used.
831 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
832 checkers that were complaining about end of non-void function with no
835 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
836 This was a regression introduced in 4.83 by another bugfix.
838 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
840 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
841 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
847 TF/01 Correctly close the server side of TLS when forking for delivery.
849 When a message was received over SMTP with TLS, Exim failed to clear up
850 the incoming connection properly after forking off the child process to
851 deliver the message. In some situations the subsequent outgoing
852 delivery connection happened to have the same fd number as the incoming
853 connection previously had. Exim would try to use TLS and fail, logging
854 a "Bad file descriptor" error.
856 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
857 utilities have not been installed.
859 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
860 temporary space as the ACL may create new global variables.
862 TL/01 LDAP support uses per connection or global context settings, depending
863 upon the detected version of the libraries at build time.
865 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
866 to extract and use the src ip:port in logging and expansions as if it
867 were a direct connection from the outside internet. PPv2 support was
868 updated based on HAProxy spec change in May 2014.
870 JH/02 Add ${listextract {number}{list}{success}{fail}}.
872 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
873 Properly escape header and check for NULL return.
875 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
878 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
880 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
881 characters in header names, implemented as a verify condition.
882 Contributed by Michael Fischer v. Mollard.
884 TL/05 Rename SPF condition results err_perm and err_temp to standardized
885 results permerror and temperror. Previous values are deprecated but
886 still accepted. In a future release, err_perm and err_temp will be
887 completely removed, which will be a backward incompatibility if the
888 ACL tests for either of these two old results. Patch contributed by
889 user bes-internal on the mailing list.
891 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
893 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
894 selectors, in both main and reject logs.
896 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
899 JH/07 Add malware type "sock" for talking to simple daemon.
901 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
903 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
904 routers/transports under cutthrough routing.
906 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
907 numbers. Touch up "bool" conditional to keep the same definition.
909 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
911 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
913 JH/12 Expand items in router/transport headers_add or headers_remove lists
914 individually rather than the list as a whole. Bug 1452.
916 Required for reasonable handling of multiple headers_ options when
917 they may be empty; requires that headers_remove items with embedded
918 colons must have them doubled (or the list-separator changed).
920 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
921 view the policy declared in the DMARC record. Currently, $dmarc_status
922 is a combined value of both the record presence and the result of the
925 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
927 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
928 dnslookup router and the smtp transport (applying to the forward
931 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
932 of ldap servers used for a specific lookup. Patch provided by Heiko
935 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
936 New variable $lookup_dnssec_authenticated for observability.
938 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
939 Patch submitted by Lars Timman.
941 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
943 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
944 Requires trusted mode and valid format message id, aborts otherwise.
945 Patch contributed by Heiko Schlichting.
947 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
948 certextract with support for various fields. Bug 1358.
950 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
951 is requested by default, modifiable by smtp transport option
954 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
955 operate on certificate variables to give certificate fingerprints
956 Also new ${sha256:cert_variable}.
958 JH/23 The PRDR feature is moved from being Experimental into the mainline.
960 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
961 Christian Aistleitner.
963 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
965 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
966 file. Patch from Wolfgang Breyha.
968 JH/25 Expand the coverage of the delivery $host and $host_address to
969 client authenticators run in verify callout. Bug 1476.
971 JH/26 Port service names are now accepted for tls_on_connect_ports, to
972 align with daemon_smtp_ports. Bug 72.
974 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
975 support and error reporting did not work properly.
977 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
978 and is readable. Patch from Andrew Colin Kissa.
980 TL/14 Enhance documentation of ${run expansion and how it parses the
981 commandline after expansion, particularly in the case when an
982 unquoted variable expansion results in an empty value.
984 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
986 PP/02 Fix internal collision of T_APL on systems which support RFC3123
987 by renaming away from it. Addresses GH issue 15, reported by
990 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
992 TL/15 SECURITY: prevent double expansion in math comparison functions
993 (can expand unsanitized data). Not remotely exploitable.
1000 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1002 PP/02 Make -n do something, by making it not do something.
1003 When combined with -bP, the name of an option is not output.
1005 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1008 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1009 $sender_host_name and config options to manage this, and basic check
1012 PP/05 DSCP support for outbound connections and control modifier for inbound.
1014 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1015 (Only plugin which currently uses this is kerberos4, which nobody should
1016 be using, but we should make it available and other future plugins might
1017 conceivably use it, even though it would break NAT; stuff *should* be
1018 using channel bindings instead).
1020 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1021 name; added for Sendmail compatibility; requires admin caller.
1022 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1023 just ignore it); requires trusted caller.
1024 Also parse but ignore: -Ac -Am -X<logfile>
1027 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1029 TL/02 Add +smtp_confirmation as a default logging option.
1031 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1032 Patch by Magnus Holmgren from 2007-02-20.
1034 TL/04 Bugzilla 1281 - Spec typo.
1035 Bugzilla 1283 - Spec typo.
1036 Bugzilla 1290 - Spec grammar fixes.
1038 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1040 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1042 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1043 mailing list by by Dmitry Isaikin.
1045 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1047 JH/02 Support "G" suffix to numbers in ${if comparisons.
1049 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1051 NM/01 Bugzilla 1197 - Spec typo
1052 Bugzilla 1196 - Spec examples corrections
1054 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1056 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1057 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1060 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1061 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1063 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1064 "acl {{name}{arg}...}", and optional args on acl condition
1067 JH/05 Permit multiple router/transport headers_add/remove lines.
1069 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1071 JH/07 Avoid using a waiting database for a single-message-only transport.
1072 Performance patch from Paul Fisher. Bugzilla 1262.
1074 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1077 JH/09 Add $headers_added variable, with content from use of ACL modifier
1078 add_header (but not yet added to the message). Bugzilla 199.
1080 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1081 Pulled from Bugzilla 817 by Wolfgang Breyha.
1083 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1085 (nb: this is the same fix as in Exim 4.80.1)
1087 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1090 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1093 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1095 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1096 advertise SMTP AUTH mechanism to us, instead of a generic
1097 protocol violation error. Also, make Exim more robust to bad
1098 data from the Dovecot auth socket.
1100 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1102 When a queue runner is handling a message, Exim first routes the
1103 recipient addresses, during which it prunes them based on the retry
1104 hints database. After that it attempts to deliver the message to
1105 any remaining recipients. It then updates the hints database using
1108 So if a recipient address works intermittently, it can get repeatedly
1109 deferred at routing time. The retry hints record remains fresh so the
1110 address never reaches the final cutoff time.
1112 This is a fairly common occurrence when a user is bumping up against
1113 their storage quota. Exim had some logic in its local delivery code
1114 to deal with this. However it did not apply to per-recipient defers
1115 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1117 This change adds a proper retry rule check during routing so that the
1118 final cutoff time is checked against the message's age. We only do
1119 this check if there is an address retry record and there is not a
1120 domain retry record; this implies that previous attempts to handle
1121 the address had the retry_use_local_parts option turned on. We use
1122 this as an approximation for the destination being like a local
1123 delivery, as in LMTP.
1125 I suspect this new check makes the old local delivery cutoff check
1126 redundant, but I have not verified this so I left the code in place.
1128 TF/02 Correct gecos expansion when From: is a prefix of the username.
1130 Test 0254 submits a message to Exim with the header
1134 When I ran the test suite under the user fanf2, Exim expanded
1135 the header to contain my full name, whereas it should have added
1136 a Resent-Sender: header. It erroneously treats any prefix of the
1137 username as equal to the username.
1139 This change corrects that bug.
1141 GF/01 DCC debug and logging tidyup
1142 Error conditions log to paniclog rather than rejectlog.
1143 Debug lines prefixed by "DCC: " to remove any ambiguity.
1145 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1147 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1148 Bug spotted by Jeremy Harris; was flawed since initial commit.
1149 Would have resulted in OCSP responses post-SNI triggering an Exim
1150 NULL dereference and crash.
1152 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1154 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1155 Bug detection, analysis and fix by Samuel Thibault.
1156 Bugzilla 1331, Debian bug #698092.
1158 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1160 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1161 Server implementation by Todd Lyons, client by JH.
1162 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1163 config variable "prdr_enable" controls whether the server
1164 advertises the facility. If the client requests PRDR a new
1165 acl_data_smtp_prdr ACL is called once for each recipient, after
1166 the body content is received and before the acl_smtp_data ACL.
1167 The client is controlled by both of: a hosts_try_prdr option
1168 on the smtp transport, and the server advertisement.
1169 Default client logging of deliveries and rejections involving
1170 PRDR are flagged with the string "PRDR".
1172 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1173 fclose(). Diagnosis by Todd Lyons.
1175 PP/17 Update configure.default to handle IPv6 localhost better.
1176 Patch by Alain Williams (plus minor tweaks).
1179 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1180 This is now consistent with GnuTLS, and is now documented: the
1181 previous undocumented portable approach to treating the option as
1182 unset was to force an expansion failure. That still works, and
1183 an empty string is now equivalent.
1185 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1186 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1187 not performing validation itself.
1189 PP/20 Added force_command boolean option to pipe transport.
1190 Patch from Nick Koston, of cPanel Inc.
1192 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1195 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1197 PP/21 Fix eximon continuous updating with timestamped log-files.
1198 Broken in a format-string cleanup in 4.80, missed when I repaired the
1199 other false fix of the same issue.
1200 Report and fix from Heiko Schlichting.
1203 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1204 Report from Prashanth Katuri.
1206 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1207 It's SecureTransport, so affects any MacOS clients which use the
1208 system-integrated TLS libraries, including email clients.
1210 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1211 using a MIME ACL for non-SMTP local injection.
1212 Report and assistance in diagnosis by Warren Baker.
1214 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1216 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1218 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1219 last id that failed so it may be referenced in subsequent ACL's.
1221 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1224 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1225 ldap library initialization, allowing self-signed CA's to be
1226 used. Also properly sets require_cert option later in code by
1227 using NULL (global ldap config) instead of ldap handle (per
1228 session). Bug diagnosis and testing by alxgomz.
1230 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1231 the src/util/ subdirectory.
1233 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1234 renamed to Transport Post Delivery Action by Jeremy Harris, as
1237 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1238 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1239 redis_servers = needs to be configured which will be used by the redis
1240 lookup. Patch from Warren Baker, of The Packet Hub.
1242 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1244 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1245 hostname or reverse DNS when processing a host list. Used suggestions
1246 from multiple comments on this bug.
1248 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1250 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1251 Missed a few lines, added it to make the runtest require no keyboard
1254 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1255 contains upper case chars. Make router use caseful_local_part.
1257 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1258 support when GnuTLS has been built with p11-kit.
1264 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1266 This, or similar/improved, will also be change PP/11 of 4.82.
1272 PP/01 Handle short writes when writing local log-files.
1273 In practice, only affects FreeBSD (8 onwards).
1274 Bugzilla 1053, with thanks to Dmitry Isaikin.
1276 NM/01 Bugzilla 949 - Documentation tweak
1278 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1281 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1283 PP/02 Implemented gsasl authenticator.
1285 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1287 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1288 `pkg-config foo` for cflags/libs.
1290 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1291 with rest of GSASL and with heimdal_gssapi.
1293 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1294 `pkg-config foo` for cflags/libs for the TLS implementation.
1296 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1297 properties get this fed in as external SSF. A number of robustness
1298 and debugging improvements to the cyrus_sasl authenticator.
1300 PP/08 cyrus_sasl server now expands the server_realm option.
1302 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1303 Patch by Jeremy Harris.
1305 PP/10 Added dbmjz lookup type.
1307 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1309 PP/12 MAIL args handles TAB as well as SP, for better interop with
1310 non-compliant senders.
1311 Analysis and variant patch by Todd Lyons.
1313 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1314 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1315 Patch from Dirk Mueller <dmueller@suse.com>
1317 PP/13 tls_peerdn now print-escaped for spool files.
1318 Observed some $tls_peerdn in wild which contained \n, which resulted
1319 in spool file corruption.
1321 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1322 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1323 or write after TLS renegotiation, which otherwise led to messages
1326 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1327 as a tracking header (ie: a signed header comes before the signature).
1328 Patch from Wolfgang Breyha.
1330 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1331 comma-sep list; embedded commas doubled.
1333 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1335 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1337 Report and patch from Dmitry Banschikov.
1339 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1340 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1341 failures. We appear to now support TLS1.1+ with Exim.
1343 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1344 lets Exim select keys and certificates based upon TLS SNI from client.
1345 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1346 before an outbound SMTP session. New log_selector, +tls_sni.
1348 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1349 NULL dereference. Report and patch from Alun Jones.
1351 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1352 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1353 Not seeing resolver debug output on NetBSD, but suspect this is a
1354 resolver implementation change.
1356 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1357 Left warnings. Added "eximon gdb" invocation mode.
1359 PP/21 Defaulting "accept_8bitmime" to true, not false.
1361 PP/22 Added -bw for inetd wait mode support.
1363 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1364 locate the relevant includes and libraries. Made this the default.
1366 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1367 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1369 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1370 This may cause build issues on older platforms.
1372 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1373 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1374 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1375 Added SNI support via GnuTLS too.
1376 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1378 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1380 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1381 Applied second patch from Janne, implementing suggestion to default
1382 multiple-strings-in-record handling to match SPF spec.
1384 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1386 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1387 read-only, out of scope).
1388 Patch from Wolfgang Breyha, report from Stuart Northfield.
1390 PP/29 Fix three issues highlighted by clang analyser static analysis.
1391 Only crash-plausible issue would require the Cambridge-specific
1392 iplookup router and a misconfiguration.
1393 Report from Marcin Mirosław.
1395 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1397 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1398 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1399 As part of this, removing so much warning spew let me fix some minor
1400 real issues in debug logging.
1402 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1403 assignment on my part. Fixed.
1405 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1406 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1407 Janne Snabb (who went above and beyond: thank you).
1409 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1410 string otherwise requires a connection and a bunch more work and it's
1411 relatively easy to get wrong. Should also expose TLS library linkage
1414 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1415 64-bit ${eval} (JH/03).
1417 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1418 GNU libc to support some of the 64-bit stuff, should not lead to
1419 conflicts. Defined before os.h is pulled in, so if a given platform
1420 needs to override this, it can.
1422 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1423 protection layer was required, which is not implemented.
1424 Bugzilla 1254, patch from Wolfgang Breyha.
1426 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1427 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1428 tls_dhparam take prime identifiers. Also unbreak combination of
1429 OpenSSL+DH_params+TLSSNI.
1431 PP/39 Disable SSLv2 by default in OpenSSL support.
1437 PP/01 Solaris build fix for Oracle's LDAP libraries.
1438 Bugzilla 1109, patch from Stephen Usher.
1440 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1442 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1445 TF/02 Fix a couple more cases where we did not log the error message
1446 when unlink() failed. See also change 4.74-TF/03.
1448 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1449 lock up or crash if it happened to be inside a call to libc when it
1450 got a SIGUSR1 from exiwhat.
1452 The SIGUSR1 handler appends the current process status to the process
1453 log which is later printed by exiwhat. It used to use the general
1454 purpose logging code to do this, but several functions it calls are
1455 not safe for signals.
1457 The new output code in the SIGUSR1 handler is specific to the process
1458 log, and simple enough that it's easy to inspect for signal safety.
1459 Removing some special cases also simplifies the general logging code.
1460 Removing the spurious timestamps from the process log simplifies
1463 TF/04 Improved ratelimit ACL condition.
1465 The /noupdate option has been deprecated in favour of /readonly which
1466 has clearer semantics. The /leaky, /strict, and /readonly update modes
1467 are mutually exclusive. The update mode is no longer included in the
1468 database key; it just determines when the database is updated. (This
1469 means that when you upgrade Exim will forget old rate measurements.)
1471 Exim now checks that the per_* options are used with an update mode that
1472 makes sense for the current ACL. For example, when Exim is processing a
1473 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1474 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1475 must specify per_mail/readonly. If you omit the update mode it defaults to
1476 /leaky where that makes sense (as before) or /readonly where required.
1478 The /noupdate option is now undocumented but still supported for
1479 backwards compatibility. It is equivalent to /readonly except that in
1480 ACLs where /readonly is required you may specify /leaky/noupdate or
1481 /strict/noupdate which are treated the same as /readonly.
1483 A useful new feature is the /count= option. This is a generalization
1484 of the per_byte option, so that you can measure the throughput of other
1485 aggregate values. For example, the per_byte option is now equivalent
1486 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1488 The per_rcpt option has been generalized using the /count= mechanism
1489 (though it's more complicated than the per_byte equivalence). When it is
1490 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1491 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1492 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1493 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1494 Note that using per_rcpt with a non-readonly update mode in more than
1495 one ACL will cause the recipients to be double-counted. (The per_mail
1496 and per_byte options don't have this problem.)
1498 The handling of very low rates has changed slightly. If the computed rate
1499 is less than the event's count (usually one) then this event is the first
1500 after a long gap. In this case the rate is set to the same as this event's
1501 count, so that the first message of a spam run is counted properly.
1503 The major new feature is a mechanism for counting the rate of unique
1504 events. The new per_addr option counts the number of different
1505 recipients that someone has sent messages to in the last time period. It
1506 behaves like per_rcpt if all the recipient addresses are different, but
1507 duplicate recipient addresses do not increase the measured rate. Like
1508 the /count= option this is a general mechanism, so the per_addr option
1509 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1510 example, measure the rate that a client uses different sender addresses
1511 with the options per_mail/unique=$sender_address. There are further
1512 details in the main documentation.
1514 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1516 TF/06 Removed a few PCRE remnants.
1518 TF/07 Automatically extract Exim's version number from tags in the git
1519 repository when doing development or release builds.
1521 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1522 Bugzilla 879. Patch from Paul Fisher.
1524 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1525 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1528 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1530 PP/05 Variable $av_failed, true if the AV scanner deferred.
1531 Bugzilla 1078. Patch from John Horne.
1533 PP/06 Stop make process more reliably on build failure.
1534 Bugzilla 1087. Patch from Heiko Schlittermann.
1536 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1537 Bugzilla 1089. Patch from Heiko Schlittermann.
1539 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1540 Bugzilla 1131. Patch from Holger Weiß.
1542 PP/09 Handle IPv6 addresses with SPF.
1543 Bugzilla 860. Patch from Wolfgang Breyha.
1545 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1547 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1550 PP/11 match_* no longer expand right-hand-side by default.
1551 New compile-time build option, EXPAND_LISTMATCH_RHS.
1552 New expansion conditions, "inlist", "inlisti".
1554 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1556 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1558 PP/14 fix log_write() format string regression from TF/03.
1559 Bugzilla 1152. Patch from Dmitry Isaikin.
1565 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1567 PP/02 Harmonised TLS library version reporting; only show if debugging.
1568 Layout now matches that introduced for other libraries in 4.74 PP/03.
1570 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1572 PP/04 New "dns_use_edns0" global option.
1574 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1577 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1578 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1580 TK/01 Updated PolarSSL code to 0.14.2.
1581 Bugzilla 1097. Patch from Andreas Metzler.
1583 PP/07 Catch divide-by-zero in ${eval:...}.
1584 Fixes bugzilla 1102.
1586 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1589 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1590 format-string attack -- SECURITY: remote arbitrary code execution.
1592 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1593 time unintentionally subject to list matching rules, letting the header
1594 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1595 arbitrary string expansion). This allowed for information disclosure.
1597 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1598 INT_MIN/-1 -- value coerced to INT_MAX.
1604 NM/01 Workaround for PCRE version dependency in version reporting
1607 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1608 This fixes portability to compilers other than gcc, notably
1609 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1611 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1612 makefiles for portability to HP-UX and POSIX correctness.
1614 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1615 Also via indented variable definition in the Makefile.
1616 (Debugging by Oliver Heesakkers).
1618 PP/02 Restore caching of spamd results with expanded spamd_address.
1619 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1621 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1622 Improves build reliability. Fix from: Frank Elsner
1624 NM/02 Fix wide character breakage in the rfc2047 coding
1625 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1627 NM/03 Allow underscore in dnslist lookups
1628 Fixes bug 1026. Patch from Graeme Fowler
1630 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1631 Code patches from Adam Ciarcinski of NetBSD.
1633 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1636 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1637 is logged, to avoid truncation. Patch from John Horne.
1639 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1640 Patch from Jakob Hirsch.
1642 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1643 SQL string expansion failure details.
1644 Patch from Andrey Oktyabrski.
1646 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1647 Patch from Simon Arlott.
1649 PP/09 New lookups functionality failed to compile on old gcc which rejects
1650 extern declarations in function scope.
1651 Patch from Oliver Fleischmann
1653 PP/10 Use sig_atomic_t for flags set from signal handlers.
1654 Check getgroups() return and improve debugging.
1655 Fixed developed for diagnosis in bug 927 (which turned out to be
1658 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1659 Patch from Mark Zealey.
1661 PP/12 Bugzilla 1056: Improved spamd server selection.
1662 Patch from Mark Zealey.
1664 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1665 Based on patch from Heiko Schlittermann.
1667 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1668 Patch from Uwe Doering, sign-off by Michael Haardt.
1670 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1671 variable declaration deep within a block. Bug and patch from
1674 PP/15 lookups-Makefile IRIX compatibility coercion.
1676 PP/16 Make DISABLE_DKIM build knob functional.
1678 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1679 Patch by Simon Arlott
1681 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1682 variable argument macros. Our copy now differs from upstream.
1688 TF/01 Failure to get a lock on a hints database can have serious
1689 consequences so log it to the panic log.
1691 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1692 controlled using the smtp_confirmation log selector.
1694 TF/03 Include the error message when we fail to unlink a spool file.
1696 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1697 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1698 for maintaining out-of-tree patches for some time.
1700 PP/01 Bugzilla 139: Documentation and portability issues.
1701 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1702 Handle per-OS dynamic-module compilation flags.
1704 PP/02 Let /dev/null have normal permissions.
1705 The 4.73 fixes were a little too stringent and complained about the
1706 permissions on /dev/null. Exempt it from some checks.
1707 Reported by Andreas M. Kirchwitz.
1709 PP/03 Report version information for many libraries, including
1710 Exim version information for dynamically loaded libraries. Created
1711 version.h, now support a version extension string for distributors
1712 who patch heavily. Dynamic module ABI change.
1714 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1715 privilege escalation vulnerability whereby the Exim run-time user
1716 can cause root to append content of the attacker's choosing to
1719 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1722 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1723 If dropping privileges for untrusted macros, we disabled normal logging
1724 on the basis that it would fail; for the Exim run-time user, this is not
1725 the case, and it resulted in successful deliveries going unlogged.
1726 Fixed. Reported by Andreas Metzler.
1732 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1733 only prepend for the Resent-* case. Fixes regression introduced in
1734 Exim 4.70 by NM/22 for Bugzilla 607.
1736 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1737 increasing numbers of administrators be bitten by this.
1739 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1741 PP/03 Bugzilla 994: added openssl_options main configuration option.
1743 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1745 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1747 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1749 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1750 without a peer certificate, leading to a segfault because of an
1751 assumption that peers always have certificates. Be a little more
1752 paranoid. Problem reported by Martin Tscholak.
1754 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1755 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1756 NB: ClamAV planning to remove STREAM in "middle of 2010".
1757 CL also introduces -bmalware, various -d+acl logging additions and
1758 more caution in buffer sizes.
1760 PP/09 Implemented reverse_ip expansion operator.
1762 PP/10 Bugzilla 937: provide a "debug" ACL control.
1764 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1766 PP/12 Bugzilla 973: Implement --version.
1768 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1770 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1772 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1774 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1775 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1776 ignore trailing whitespace.
1778 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1780 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1783 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1784 Notification from Dr Andrew Aitchison.
1786 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1787 ExtendedDetectionInfo response format.
1788 Notification from John Horne.
1790 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1793 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1794 XSL and documented dependency on system catalogs, with examples of how
1797 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1800 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1801 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1802 configuration file which is writeable by the Exim user or group.
1804 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1805 of configuration files to cover files specified with the -C option if
1806 they are going to be used with root privileges, not just the default
1809 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1810 option (effectively making it always true).
1812 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1813 files to be used while preserving root privileges.
1815 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1816 that rogue child processes cannot use them.
1818 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1819 run-time user, instead of root.
1821 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1822 Exim run-time user without dropping privileges.
1824 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1825 result string, instead of calling string_vformat() twice with the same
1828 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1829 for other users. Others should always drop root privileges if they use
1830 -C on the command line, even for a whitelisted configure file.
1832 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1834 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1840 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1841 $data_path, and $header_path variables; fixed documentation bugs and
1844 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1845 exipick to access non-standard spools, including the "frozen" queue
1848 NM/01 Bugzilla 965: Support mysql stored procedures.
1849 Patch from Alain Williams
1851 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1853 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1854 Patch from Andreas Metzler
1856 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1857 Patch from Kirill Miazine
1859 NM/05 Bugzilla 671: Added umask to procmail example.
1861 JJ/03 installed exipick 20100323.0, fixing doc bug
1863 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1864 directory. Notification and patch from Dan Rosenberg.
1866 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1868 TK/02 Improve log output when DKIM signing operation fails.
1870 MH/01 Treat the transport option dkim_domain as a colon separated
1871 list, not as a single string, and sign the message with each element,
1872 omitting multiple occurences of the same signer.
1874 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1875 Bugzilla 985, 986. Patch by Simon Arlott
1877 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1878 Patch by Simon Arlott
1880 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1881 MBX locking. Notification from Dan Rosenberg.
1887 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1889 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1891 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1893 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1895 NM/04 Bugzilla 915: Flag broken perl installation during build.
1901 TK/01 Added patch by Johannes Berg that expands the main option
1902 "spamd_address" if it starts with a dollar sign.
1904 TK/02 Write list of recipients to X-Envelope-Sender header when building
1905 the mbox-format spool file for content scanning (suggested by Jakob
1908 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1909 (http://www.dcc-servers.net/) support via dccifd. Activated by
1910 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1912 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1913 by Mark Daniel Reidel <mr@df.eu>.
1915 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1916 When building exim an external PCRE library is now needed -
1917 PCRE is a system library on the majority of modern systems.
1918 See entry on PCRE_LIBS in EDITME file.
1920 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1921 conversation. Added nologin parameter to request.
1922 Patch contributed by Kirill Miazine.
1924 TF/01 Do not log submission mode rewrites if they do not change the address.
1926 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1928 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1929 log files in place. Contributed by Roberto Lima.
1931 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1933 TF/03 Bugzilla 615: When checking the local_parts router precondition
1934 after a local_part_suffix or local_part_prefix option, Exim now
1935 does not use the address's named list lookup cache, since this
1936 contains cached lookups for the whole local part.
1938 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1939 Robert Millan. Documentation is in experimental-spec.txt.
1941 TF/04 Bugzilla 668: Fix parallel build (make -j).
1943 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1945 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1946 Patch provided by Jan Srzednicki.
1948 TF/05 Leading white space used to be stripped from $spam_report which
1949 wrecked the formatting. Now it is preserved.
1951 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1952 that they are available at delivery time.
1954 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1956 TF/08 TLS error reporting now respects the incoming_interface and
1957 incoming_port log selectors.
1959 TF/09 Produce a more useful error message if an SMTP transport's hosts
1960 setting expands to an empty string.
1962 NM/06 Bugzilla 744: EXPN did not work under TLS.
1963 Patch provided by Phil Pennock.
1965 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1966 Patch provided by Richard Godbee.
1968 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1969 acl_smtp_notquit, added index entry.
1971 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1972 Patch provided by Eugene Bujak.
1974 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1975 accept(). Patch provided by Maxim Dounin.
1977 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1978 Patch provided by Phil Pennock.
1980 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1982 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1983 Patch provided by Brad "anomie" Jorsch.
1985 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1986 Patch provided by Dean Brooks.
1988 TK/05 Add native DKIM support (does not depend on external libraries).
1990 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1991 Patch provided by Graeme Fowler.
1993 NM/16 Bugzilla 851: Documentation example syntax fix.
1995 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1997 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2000 NM/19 Bugzilla 745: TLS version reporting.
2001 Patch provided by Phil Pennock.
2003 NM/20 Bugzilla 167: bool: condition support.
2004 Patch provided by Phil Pennock.
2006 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2007 clients. Patch provided by Phil Pennock.
2009 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2010 Patch provided by Brad "anomie" Jorsch.
2012 NM/23 Bugzilla 687: Fix misparses in eximstats.
2013 Patch provided by Heiko Schlittermann.
2015 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2016 Patch provided by Heiko Schlittermann.
2018 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2019 plus update to original patch.
2021 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2023 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2024 Patch provided by David Brownlee.
2026 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2028 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2030 NM/30 Bugzilla 888: TLS documentation bugfixes.
2032 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2034 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2035 Unlike the original bugzilla I have changed all shell scripts in src tree.
2037 NM/33 Bugzilla 898: Transport filter timeout fix.
2038 Patch by Todd Rinaldo.
2040 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2041 Patch by Serge Demonchaux.
2043 NM/35 Bugzilla 39: Base64 decode bug fixes.
2044 Patch by Jakob Hirsch.
2046 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2048 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2050 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2052 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2058 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2059 ALT-N's libdkim that I have put here:
2060 http://duncanthrax.net/exim-experimental/
2062 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2063 were called 'true' and it seems that C99 defines that as a reserved
2064 keyword to be used with 'bool' variable types. That means you could
2065 not include C99-style headers which use bools without triggering
2066 build errors in sieve.c.
2068 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2069 as mailq or other aliases. Changed the --help handling significantly
2070 to do whats expected. exim_usage() emits usage/help information.
2072 SC/01 Added the -bylocaldomain option to eximstats.
2074 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2076 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2078 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2084 PH/01 Another patch from the Sieve maintainer.
2086 PH/02 When an IPv6 address is converted to a string for single-key lookup
2087 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2088 dots are used instead of colons so that keys in lsearch files need not
2089 contain colons. This was done some time before quoting was made available
2090 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2091 (notated using the quote facility) so as to distinguish them from IPv4
2092 keys. This meant that lookups for IP addresses in host lists did not work
2093 for iplsearch lookups.
2095 This has been fixed by arranging for IPv6 addresses to be expressed with
2096 colons if the lookup type is iplsearch. This is not incompatible, because
2097 previously such lookups could never work.
2099 The situation is now rather anomalous, since one *can* have colons in
2100 ordinary lsearch keys. However, making the change in all cases is
2101 incompatible and would probably break a number of configurations.
2103 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2106 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2107 conversion specification without a maximum field width, thereby enabling
2108 a rogue spamd server to cause a buffer overflow. While nobody in their
2109 right mind would setup Exim to query an untrusted spamd server, an
2110 attacker that gains access to a server running spamd could potentially
2111 exploit this vulnerability to run arbitrary code as the Exim user.
2113 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2114 $primary_hostname instead of what libspf2 thinks the hosts name is.
2116 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2117 a directory entry by the name of the lookup key. Previously, if a
2118 symlink pointed to a non-existing file or a file in a directory that
2119 Exim lacked permissions to read, a lookup for a key matching that
2120 symlink would fail. Now it is enough that a matching directory entry
2121 exists, symlink or not. (Bugzilla 503.)
2123 PH/03 The body_linecount and body_zerocount variables are now exported in the
2126 PH/04 Added the $dnslist_matched variable.
2128 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2129 This means they are set thereafter only if the connection becomes
2132 PH/06 Added the client_condition to authenticators so that some can be skipped
2133 by clients under certain conditions.
2135 PH/07 The error message for a badly-placed control=no_multiline_responses left
2136 "_responses" off the end of the name.
2138 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2140 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2141 (without spaces) instead of just copying the configuration text.
2143 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2145 PH/11 Added $max_received_linelength.
2147 PH/12 Added +ignore_defer and +include_defer to host lists.
2149 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2150 way in which PCRE > 7.0 is built.
2152 PH/14 Implemented queue_only_load_latch.
2154 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2155 MAIL command. The effect was to mangle the value on 64-bit systems.
2157 PH/16 Another patch from the Sieve maintainer.
2159 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2161 PH/18 If a system quota error occurred while trying to create the file for
2162 a maildir delivery, the message "Mailbox is full" was not appended to the
2163 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2164 only to a quota excession during the actual writing of the file.
2166 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2167 characters?) which causes problems in log lines. The DN values are now
2168 passed through string_printing() before being added to log lines.
2170 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2171 and InterBase are left for another time.)
2173 PH/21 Added message_body_newlines option.
2175 PH/22 Guard against possible overflow in moan_check_errorcopy().
2177 PH/23 POSIX allows open() to be a macro; guard against that.
2179 PH/24 If the recipient of an error message contained an @ in the local part
2180 (suitably quoted, of course), incorrect values were put in $domain and
2181 $local_part during the evaluation of errors_copy.
2187 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2188 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2191 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2192 issue a MAIL command.
2194 PH/02 In an ACL statement such as
2196 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2198 if a client was not listed at all, or was listed with a value other than
2199 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2200 the condition was not true (as it should be), so access was not denied.
2201 The bug was that the ! inversion was incorrectly passed on to the second
2202 item. This has been fixed.
2204 PH/03 Added additional dnslists conditions == and =& which are different from
2205 = and & when the dns lookup returns more than one IP address.
2207 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2208 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2210 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2211 FSYNC, which compiles an option called disable_fsync that allows for
2212 bypassing fsync(). The documentation is heavily laced with warnings.
2214 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2216 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2217 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2218 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2219 including adding "make clean"; (3) Added -fPIC when compiling the test
2220 dynamically loaded module, to get rid of a warning.
2222 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2223 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2224 = 0s. The bug is otherwise harmless.
2226 PH/07 There was a bug in the dovecot authenticator such that the value of
2227 $auth1 could be overwritten, and so not correctly preserved, after a
2228 successful authentication. This usually meant that the value preserved by
2229 the server_setid option was incorrect.
2231 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2233 PH/09 Installed PCRE release 7.0.
2235 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2236 run for batched SMTP input. It is now run at the start of every message
2237 in the batch. While fixing this I discovered that the process information
2238 (output by running exiwhat) was not always getting set for -bs and -bS
2239 input. This is fixed, and it now also says "batched" for BSMTP.
2241 PH/11 Added control=no_pipelining.
2243 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2244 patch, slightly modified), and move the expansion of helo_data till after
2245 the connection is made in the smtp transport (so it can use these
2248 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2250 PH/14 Added log_selector = +pid.
2252 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2254 PH/16 Add ${if forany and ${if forall.
2256 PH/17 Added dsn_from option to vary the From: line in DSNs.
2258 PH/18 Flush SMTP output before performing a callout, unless control =
2259 no_callout_flush is set.
2261 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2262 was true (the default) a successful delivery failed to delete the retry
2263 item, thus causing premature timeout of the address. The bug is now
2266 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2268 PH/21 Long custom messages for fakedefer and fakereject are now split up
2269 into multiline responses in the same way that messages for "deny" and
2270 other ACL rejections are.
2272 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2273 with slight modification.
2275 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2276 draft, changing the syntax and factoring some duplicate code".
2278 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2279 for deliveries of the second and subsequent messages over the same SMTP
2282 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2283 ${reduce, with only minor "tidies".
2285 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2287 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2288 expansion side effects.
2290 PH/27 When a message times out after an over-quota error from an Exim-imposed
2291 quota, the bounce message says "mailbox is full". This message was not
2292 being given when it was a system quota that was exceeded. It now should
2295 MH/03 Made $recipients available in local_scan(). local_scan() already has
2296 better access to the recipient list through recipients_list[], but
2297 $recipients can be useful in postmaster-provided expansion strings.
2299 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2300 in the case of a MAIL command with additional options following the
2301 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2302 were accidentally chopped off.
2304 PH/29 SMTP synchronization checks are implemented when a command is read -
2305 there is a check that no more input is waiting when there shouldn't be
2306 any. However, for some commands, a delay in an ACL can mean that it is
2307 some time before the response is written. In this time, more input might
2308 arrive, invalidly. So now there are extra checks after an ACL has run for
2309 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2310 pipelining has not been advertised.
2312 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2314 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2315 returned in the SMTP response when smtp_return_error_details was false.
2316 This has been fixed.
2318 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2319 instead of the C I/O that was originally supplied, because problems were
2320 reported on Solaris.
2322 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2323 Exim which did not show up earlier: it was assuming that a call to
2324 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2325 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2326 was a macro that became an assignment, so it seemed to work. This has
2327 changed to a proper function call with a void return, hence the compile
2328 error. Exim's code has been fixed.
2330 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2333 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2335 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2337 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2338 because it turns out to be too big or there is a timeout) there is no
2339 "Completed" line in the log. When some messages of this type were
2340 selected by exigrep, they were listed as "not completed". Others were
2341 picked up by some special patterns. I have improved the selection
2342 criteria to be more general.
2344 PH/38 The host_find_failed option in the manualroute router can now be set
2345 to "ignore", to completely ignore a host whose IP address cannot be
2346 found. If all hosts are ignored, the behaviour is controlled by the new
2347 host_all_ignored option.
2349 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2350 homing or because of multiple MX records with /mx) generated more than
2351 one IP address, and the following item turned out to be the local host,
2352 all the secondary addresses of the first item were incorrectly removed
2353 from the list, along with the local host and any following hosts (which
2354 is what is supposed to happen).
2356 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2357 whoever called Exim into the -H file. In the case of the daemon it was
2358 behaving confusingly. When first started, it used values for whoever
2359 started the daemon, but after a SIGHUP it used the Exim user (because it
2360 calls itself on a restart). I have changed the code so that it now always
2363 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2364 message are rejected with the same error (e.g. no authentication or bad
2365 sender address), and a DATA command is nevertheless sent (as can happen
2366 with PIPELINING or a stupid MUA), the error message that was given to the
2367 RCPT commands is included in the rejection of the DATA command. This is
2368 intended to be helpful for MUAs that show only the final error to their
2371 PH/42 Another patch from the Sieve maintainer.
2373 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2374 Eximstats - Fixed some broken HTML links and added missing column headers
2376 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2377 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2379 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2381 PH/43 Yet another patch from the Sieve maintainer.
2383 PH/44 I found a way to check for a TCP/IP connection going away before sending
2384 the response to the final '.' that terminates a message, but only in the
2385 case where the client has not sent further data following the '.'
2386 (unfortunately, this is allowed). However, in many cases there won't be
2387 any further data because there won't be any more messages to send. A call
2388 to select() can be used: if it shows that the input is "ready", there is
2389 either input waiting, or the socket has been closed. An attempt to read
2390 the next input character can distinguish the two cases. Previously, Exim
2391 would have sent an OK response which the client would never have see.
2392 This could lead to message repetition. This fix should cure that, at
2393 least in a lot of common cases.
2395 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2396 advertised in response to EHLO.
2402 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2403 fixed by 4.65/MH/01 (is this a record?) are fixed:
2405 (i) An empty string was always treated as zero by the numeric comparison
2406 operators. This behaviour has been restored.
2408 (ii) It is documented that the numeric comparison operators always treat
2409 their arguments as decimal numbers. This was broken in that numbers
2410 starting with 0 were being interpreted as octal.
2412 While fixing these problems I realized that there was another issue that
2413 hadn't been noticed. Values of message_size_limit (both the global option
2414 and the transport option) were treated as octal if they started with 0.
2415 The documentation was vague. These values are now always treated as
2416 decimal, and I will make that clear in the documentation.
2422 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2423 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2426 MH/01 Don't check that the operands of numeric comparison operators are
2427 integers when their expansion is in "skipping" mode (fixes bug
2428 introduced by 4.64-PH/07).
2430 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2431 child addresses, Exim now panics and dies. Previously, because the count
2432 is held in a short int, deliveries were likely to be lost. As such a
2433 large number of recipients for a single message is ridiculous
2434 (performance will be very, very poor), I have chosen to impose a limit
2435 rather than extend the field.
2441 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2442 leftover -K file (the existence of which was triggered by #402).
2443 While we were at it, introduced process PID as part of the -K
2444 filename. This should rule out race conditions when creating
2447 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2448 processing considerably. Previous code took too long for large mails,
2449 triggering a timeout which in turn triggers #401.
2451 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2452 in the DK code in transports.c. sendfile() is not really portable,
2453 hence the _LINUX specificness.
2455 TF/01 In the add_headers option to the mail command in an Exim filter,
2456 there was a bug that Exim would claim a syntax error in any
2457 header after the first one which had an odd number of characters
2460 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2461 callout verification, Exim cached a "reject" for the entire domain. This
2462 is correct for most verifications, but it is not correct for a recipient
2463 verification with use_sender or use_postmaster set, because in that case
2464 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2465 case of MAIL FROM:<> rejection from other early rejections (e.g.
2466 rejection of HELO). When verifying a recipient using a non-null MAIL
2467 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2468 Whatever the result of the callout, the value of the domain cache is
2469 left unchanged (for any other kind of callout, getting as far as trying
2470 RCPT means that the domain itself is ok).
2472 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2475 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2476 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2477 session, the daemon ignores ECONNECT errors and logs others; it now
2478 ignores EPIPE as well.
2480 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2481 (quoted-printable decoding).
2483 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2484 later the small subsequent patch to fix an introduced bug.
2486 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2488 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2490 PH/08 An error is now given if message_size_limit is specified negative.
2492 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2493 to be given (somewhat) arbitrary names.
2495 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2498 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2499 miscellaneous code fixes
2501 PH/10 Added the log_reject_target ACL modifier to specify where to log
2504 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2505 hostname. This is wrong, because it relates to the incoming message (and
2506 probably the interface on which it is arriving) and not to the outgoing
2507 callout (which could be using a different interface). This has been
2508 changed to use the value of the helo_data option from the smtp transport
2509 instead - this is what is used when a message is actually being sent. If
2510 there is no remote transport (possible with a router that sets up host
2511 addresses), $smtp_active_hostname is used.
2513 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2514 tweaks were necessary in order to get it to work (see also 21 below):
2515 (a) The code assumed that strncpy() returns a negative number on buffer
2516 overflow, which isn't the case. Replaced with Exim's string_format()
2518 (b) There were several signed/unsigned issues. I just did the minimum
2519 hacking in of casts. There is scope for a larger refactoring.
2520 (c) The code used strcasecmp() which is not a standard C function.
2521 Replaced with Exim's strcmpic() function.
2522 (d) The code set only $1; it now sets $auth1 as well.
2523 (e) A simple test gave the error "authentication client didn't specify
2524 service in request". It would seem that Dovecot has changed its
2525 interface. Fortunately there's a specification; I followed it and
2526 changed what the client sends and it appears to be working now.
2528 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2531 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2532 address A is aliased to B and C, where B exists and C does not. Without
2533 -v the output is "A verified" because verification stops after a
2534 successful redirection if more than one address is generated. However,
2535 with -v the child addresses are also verified. Exim was outputting "A
2536 failed to verify" and then showing the successful verification for C,
2537 with its parentage. It now outputs "B failed to verify", showing B's
2538 parentage before showing the successful verification of C.
2540 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2541 look up a TXT record in a specific list after matching in a combined
2544 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2545 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2546 they consult the DNS. I had assumed they would set it the way they
2547 wanted; and indeed my experiments on Linux seem to show that in some
2548 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2549 To be on the safe side, however, I have now made the interface to
2550 host_find_byname() similar to host_find_bydns(), with an argument
2551 containing the DNS resolver options. The host_find_byname() function now
2552 sets these options at its start, just as host_find_bydns() does. The smtp
2553 transport options dns_qualify_single and dns_search_parents are passed to
2554 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2555 of host_find_byname() use the default settings of RES_DEFNAMES
2556 (qualify_single) but not RES_DNSRCH (search_parents).
2558 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2559 spool_read_header() do less string testing, by means of a preliminary
2560 switch on the second character of optional "-foo" lines. (This is
2561 overdue, caused by the large number of possibilities that now exist.
2562 Originally there were few.) While I was there, I also converted the
2563 str(n)cmp tests so they don't re-test the leading "-" and the first
2564 character, in the hope this might squeeze out yet more improvement.
2566 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2567 flag allowing group syntax was set by the header_syntax check but not
2568 turned off, possible causing trouble later; (2) The flag was not being
2569 set at all for the header_verify test, causing "group"-style headers to
2570 be rejected. I have now set it in this case, and also caused header_
2571 verify to ignore an empty address taken from a group. While doing this, I
2572 came across some other cases where the code for allowing group syntax
2573 while scanning a header line wasn't quite right (mostly, not resetting
2574 the flag correctly in the right place). These bugs could have caused
2575 trouble for malformed header lines. I hope it is now all correct.
2577 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2578 with the "reply" argument non-NULL. The code, however (which originally
2579 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2580 but it didn't always do it. This confused somebody who was copying the
2581 code for some other use. I have removed all the tests.
2583 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2584 feature that was used to support insecure browsers during the U.S. crypto
2585 embargo. It requires special client support, and Exim is probably the
2586 only MTA that supported it -- and would never use it because real RSA is
2587 always available. This code has been removed, because it had the bad
2588 effect of slowing Exim down by computing (never used) parameters for the
2589 RSA_EXPORT functionality.
2591 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2592 authenticator to fail if there's a tab character in the incoming data
2593 (there should never be unless someone is messing about, as it's supposed
2594 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2595 if the connection is using TLS or if the remote IP is the same as the
2596 local IP, and the "valid-client-cert option" if a client certificate has
2599 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2600 authenticators. This can be used for authorization after authentication
2601 succeeds. (In the case of plaintext, it servers for both authentication
2604 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2605 if any retry times were supplied.
2607 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2608 connection, where there is no client IP address to check. In this
2609 situation, the verify now always succeeds.
2611 PH/25 Applied John Jetmore's -Mset patch.
2613 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2615 PH/27 In a string expansion for a processed (not raw) header when multiple
2616 headers of the same name were present, leading whitespace was being
2617 removed from all of them, but trailing whitespace was being removed only
2618 from the last one. Now trailing whitespace is removed from each header
2619 before concatenation. Completely empty headers in a concatenation (as
2620 before) are ignored.
2622 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2623 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2625 PH/29 [Removed. This was a change that I later backed out, and forgot to
2626 correct the ChangeLog entry (that I had efficiently created) before
2627 committing the later change.]
2629 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2630 address errors (4xx response to RCPT) over the same connection as other
2631 messages routed to the same hosts. Such deliveries are always "forced",
2632 so retry times are not inspected. This resulted in far too many retries
2633 for the affected addresses. The effect occurred only when there were more
2634 hosts than the hosts_max_try setting in the smtp transport when it had
2635 the 4xx errors. Those hosts that it had tried were not added to the list
2636 of hosts for which the message was waiting, so if all were tried, there
2637 was no problem. Two fixes have been applied:
2639 (i) If there are any address or message errors in an SMTP delivery, none
2640 of the hosts (tried or untried) are now added to the list of hosts
2641 for which the message is waiting, so the message should not be a
2642 candidate for sending over the same connection that was used for a
2643 successful delivery of some other message. This seems entirely
2644 reasonable: after all the message is NOT "waiting for some host".
2645 This is so "obvious" that I'm not sure why it wasn't done
2646 previously. Hope I haven't missed anything, but it can't do any
2647 harm, as the worst effect is to miss an optimization.
2649 (ii) If, despite (i), such a delivery is accidentally attempted, the
2650 routing retry time is respected, so at least it doesn't keep
2651 hammering the server.
2653 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2654 in ${readsocket because some servers need this prod.
2656 PH/32 Added some extra debug output when updating a wait-xxx database.
2658 PH/33 The hint "could be header name not terminated by colon", which has been
2659 given for certain expansion errors for a long time, was not being given
2660 for the ${if def:h_colon_omitted{... case.
2662 PH/34 The spec says: "With one important exception, whenever a domain list is
2663 being scanned, $domain contains the subject domain." There was at least
2664 one case where this was not true.
2666 PH/35 The error "getsockname() failed: connection reset by peer" was being
2667 written to the panic log as well as the main log, but it isn't really
2668 panic-worthy as it just means the connection died rather early on. I have
2669 removed the panic log writing for the ECONNRESET error when getsockname()
2672 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2673 runs only) independently of the message's sender address. This meant
2674 that, if the 4xx error was in fact related to the sender, a different
2675 message to the same recipient with a different sender could confuse
2676 things. In particular, this can happen when sending to a greylisting
2677 server, but other circumstances could also provoke similar problems.
2678 I have changed the default so that the retry time for these errors is now
2679 based a combination of the sender and recipient addresses. This change
2680 can be overridden by setting address_retry_include_sender=false in the
2683 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2684 remote server are returned as part of bounce messages. This was not
2685 happening for LMTP over a pipe (the lmtp transport), but now it is the
2686 same for both kinds of LMTP.
2688 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2689 in header lines that were in fact CNAMEs. This is no longer the case.
2691 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2692 and queue runs started by the daemon processed all messages. This has
2693 been fixed so that -R and -S can now usefully be given with -q<time>.
2695 PH/40 Import PCRE release 6.7 (fixes some bugs).
2697 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2699 PH/42 Give an error if -q is specified more than once.
2701 PH/43 Renamed the variables $interface_address and $interface_port as
2702 $received_ip_address and $received_port, to make it clear that these
2703 values apply to message reception, and not to the outgoing interface when
2704 a message is delivered. (The old names remain recognized, of course.)
2706 PH/44 There was no timeout on the connect() call when using a Unix domain
2707 socket in the ${readsocket expansion. There now is.
2709 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2710 be meaningful with "accept".
2712 SC/01 Eximstats V1.43
2713 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2715 SC/02 Eximstats V1.44
2716 Use a glob alias rather than an array ref in the generated
2717 parser. This improves both readability and performance.
2719 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2720 Collect SpamAssassin and rejection statistics.
2721 Don't display local sender or destination tables unless
2722 there is data to show.
2723 Added average volumes into the top table text output.
2725 SC/04 Eximstats V1.46
2726 Collect data on the number of addresses (recipients)
2727 as well as the number of messages.
2729 SC/05 Eximstats V1.47
2730 Added 'Message too big' to the list of mail rejection
2731 reasons (thanks to Marco Gaiarin).
2733 SC/06 Eximstats V1.48
2734 Mainlog lines which have GMT offsets and are too short to
2735 have a flag are now skipped.
2737 SC/07 Eximstats V1.49 (Alain Williams)
2738 Added the -emptyok flag.
2740 SC/08 Eximstats V1.50
2741 Fixes for obtaining the IP address from reject messages.
2743 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2744 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2745 whitespace changes from 4.64-PH/27
2747 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2750 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2751 are found, allow negative numbers in numeric criteria)
2753 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2755 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2756 to match changes made in 4.64-PH/43
2758 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2760 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2761 "domain" is actually an IP address, and force a failure. This locks out
2762 those revolvers/nameservers that support "A-for-A" lookups, in
2763 contravention of the specifications.
2765 PH/48 When a host name was looked up from an IP address, and the subsequent
2766 forward lookup of the name timed out, the host name was left in
2767 $sender_host_name, contrary to the specification.
2769 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2770 restricted to single-key lookups, Exim was not diagnosing an error if
2771 * or *@ was used with a query-style lookup.
2773 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2775 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2776 long ago, but noone interested enough thought of it. Let's just say that
2777 the "1.1" means that there are some new functions that weren't there at
2778 some point in the past.
2780 PH/51 Error processing for expansion failure of helo_data from an smtp
2781 transport during callout processing was broken.
2783 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2784 tested/used via the -bh/-bhc/-bs options.
2786 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2787 bug, fixed in subsequent PCRE releases).
2789 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2790 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2792 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2798 SC/01 Use a glob alias rather than an array ref in eximstats generated
2799 parser. This improves both readability and performance.
2801 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2802 Don't display local sender or destination tables in eximstats unless
2803 there is data to show.
2804 Added average volumes into the eximstats top table text output.
2806 SC/03 Collect data on the number of addresses (recipients) as well
2807 as the number of messages in eximstats.
2809 TF/01 Correct an error in the documentation for the redirect router. Exim
2810 does (usually) call initgroups() when daemonizing.
2812 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2813 with consistent privilege compared to when running as a daemon.
2815 TF/03 Note in the spec that $authenticated_id is not set for local
2816 submissions from trusted users.
2818 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2819 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2821 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2822 by adding some example configuration directives to the default
2823 configuration file. A little bit of work is required to uncomment the
2824 directives and define how usernames and passwords are checked, but
2825 there is now a framework to start from.
2827 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2828 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2829 without this. I don't know how relevant this is to other LDAP libraries.
2831 PH/02 Add the verb name to the "unknown ACL verb" error.
2833 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2835 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2837 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2838 directory not expanded when it should be if an expanded home directory
2839 was set for the address (which is overridden by the transport).
2841 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2844 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2845 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2846 because it is too late at that time, and has no effect.
2848 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2849 security issue with \' (bugzilla #107). I could not use the
2850 PQescapeStringConn() function, because it needs a PGconn value as one of
2853 PH/08 When testing addresses using -bt, indicate those final addresses that
2854 are duplicates that would not cause an additional delivery. At least one
2855 person was confused, thinking that -bt output corresponded to deliveries.
2856 (Suppressing duplicates isn't a good idea as you lose the information
2857 about possibly different redirections that led to the duplicates.)
2859 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2860 systems where poll() doesn't work, in particular OS X.
2862 PH/10 Added more information to debugging output for retry time not reached.
2864 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2865 operations in malware.c.
2867 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2870 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2871 syslog_duplication set false, log lines that would normally be written
2872 both the the main log and to the reject log were not written to syslog at
2875 PH/14 In the default configuration, change the use of "message" in ACL warn
2876 statements to "add_header".
2878 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2879 not followed by a command (e.g. "seen endif").
2881 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2882 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2885 PH/17 Added extra conditions to the default value of delay_warning_condition
2889 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2890 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2891 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2894 The Auto-Submitted: and various List- headers are standardised, whereas I
2895 don't think Precedence: ever was.
2897 PH/18 Refactored debugging code in route_finduser() to show more information,
2898 in particular, the error code if getpwnam() issues one.
2900 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2901 This is apparently needed in addition to the PH/07 change above to avoid
2902 any possible encoding problems.
2904 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2905 but not after initializing Perl.
2907 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2908 output them only if debugging. By default they are written stderr,
2909 apparently, which is not desirable.
2911 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2914 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2917 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2919 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2920 authenticated or an ident call has been made. Suppress the default
2921 values for $authenticated_id and $authenticated_sender (but permit -oMai
2922 and -oMas) when testing with -bh.
2924 PH/24 Re-jigged the order of the tests in the default configuration so that the
2925 tests for valid domains and recipients precede the DNS black list and CSA
2926 tests, on the grounds that those ones are more expensive.
2928 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2929 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2930 This bug exists in every version of Exim that I still have, right back to
2933 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2934 However, an attempt to turn on case-sensitivity in a regex key by
2935 including (?-i) didn't work because the subject string was already
2936 lowercased, and the effects were non-intuitive. It turns out that a
2937 one-line patch can be used to allow (?-i) to work as expected.
2943 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2944 other effects) broke the use of negated acl sub-conditions.
2946 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2949 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2950 "Deny" causes Exim to reject the incoming connection with a 554 error.
2951 Unfortunately, if there is a major crisis, such as a disk failure,
2952 tcp-wrappers gives "deny", whereas what one would like would be some
2953 kind of temporary error. A kludge has been added to help with this.
2954 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2955 554 error is used if errno is still zero or contains ENOENT (which occurs
2956 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2959 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2961 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2962 errors. Otherwise a message that provokes a temporary error (when other
2963 messages do not) can cause a whole host to time out.
2965 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2966 addresses were routed directly to files or pipes from a redirect router.
2967 File deliveries just didn't batch; pipe deliveries might have suffered
2970 PH/06 A failure to get a lock for a hints database would erroneously always say
2971 "Failed to get write lock", even when it was really a read lock.
2973 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2974 of 0600. This has been changed to use the value of the lockfile_mode
2975 option (which defaults to 0600).
2977 PH/08 Applied small patch from the Sieve maintainer.
2979 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2980 folder from quota calculations, a direct delivery into this folder messed
2981 up the contents of the maildirsize file. This was because the regex was
2982 used only to exclude .Trash (or whatever) when the size of the mailbox
2983 was calculated. There was no check that a delivery was happening into an
2984 excluded directory. This bug has been fixed by ignoring all quota
2985 processing for deliveries into excluded directories.
2987 PH/10 Added the maildirfolder_create_regex option to appendfile.
2993 PH/01 The code for finding all the local interface addresses on a FreeBSD
2994 system running IPv6 was broken. This may well have applied to all BSD
2995 systems, as well as to others that have similar system calls. The broken
2996 code found IPv4 interfaces correctly, but gave incorrect values for the
2997 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2998 that it would not match correctly against @[] and not recognize the IPv6
3001 PH/02 The ipliteral router was not recognizing addresses of the form user@
3002 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3004 PH/03 Added disable_ipv6.
3006 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3007 decoded form, because it is most often used to construct To: headers
3008 lines in autoreplies, and the decoded form may well be syntactically
3009 invalid. However, $reply_address has leading white space removed, and all
3010 newlines turned into spaces so that the autoreply transport does not
3013 PH/05 If group was specified without a user on a router, and no group or user
3014 was specified on a transport, the group from the router was ignored.
3016 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3017 for visible compile-time settings that can be used to change these
3018 numbers, for those that want even more. Backwards compatibility with old
3019 spool files has been maintained. However, going back to a previous Exim
3020 release will lost any variables that are in spool files.
3022 PH/07 Two small changes when running in the test harness: increase delay when
3023 passing a TCP/IP connection to a new process, in case the original
3024 process has to generate a bounce, and remove special handling of
3025 127.0.0.2 (sic), which is no longer necessary.
3027 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3028 be the same on different OS.
3030 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3033 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3034 whether --show-vars was specified or not
3036 JJ/02 exipick: Added support for new ACL variable spool format introduced
3039 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3040 syntactically invalid From: or Reply-to: line, and a filter used this to
3041 generate an autoreply, and therefore failed to obtain an address for the
3042 autoreply, Exim could try to deliver to a non-existent relative file
3043 name, causing unrelated and misleading errors. What now happens is that
3044 it logs this as a hard delivery error, but does not attempt to create a
3047 PH/11 The exinext utility has a -C option for testing purposes, but although
3048 the given file was scanned by exinext itself; it wasn't being passed on
3049 when Exim was called.
3051 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3052 an end-of-file indication when reading a command response.
3054 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3055 compiled. In many other places in Exim, IPv6 addresses are always
3056 recognized, so I have changed this. It also means that IPv4 domain
3057 literals of the form [IPV4:n.n.n.n] are now always recognized.
3059 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3060 used if the router is not running as root, for example, when verifying at
3061 ACL time, or when using -bh. The debugging output from this situation was
3062 non-existent - all you got was a failure to exec. I have made two
3065 (a) Failures to set uid/gid, the current directory, or a process leader
3066 in a subprocess such as that created by queryprogram now generate
3067 suitable debugging output when -d is set.
3069 (b) The queryprogram router detects when it is not running as root,
3070 outputs suitable debugging information if -d is set, and then runs
3071 the subprocess without attempting to change uid/gid.
3073 PH/15 Minor change to Makefile for building test_host (undocumented testing
3076 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3077 additional section of a DNS packet that returns MX or SRV records.
3078 Instead, it always explicitly searches for A/AAAA records. This avoids
3079 major problems that occur when a DNS server includes only records of one
3080 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3081 fixed another bug: if SRV records were looked up and the corresponding
3082 address records were *not* found in the additional section, the port
3083 values from the SRV records were lost.
3085 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3086 using the correct key (the original address) when searching the retry
3087 rules in order to find which one to use for generating the retry hint.
3089 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3090 adding the default one. Similarly, if it contains a Reply-To: header, the
3091 errors_reply_to option, if set, is not used.
3093 PH/19 When calculating a retry time, Exim used to measure the "time since
3094 failure" by looking at the "first failed" field in the retry record. Now
3095 it does not use this if it is later than than the arrival time of the
3096 message. Instead it uses the arrival time. This makes for better
3097 behaviour in cases where some deliveries succeed, thus re-setting the
3098 "first failed" field. An example is a quota failure for a huge message
3099 when small messages continue to be delivered. Without this change, the
3100 "time since failure" will always be short, possible causing more frequent
3101 delivery attempts for the huge message than are intended.
3102 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3104 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3105 $1, $2, $3) because the numerical variables can be reset during some
3106 expansion items (e.g. "match"), thereby losing the authentication data.
3108 PH/21 Make -bV show the size of off_t variables so that the test suite can
3109 decide whether to run tests for quotas > 2G.
3111 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3112 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3113 filecount value is greater than 2G or if a quota value is greater than 2G
3114 on a system where the size of off_t is not greater than 4, a panic error
3117 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3118 never match. The debug and -bh output now contains an explicit error
3119 message indicating a malformed IPv4 address or mask.
3121 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3122 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3123 PH/23 above applies.
3125 PH/25 Do not write to syslog when running in the test harness. The only
3126 occasion when this arises is a failure to open the main or panic logs
3127 (for which there is an explicit test).
3129 PH/26 Added the /no_tell option to "control=freeze".
3131 PH/27 If a host name lookup failed very early in a connection, for example, if
3132 the IP address matched host_lookup and the reverse lookup yielded a name
3133 that did not have a forward lookup, an error message of the form "no IP
3134 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3135 could be logged. Now it outputs the IP address instead of "NULL".
3137 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3138 allows the sender and the authenticated sender to be set when
3139 submitting a message from within Exim. Since child_open_exim() is
3140 documented for local_scan(), the new function should be too.
3142 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3143 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3144 results in an empty string is now treated as unset.
3146 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3148 PH/31 Added sender_verify_fail logging option.
3150 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3151 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3152 tidied the source and removed it altogether.
3154 PH/33 When a queue run was abandoned because the load average was too high, a
3155 log line was always written; now it is written only if the queue_run log
3156 selector is set. In addition, the log line for abandonment now contains
3157 information about the queue run such as the pid. This is always present
3158 in "start" and "stop" lines but was omitted from the "abandon" line.
3160 PH/34 Omit spaces between a header name and the colon in the error message that
3161 is given when verify = headers_syntax fails (if there are lots of them,
3162 the message gets confusing).
3164 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3165 names, as there are now some PTR records that contain slashes. This check
3166 is only to protect against broken name servers that fall over on strange
3167 characters, so the fact that it applies to all lookups doesn't matter.
3169 PH/36 Now that the new test suite is complete, we can remove some of the
3170 special code in Exim that was needed for the old test suite. For example,
3171 sorting DNS records because real resolvers return them in an arbitrary
3172 order. The new test suite's fake resolver always returns records in the
3175 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3176 bounces) except when queue_only is set, to avoid logging races between
3177 the different processes.
3179 PH/38 Panic-die if .include specifies a non-absolute path.
3181 PH/39 A tweak to the "H" retry rule from its user.
3183 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3184 a label. They prevented compilation on older perls.
3186 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3187 a warning to be raised on newish perls.
3189 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3190 on queue. Changes to match documented behaviour of showing count of
3191 messages matching specified criteria.
3193 PH/40 Changed the default ident timeout from 30s to 5s.
3195 PH/41 Added support for the use of login_cap features, on those BSD systems
3196 that have them, for controlling the resources used by pipe deliveries.
3198 PH/42 The content-scanning code uses fopen() to create files in which to put
3199 message data. Previously it was not paying any attention to the mode of
3200 the files. Exim runs with umask(0) because the rest of the code creates
3201 files with open(), and sets the required mode explicitly. Thus, these
3202 files were ending up world-writeable. This was not a big issue, because,
3203 being within the spool directory, they were not world-accessible. I have
3204 created a function called modefopen, which takes an additional mode
3205 argument. It sets umask(777), creates the file, chmods it to the required
3206 mode, then resets the umask. All the relevant calls to fopen() in the
3207 content scanning code have been changed to use this function.
3209 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3210 to 24 hours. This avoids potential overflow problems when processing G
3211 and H retry rules. I suspect nobody ever tinkers with this value.
3213 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3215 PH/45 When the plaintext authenticator is running as a client, the server's
3216 challenges are checked to ensure they are valid base64 strings. By
3217 default, the authentication attempt is cancelled if an invalid string is
3218 received. Setting client_ignore_invalid_base64 true ignores these errors.
3219 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3220 they are received. Thus, the responses can be made to depend on the
3221 challenges. If an invalid string is ignored, an empty string is placed in
3224 PH/46 Messages that are created by the autoreply transport now contains a
3225 References: header, in accordance with RFCs 2822 and 3834.
3227 PH/47 Added authenticated_sender_force to the smtp transport.
3229 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3231 PH/49 Installed latest patch from the Sieve maintainer.
3233 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3234 was also set, the appendfile transport was unnecessarily scanning a
3235 directory of message files (e.g. for maildir delivery) to find the count
3236 of files (along with the size), even though it did not need this
3237 information. It now does the scan only if it needs to find either the
3238 size of the count of files.
3240 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3242 PH/52 Two bugs concerned with error handling when the smtp transport is
3245 (i) Exim was not creating retry information for temporary errors given
3246 for individual recipients after the DATA command when the smtp transport
3247 was used in LMTP mode. This meant that they could be retried too
3248 frequently, and not timed out correctly.
3250 (ii) Exim was setting the flag that allows error details to be returned
3251 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3252 recipients that were returned after the DATA command.
3254 PH/53 This is related to PH/52, but is more general: for any failing address,
3255 when detailed error information was permitted to be returned to the
3256 sender, but the error was temporary, then after the final timeout, only
3257 "retry timeout exceeded" was returned. Now it returns the full error as
3258 well as "retry timeout exceeded".
3260 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3261 do this, and (what is worse) MTAs that accept it.
3263 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3264 will now be deprecated.
3266 PH/56 New os.c-cygwin from the Cygwin maintainer.
3268 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3269 formats (previously only available in exim formats via -bpr, -bpru,
3270 and -bpra. Now also available in native and exiqgrep formats)
3272 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3273 with very large, slow to parse queues
3275 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3277 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3279 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3280 responses to authentication challenges, though it was showing the
3281 challenges; (ii) I've removed the CR characters from the debug output for
3284 PH/58 Allow for the insertion of a newline as well as a space when a string
3285 is turned into more than one encoded-word during RFC 2047 encoding. The
3286 Sieve code now uses this.
3288 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3289 data_4xx, lost_connection, tls_required.
3291 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3292 message was being sent as an SMTP response.
3294 PH/61 Add -l and -k options to exicyclog.
3296 PH/62 When verifying, if an address was redirected to one new address, so that
3297 verification continued, and the new address failed or deferred after
3298 having set something in $address_data, the value of $address_data was not
3299 passed back to the ACL. This was different to the case when no
3300 redirection occurred. The value is now passed back in both cases.
3302 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3303 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3304 use login_cap.h, so on its own it isn't the distinguishing feature. The
3305 new name refers directly to the setclassresources() function.
3307 PH/65 Added configuration files for NetBSD3.
3309 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3311 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3312 is preferred over IPv4.
3314 PH/68 The bounce_return_message and bounce_return_body options were not being
3315 honoured for bounces generated during the reception of non-SMTP messages.
3316 In particular, this applied to messages rejected by the ACL. This bug has
3317 been fixed. However, if bounce_return_message is true and bounce_return_
3318 body is false, the headers that are returned for a non-SMTP message
3319 include only those that have been read before the error was detected.
3320 (In the case of an ACL rejection, they have all been read.)
3322 PH/69 The HTML version of the specification is now built in a directory called
3323 spec_html instead of spec.html, because the latter looks like a path with
3324 a MIME-type, and this confuses some software.
3326 PH/70 Catch two compiler warnings in sieve.c.
3328 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3329 function verify_get_ident() calls ip_connect() to connect a socket, but
3330 if the "connect()" function timed out, ip_connect() used to close the
3331 socket. However, verify_get_ident() also closes the socket later, and in
3332 between Exim writes to the log, which may get opened at this point. When
3333 the socket was closed in ip_connect(), the log could get the same file
3334 descriptor number as the socket. This naturally causes chaos. The fix is
3335 not to close the socket in ip_connect(); the socket should be closed by
3336 the function that creates it. There was only one place in the code where
3337 this was missing, in the iplookup router, which I don't think anybody now
3338 uses, but I've fixed it anyway.
3340 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3341 well as to direct DNS lookups. Otherwise the handling of names in host
3342 lists is inconsistent and therefore confusing.
3348 PH/01 Two changes to the default runtime configuration:
3350 (1) Move the checks for relay_from_hosts and authenticated clients from
3351 after to before the (commented out) DNS black list checks.
3353 (2) Add control=submission to the relay_from_hosts and authenticated
3354 clients checks, on the grounds that messages accepted by these
3355 statements are most likely to be submissions.
3357 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3359 (1) Generate an error if the third argument for the ${prvs expansion is
3362 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3365 (3) Reset the variables that are obtained from the first argument of
3366 ${prvscheck and used in the second argument before leaving the code,
3367 because their memory is reclaimed, so using them afterwards may do
3370 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3371 one (it's much easier than Tom thought :-).
3373 (5) Because of (4), we can now allow for the use of $prvscheck_result
3374 inside the third argument.
3376 PH/03 For some reason, the default setting of PATH when running a command from
3377 a pipe transport was just "/usr/bin". I have changed it to
3380 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3381 anything to be listed in the output from -bV.
3383 PH/05 When a filter generated an autoreply, the entire To: header line was
3384 quoted in the delivery log line, like this:
3386 => >A.N.Other <ano@some.domain> <original@ddress> ...
3388 This has been changed so that it extracts the operative address. There
3389 may be more than one such address. If so, they are comma-separated, like
3392 => >ano@some.domain,ona@other.domain <original@ddress> ...
3394 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3395 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3396 not being looked up in the rDNS to get a host name, Exim was showing the
3397 IP address twice in Received: lines, even though the IP addresses were
3398 identical. For example:
3400 Received: from [1.2.3.4] (helo=[1.2.3.4])
3402 However, if the real host name was known, it was omitting the HELO data
3403 if it matched the actual IP address. This has been tidied up so that it
3404 doesn't show the same IP address twice.
3406 PH/07 When both +timestamp and +memory debugging was on, the value given by
3407 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3408 called by the debug printing, thereby overwriting the timestamp buffer.
3409 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3411 PH/08 When the original message was included in an autoreply transport, it
3412 always said "this is a copy of the message, including all the headers",
3413 even if body_only or headers_only was set. It now gives an appropriate
3416 PH/09 Applied a patch from the Sieve maintainer which:
3418 o fixes some comments
3419 o adds the (disabled) notify extension core
3420 o adds some debug output for the result of if/elsif tests
3421 o points to the current vacation draft in the documentation
3422 and documents the missing references header update
3426 o fixes a bug in processing the envelope test (when testing
3427 multiple envelope elements, the last element determined the
3430 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3431 Electronic Mail") by including:
3433 Auto-submitted: auto-generated
3435 in the messages that it generates (bounce messages and others, such as
3436 warnings). In the case of bounce messages for non-SMTP messages, there was
3437 also a typo: it was using "Auto_submitted" (underscore instead of
3438 hyphen). Since every message generated by Exim is necessarily in response
3439 to another message, thes have all been changed to:
3441 Auto-Submitted: auto-replied
3443 in accordance with these statements in the RFC:
3445 The auto-replied keyword:
3447 - SHOULD be used on messages sent in direct response to another
3448 message by an automatic process,
3450 - MUST NOT be used on manually-generated messages,
3452 - MAY be used on Delivery Status Notifications (DSNs) and Message
3453 Disposition Notifications (MDNs),
3455 - MUST NOT be used on messages generated by automatic or periodic
3456 processes, except for messages which are automatic responses to
3459 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3460 to the default Received: header definition.
3462 PH/12 Added log selector acl_warn_skipped (default on).
3464 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3465 variables because (a) they are in the wrong storage pool and (b) even if
3466 they were copied, it wouldn't work properly because of the caching.
3468 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3469 checking when decoding. Apparently there are clients that generate
3470 overlong encoded strings. Why am I not surprised?
3472 PH/15 If the first argument of "${if match_address" was not empty, but did not
3473 contain an "@" character, Exim crashed. Now it writes a panic log message
3474 and treats the condition as false.
3476 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3478 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3479 extension "envelope-auth". The code is finished and in agreement with
3480 other implementations, but there is no documentation so far and in fact,
3481 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3482 not changing the active code.
3484 Print executed "if" and "elsif" statements when debugging is used. This
3485 helps a great deal to understand what a filter does.
3487 Document more things not specified clearly in RFC3028. I had all this
3488 sorted out, when out of a sudden new issues came to my mind. Oops."
3490 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3493 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3494 canonical form (as documented). However, after a host name lookup from
3495 the IP address, check_host() was doing a simple string comparison with
3496 addresses acquired from the DNS when checking that the found name did
3497 have the original IP as one of its addresses. Since any found IPv6
3498 addresses are likely to be in abbreviated form, the comparison could
3499 fail. Luckily, there already exists a function for doing the comparison
3500 by converting both addresses to binary, so now that is used instead of
3501 the text comparison.
3503 PH/20 There was another similar case to PH/19, when a complete host name was
3504 given in a host list; looking up its IP address could give an abbreviated
3505 form, whereas the current host's name might or might not be abbreviated.
3506 The same fix has been applied.
3512 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3513 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3516 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3517 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3519 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3521 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3522 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3523 Also, for "Auto-Submitted", treat anything other than "no" as
3524 non-personal, in accordance with RFC 3834. (Previously it treated
3525 anything starting "auto-" as non-personal.)
3527 TF/01 The control=submission/name=... option had a problem with syntax
3528 errors if the name included a slash character. The /name= option
3529 now slurps the rest of the string, so it can include any characters
3530 but it must come last in the list of options (after /sender_retain
3533 PH/05 Some modifications to the interface to the fake nameserver for the new
3541 TK/01 Added the "success_on_redirect" address verification option. See
3542 NewStuff for rationale and an example.
3544 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3546 PH/02 Patch to exigrep to allow it to work on syslog lines.
3548 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3549 fread() to skip over the body file's header line, because in Cygwin the
3550 header line is locked and is inaccessible.
3552 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3553 co-exist for some time) to make it clear that it is the Exim ID that is
3554 referenced, not the Message-ID: header line.
3556 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3557 string_format() function, because snprintf() does not exist on all
3560 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3561 ${stat: expansion item.
3563 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3564 protocol synchronization error", to keep the pedants happy.
3566 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3567 well as for IRIX systems, when gcc is being used. See the host.c source
3570 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3572 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3575 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3576 transport and to the smtp transport in LMTP mode.
3578 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3580 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3581 run a filter in a subprocess. This could lead to confusion in subsequent
3582 lookups in the parent process. There should also be a search_tidyup() at
3583 the end of the subprocess.
3585 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3586 only if the host matched helo_try_verify_hosts, which caused the
3587 verification to occur when the EHLO/HELO command was issued. The ACL just
3588 tested the remembered result. Now, if a previous verification attempt has
3589 not happened, "verify = helo" does it there and then.
3591 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3593 TK/03 Fix log output including CR from clamd.
3595 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3596 exist provoked a memory error which could cause a segfault.
3598 PH/15 Installed PCRE 6.2
3600 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3602 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3603 of the problem. Specifically, suggested +O2 rather than +O1 for the
3606 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3608 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3609 record, the port was not correctly logged when the outgoing_port log
3610 selector was set (it logged the transort's default port).
3612 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3613 fallback_hosts, and "hosts" in the smtp transport.
3615 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3616 host errors such as "Connection refused".
3618 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3619 authentication with radiusclient 0.4.9:
3621 - Error returned from rc_read_config was caught wrongly
3622 - Username/password not passed on to radius server due to wrong length.
3624 The presumption is that some radiusclient API changes for 4.51/PH/17
3625 were not taken care of correctly. The code is still untested by me (my
3626 Linux distribution still has 0.3.2 of radiusclient), but it was
3627 contributed by a Radius user.
3629 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3630 expanding the "port" option of the smtp transport.
3632 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3633 while reading a MIME header. Thanks to Tom Hughes for a patch.
3635 PH/24 Include config.h inside local_scan.h so that configuration settings are
3638 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3639 that in an ACL for RCPT (for example), you can examine exactly what was
3642 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3643 commands, but it was not correctly comparing the address with the actual
3644 client host address. Thus, it would show the EHLO address in Received:
3645 header lines when this was not necessary.
3647 PH/27 Added the % operator to ${eval:}.
3649 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3650 it should be ignoring failures (because with -C, for example, it has lost
3651 privilege). It wasn't ignoring creation failures other than "already
3654 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3657 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3658 because an input line was too long, either on its own, or by virtue of
3659 too many continuations, the temporary file was not being removed, and the
3660 return code was incorrect.
3662 PH/31 Missing "BOOL" in function definition in filtertest.c.
3664 PH/32 Applied Sieve patches from the maintainer.
3666 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3668 PH/33 Added "verify = not_blind".
3670 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3671 Local/Makefile (with some defaults set). These are used in built scripts
3672 such as exicyclog, but they have never been used in the exim_install
3673 script (though there are many overriding facilities there). I have
3674 arranged that the exim_install script now takes note of these two
3677 PH/35 Installed configuration files for Dragonfly.
3679 PH/36 When a locally submitted message by a trusted user did not contain a
3680 From: header, and the sender address was obtained from -f or from an SMTP
3681 MAIL command, and the trusted user did not use -F to supply a sender
3682 name, $originator_name was incorrectly used when constructing a From:
3683 header. Furthermore, $originator_name was used for submission mode
3684 messages from external hosts without From: headers in a similar way,
3685 which is clearly wrong.
3687 PH/37 Added control=suppress_local_fixups.
3689 PH/38 When log_selector = +received_sender was set, and the addition of the
3690 sender made the log line's construction buffer exactly full, or one byte
3691 less than full, an overflow happened when the terminating "\n" was
3694 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3695 when the result of a list match is failure because a DNS lookup failed.
3697 PH/40 RM_COMMAND is now used in the building process.
3699 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3700 the "build-* directories that it finds.
3702 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3703 address in a domain literal was a prefix of an interface address.
3705 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3706 when verifying a sender address, unless rewrite_headers is false.
3708 PH/44 Wrote a long comment about why errors_to addresses are verified as
3709 recipients, not senders.
3711 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3712 the ratelimit ACL was added.
3714 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3716 PH/46 Added extra information about PostgreSQL errors to the error string.
3718 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3719 avoiding the need to install special zones in a real server. This is
3720 backwards compatible; if it can't find the fake resolver, it drops back.
3721 Thus, both old and new test suites can be run.
3723 TF/02 Added util/ratelimit.pl
3725 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3726 clock is set back in time.
3728 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3729 Candler <B.Candler@pobox.com>.
3731 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3732 OK for addresses that are the result of redirections.
3734 PH/48 A number of further additions for the benefit of the new test suite,
3735 including a fake gethostbyname() that interfaces to the fake DNS resolver
3738 TF/06 The fix for widen_domains has also been applied to qualify_single and
3739 search_parents which are the other dnslookup options that can cause
3742 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3745 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3747 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3748 service" if it encountered an MX record with an empty target hostname.
3749 The message is now "an MX or SRV record indicated no SMTP service".
3751 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3752 if the DNS of the sending site is misconfigured. This is quite a
3753 common situation. This change restores the behaviour of treating a
3754 helo verification defer as a failure.
3756 PH/51 If self=fail was set on a router, the bounce message did not include the
3757 actual error message.
3763 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3765 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3766 command itself ended in error, the underlying message about the transport
3767 filter timeout was being overwritten with the pipe command error. Now the
3768 underlying error message should be appended to the second error message.
3770 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3772 PH/02 Reduce the amount of output that "make" produces by default. Full output
3773 can still be requested.
3775 PH/03 The warning log line about a condition test deferring for a "warn" verb
3776 was being output only once per connection, rather than after each
3777 occurrence (because it was using the same function as for successful
3778 "warn" verbs). This seems wrong, so I have changed it.
3780 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3781 it should not have, which might have caused a crash in the right
3782 circumstances, but probably never did.
3784 PH/04 Installed a modified version of Tony Finch's patch to make submission
3785 mode fix the return path as well as the Sender: header line, and to
3786 add a /name= option so that you can make the user's friendly name appear
3789 TF/03 Added the control = fakedefer ACL modifier.
3791 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3792 Mark Lowes for thorough testing.
3794 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3796 TK/03 Merged latest SRS patch from Miles Wilton.
3798 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3799 with the definition in sysexits.h (which is #included earlier).
3800 Fortunately, Exim does not actually use EX_OK. The code used to try to
3801 preserve the sysexits.h value, by assuming that macro definitions were
3802 scanned for macro replacements. I have been disabused of this notion,
3803 so now the code just undefines EX_OK before #including unistd.h.
3805 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3806 in the smtp transport. When a block could not be written in a single
3807 write() function, the timeout was being re-applied to each part-write.
3808 This seems wrong - if the receiver was accepting one byte at a time it
3809 would take for ever. The timeout is now adjusted when this happens. It
3810 doesn't have to be particularly precise.
3812 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3813 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3815 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3816 without a domain if the check to <postmaster@domain> fails.
3818 SC/01 Eximstats: added -xls and the ability to specify output files
3819 (patch written by Frank Heydlauf).
3821 SC/02 Eximstats: use FileHandles for outputting results.
3823 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3825 SC/04 Eximstats: fixed display of large numbers with -nvr option
3827 SC/05 Eximstats: fixed merging of reports with empty tables.
3829 SC/06 Eximstats: added the -include_original_destination flag
3831 SC/07 Eximstats: removed tabs and trailing whitespace.
3833 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3835 TK/06 MBOX spool code: Add real "From " MBOX separator line
3836 so the .eml file is really in mbox format (even though
3837 most programs do not really care). Patch from Alex Miller.
3839 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3840 The latter is generated from $received_to and is only set if the
3841 message has one envelope recipient. SA can use these headers,
3842 obviously out-of-the-box. Patch from Alex Miller.
3844 PH/08 The ${def test on a variable was returning false if the variable's
3845 value was "0", contrary to what the specification has always said!
3846 The result should be true unless the variable is empty.
3848 PH/09 The syntax error of a character other than { following "${if
3849 def:variable_name" (after optional whitespace) was not being diagnosed.
3850 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3851 accidental colon was present, for example, could give incorrect results.
3853 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3854 result is used (not including appendfile, where other changes are about
3857 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3858 This involved changing a lot of size variables from int to off_t. It
3859 should work with maildirs and everything.
3861 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3862 spamd dying while we are connected to it.
3864 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3867 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3868 function for BDB 4.3.
3870 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3872 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3873 the sources for PCRE have been re-arranged and more files are now
3876 PH/15 The code I had for printing potentially long long variables in PH/11
3877 above was not the best (it lost precision). The length of off_t variables
3878 is now inspected at build time, and an appropriate printing format (%ld
3879 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3880 to be "long long int" or "long int". This is needed for the internal
3881 formatting function string_vformat().
3883 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3884 the configuration file to be ":syslog", then the script "guesses" where
3885 the logs files are, rather than using the compiled in default. In our
3886 case the guess is not the same as the compiled default, so the script
3887 suddenly stopped working when I started to use syslog. The patch checks
3888 to see if log_file_path is "". If so, it attempts to read it from exim
3889 with no configuration file to get the compiled in version, before it
3890 falls back to the previous guessing code."
3892 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3893 implementing BATV in an Exim configuration. See NewStuff for the gory
3896 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3897 Makefile that are specific to HP-UX.
3899 PH/18 If the "use_postmaster" option was set for a recipient callout together
3900 with the "random" option, the postmaster address was used as the MAIL
3901 FROM address for the random test, but not for the subsequent recipient
3902 test. It is now used for both.
3904 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3905 patch removes a few documentation additions to RFC 3028, because the
3906 latest draft now contains them. It adds the new en;ascii-case comparator
3907 and a new error check for 8bit text in MIME parts. Comparator and
3908 require names are now matched exactly. I enabled the subaddress
3909 extension, but it is not well tested yet (read: it works for me)."
3911 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3912 rework some of the code of TK/09 above to avoid the hardwired use of
3913 "%lld" and "long long". Replaced the call to snprintf() with a call to
3916 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3917 records point to non-existent hosts", "retry timeout exceeded", and
3918 "retry time not reached for any host after a long failure period".
3920 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3921 experimental DomainKeys support:
3923 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3924 (2) On an error such as an illegally used "control", the wrong name for
3925 the control was given.
3927 These problems did NOT occur unless DomainKeys support was compiled.
3929 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3931 PH/24 Added ${if match_ip condition.
3933 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3934 databases so that it will be absolutely obvious if a crash occurs in the
3935 DB library. This is a regular occurrence (often caused by mis-matched
3938 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3939 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3940 picked up on a user's system that detects such things. There doesn't seem
3941 to be a gcc warning option for this - only an attribute that has to be
3942 put on the function's prototype. It seems that in Fedora Core 4 they have
3943 set this on a number of new functions. No doubt there will be more in due
3946 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3947 specify a transport. However, if an address that was verified by such a
3948 router was the subject of a callout, Exim crashed because it tried to
3949 read the rcpt_include_affixes from the non-existent transport. Now it
3950 just assumes that the setting of that option is false. This bug was
3951 introduced by 4.51/PH/31.
3953 PH/28 Changed -d+all to exclude +memory, because that information is very
3954 rarely of interest, but it makes the output a lot bigger. People tend to
3955 do -d+all out of habit.
3957 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3958 code in os-type was giving problems when libc.so lives in lib64, like on
3961 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3962 aren't the modern standard, and it seems that some systems' include files
3963 don't always have them. Exim was already checking for some of the newer
3964 ones like T_AAAA, and defining it itself. I've added checks for all the
3965 record types that Exim uses.
3967 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3968 not automatically generating a new one, as it is supposed to. This
3969 prevented TLS from working. If the file did exist, but contained invalid
3970 data, a new version was generated, as expected. It was only the case of a
3971 non-existent file that was broken.
3973 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3974 with a change in libdomainkeys > 0.64.
3976 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3977 from DNS. If the selector record carries the flag, it now has
3978 precedence over the domain-wide flag.
3980 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3982 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3983 the use of a port name that isn't defined in /etc/services) occurred, the
3984 message was deferred as in a normal delivery, and thus remained on the
3985 spool, instead of being failed because of the mua_wrapper setting. This
3986 is now fixed, and I tidied up some of the mua_wrapper messages at the
3989 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3990 the messages in a hash of arrays rather than using individual hashes.
3991 This is a bit cleaner and results in dramatic memory savings, albeit
3992 at a slight CPU cost.
3994 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3995 as requested by Marc Sherman.
3997 SC/10 Eximstats: added histograms for user specified patterns as requested
4000 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4002 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4003 fopen() in the content-scanning modules that did not already have it.
4009 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4010 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4012 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4014 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4016 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4017 to test Sieve filters that use "vacation".
4019 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4020 that changes the way the GnuTLS parameters are stored in the cache file.
4021 The new format can be generated externally. For backward compatibility,
4022 if the data in the cache doesn't make sense, Exim assumes it has read an
4023 old-format file, and it generates new data and writes a new file. This
4024 means that you can't go back to an older release without removing the
4027 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4028 work if there are any delivery delays because "one_time" forces the
4029 parent to be marked "delivered", so its unseen clone is never tried
4030 again. For this reason, Exim now forbids the simultaneous setting of
4033 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4034 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4035 entry if you want to know the details. The fix, however, neglected to
4036 consider the case when local delivery batching is involved. The test for
4037 "previously delivered" was not happening when checking to see if an
4038 address could be batched with a previous (undelivered) one; under
4039 certain circumstances this could lead to multiple deliveries to the same
4042 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4043 in its include files, and this causes problems building Exim.
4045 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4046 header_syntax) but Exim was just ignoring anything given after a slash.
4047 In particular, this caused confusion with an attempt to use "verify =
4048 reverse_host_lookup/defer_ok". An error is now given when options are
4049 supplied for verify items that do not have them. (Maybe reverse_host_
4050 lookup should have a defer_ok option, but that's a different point.)
4052 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4053 defined by RFC 821) to 2048, because there were problems with some AUTH
4054 commands, and RFC 1869 says the size should be increased for extended
4055 SMTP commands that take arguments.
4057 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4060 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4061 "unknown" error; now it says that the functionality isn't in the binary.
4063 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4064 an address' error message when a string expansion fails (syntax or
4065 whatever). Otherwise the password may appear in the log. Following change
4066 PH/42 below, there is no longer a chance of it appearing in a bounce
4069 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4071 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4072 of its addresses was ever tried. (Bugzilla bug #2.)
4074 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4075 the result incorrectly in the debug output. (It correctly added a newline
4076 to what was transported.)
4078 TF/01 Added $received_time.
4080 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4081 commented out examples of how to interface to a virus scanner and to
4082 SpamAssassin. Also added commented examples of av_scanner and
4083 spamd_address settings.
4085 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4086 and controls are allowed in which ACLs. There were a couple of minor
4087 errors. Some of the entries in the conditions table (which is a table of
4088 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4089 negation of where the condition IS allowed.
4091 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4093 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4094 header file does not have a version number, so I've had to invent a new
4095 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4096 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4097 radiusclient), but it was contributed by a Radius user.
4099 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4100 files or directories, for OpenSSL.
4102 PH/19 When an Exim process that is running as root has to create an Exim log
4103 file, it does so in a subprocess that runs as exim:exim so as to get the
4104 ownership right at creation (otherwise, other Exim processes might see
4105 the file with the wrong ownership). There was no test for failure of this
4106 fork() call, which would lead to the process getting stuck as it waited
4107 for a non-existent subprocess. Forks do occasionally fail when resources
4108 run out. I reviewed all the other calls to fork(); they all seem to check
4111 PH/20 When checking for unexpected SMTP input at connect time (before writing
4112 the banner), Exim was not dealing correctly with a non-positive return
4113 from the read() function. If the client had disconnected by this time,
4114 the result was a log entry for a synchronization error with an empty
4115 string after "input=" when read() returned zero. If read() returned -1
4116 (an event I could not check), uninitialized data bytes were printed.
4117 There were reports of junk text (parts of files, etc) appearing after
4120 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4122 PH/22 Added support for macro redefinition, and (re)definition in between
4123 driver and ACL definitions.
4125 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4126 forgetting to use the resulting value; it was using the unexpanded value.
4128 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4129 hadn't been configured. The fix is from Juergen Kreileder, who
4130 understands it better than I do:
4132 "Here's what I see happening with three configured cyrus_sasl
4133 authenticators configured (plain, login, cram-md5):
4135 On startup auth_cyrus_sasl_init() gets called for each of these.
4136 This means three calls to sasl_listmech() without a specified mech_list.
4137 => SASL tests which mechs of all available mechs actually work
4138 => three warnings about OTP not working
4139 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4141 With the patch, sasl_listmech() also gets called three times. But now
4142 SASL's mech_list option is set to the server_mech specified in the the
4143 authenticator. Or in other words, the answer from sasl_listmech()
4144 gets limited to just the mech you're testing for (which is different
4146 => the return list contains just 'plain' or 'login', 'cram-md5' or
4147 nothing depending on the value of ob->server_mech.
4149 I've just tested the patch: Authentication still works fine,
4150 unavailable mechs specified in the exim configuration are still
4151 caught, and the auth.log warnings about OTP are gone."
4153 PH/25 When debugging is enabled, the contents of the command line are added
4154 to the debugging output, even when log_selector=+arguments is not
4157 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4158 answer is "GNU", and only if the return is "GNU/something" is the answer
4161 PH/27 $acl_verify_message is now set immediately after the failure of a
4162 verification in an ACL, and so is available in subsequent modifiers. In
4163 particular, the message can be preserved by coding like this:
4165 warn !verify = sender
4166 set acl_m0 = $acl_verify_message
4168 Previously, $acl_verify_message was set only while expanding "message"
4169 and "log_message" when a very denied access.
4171 PH/28 Modified OS/os.c-Linux with
4173 -#ifndef OS_LOAD_AVERAGE
4174 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4176 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4177 nomenclature these days.)
4179 PH/29 Installed patch from the Sieve maintainer that adds the options
4180 sieve_useraddress and sieve_subaddress to the redirect router.
4182 PH/30 In these circumstances:
4183 . Two addresses routed to the same list of hosts;
4184 . First host does not offer TLS;
4185 . First host accepts first address;
4186 . First host gives temporary error to second address;
4187 . Second host offers TLS and a TLS session is established;
4188 . Second host accepts second address.
4189 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4190 and peerdn, if requested) that were in fact used only for the second
4193 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4194 attention to any local part prefix or suffix that was matched by the
4195 router that accepted the address. It now behaves in the same way as it
4196 does for delivery: the affixes are removed from the local part unless
4197 rcpt_include_affixes is set on the transport.
4199 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4200 timeout during the DATA phase of an incoming message.
4202 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4203 applied a patch sanctioned by the Sieve maintainer.
4205 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4206 the uid or gid is negative. A case of a negative gid caused this to be
4207 noticed. The fix allows for either to be negative.
4209 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4210 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4213 PH/36 The change PH/12 above was broken. Fixed it.
4215 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4216 the grounds that routing the same address twice would always produce the
4217 same answer. This might have been true once, but it is certainly no
4218 longer true now. Routing a child address may depend on the previous
4219 routing that produced that child. Some complicated redirection strategies
4220 went wrong when messages had multiple recipients, and made Exim's
4221 behaviour dependent on the order in which the addresses were given.
4223 I have moved the duplicate checking until after the routing is complete.
4224 Exim scans the addresses that are assigned to local and remote
4225 transports, and removes any duplicates. This means that more work will be
4226 done, as duplicates will always all be routed, but duplicates are
4227 presumably rare, so I don't expect this is of any significance.
4229 For deliveries to pipes, files, and autoreplies, the duplicate checking
4230 still happens during the routing process, since they are not going to be
4233 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4234 It corrects a timeout issue with spamd. This is Ian's comment: "The
4235 background is that sometimes spamd either never reads data from a
4236 connection it has accepted, or it never writes response data. The exiscan
4237 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4238 blindly assumes that writes won't block so it may never time out."
4240 PH/39 Allow G after quota size as well as K and M.
4242 PH/40 The value set for $authenticated_id in an authenticator may not contain
4243 binary zeroes or newlines because the value is written to log lines and
4244 to spool files. There was no check on this. Now the value is run through
4245 the string_printing() function so that such characters are converted to
4246 printable escape sequences.
4248 PH/41 $message_linecount is a new variable that contains the total number of
4249 lines in the message. Compare $body_linecount, which is the count for the
4252 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4253 bounce and delay warning messages, except in certain special cases, which
4256 (a) An SMTP error message from a remote host;
4257 (b) A message specified in a :fail: redirection;
4258 (c) A message specified in a "fail" command in a system filter;
4259 (d) A message specified in a FAIL return from the queryprogram router;
4260 (e) A message specified by the cannot_route_message router option.
4262 In these cases only, Exim does include the error details in bounce and
4263 warning messages. There are also a few cases where bland messages such
4264 as "unrouteable address" or "local delivery error" are given.
4266 PH/43 $value is now also set for the "else" part of a ${run expansion.
4268 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4269 being worked on, but at least Exim now implements the latest version to
4272 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4273 process to complete was treated as a delivery failure, a timeout while
4274 writing the message to the pipe was logged, but erroneously treated as a
4275 successful delivery. Such timeouts include transport filter timeouts. For
4276 consistency with the overall process timeout, these timeouts are now
4277 treated as errors, giving rise to delivery failures by default. However,
4278 there is now a new Boolean option for the pipe transport called
4279 timeout_defer, which, if set TRUE, converts the failures into defers for
4280 both kinds of timeout. A transport filter timeout is now identified in
4283 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4284 systems where "make" and "gmake" are different, calling "gmake" at top
4285 level broke things. I've arranged for the value of $(MAKE) to be passed
4286 from the Makefile to this script so that it can call the same version of
4290 A note about Exim versions 4.44 and 4.50
4291 ----------------------------------------
4293 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4294 changes of various kinds. As a consequence, a big documentation update was
4295 needed. This delayed the release for rather longer than seemed good, especially
4296 in the light of a couple of (minor) security issues. Therefore, the changes
4297 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4298 release. So 4.44 and 4.50 are in effect two different branches that both start
4301 I have left the 4.50 change log unchanged; it contains all the changes since
4302 4.43. The change log for 4.44 is below; many of its items are identical to
4303 those for 4.50. This seems to be the most sensible way to preserve the
4304 historical information.
4310 1. Minor wording change to the doc/README.SIEVE file.
4312 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4313 computation of the current number of files was incorrect.
4315 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4316 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4319 4. Give more explanation in the error message when the command for a transport
4320 filter fails to execute.
4322 5. There are several places where Exim runs a non-Exim command in a
4323 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4324 was being done only for the command run by the queryprogram router. It is
4325 now done for all such subprocesses. The other cases are: ${run, transport
4326 filters, and the commands run by the lmtp and pipe transports.
4328 6. Added CONFIGURE_GROUP build-time option.
4330 7. Some older OS have a limit of 256 on the maximum number of file
4331 descriptors. Exim was using setrlimit() to set 1000 as a large value
4332 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4333 systems. I've change it so that if it can't get 1000, it tries for 256.
4335 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4336 was an oversight, and furthermore, ever since the addition of extra
4337 controls (e.g. 4.43/32), the checks on when to allow different forms of
4338 "control" were broken. There should now be diagnostics for all cases when a
4339 control that does not make sense is encountered.
4341 9. Added the /retain_sender option to "control=submission".
4343 10. $recipients is now available in the predata ACL (oversight).
4345 11. Tidy the search cache before the fork to do a delivery from a message
4346 received from the command line. Otherwise the child will trigger a lookup
4347 failure and thereby defer the delivery if it tries to use (for example) a
4348 cached ldap connection that the parent has called unbind on.
4350 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4351 of $address_data from the recipient verification was clobbered by the
4352 sender verification.
4354 13. The value of address_data from a sender verification is now available in
4355 $sender_address_data in subsequent conditions in the ACL statement.
4357 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4359 15. Added a new option "connect=<time>" to callout options, to set a different
4362 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4363 was its contents. (It was OK if the option was not defined at all.)
4365 17. A "Completed" log line is now written for messages that are removed from
4366 the spool by the -Mrm option.
4368 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4369 information about exactly what failed.
4371 19. Added -dd to debug only the daemon process.
4373 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4374 handles timeouts, both on the server side and network timeouts. Renamed the
4375 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4377 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4378 It is now set to "smtps".
4380 22. $host_address is now set to the target address during the checking of
4381 ignore_target_hosts.
4383 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4384 being passed; this would have caused $sender_host_name to have been used if
4385 matching the list had actually called for a host name (not very likely,
4386 since this list is usually IP addresses). A host name is now passed as
4389 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4390 code that specifies a non-restarting handler (typically sigaction() in
4391 modern systems) in an attempt to fix a rare and obscure crash bug.
4393 25. Narrowed the window for a race in the daemon that could cause it to ignore
4394 SIGCHLD signals. This is not a major problem, because they are used only to
4395 wake it up if nothing else does.
4397 26. A malformed maildirsize file could cause Exim to calculate negative values
4398 for the mailbox size or file count. Odd effects could occur as a result.
4399 The maildirsize information is now recalculated if the size or filecount
4402 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4403 support for a long time. Removed HAVE_SYS_VFS_H.
4405 28. Installed the latest version of exipick from John Jetmore.
4407 29. In an address list, if the pattern was not a regular expression, an empty
4408 subject address (from a bounce message) matched only if the pattern was an
4409 empty string. Non-empty patterns were not even tested. This was the wrong
4410 because it is perfectly reasonable to use an empty address as part of a
4411 database query. An empty address is now tested by patterns that are
4412 lookups. However, all the other forms of pattern expect the subject to
4413 contain a local part and a domain, and therefore, for them, an empty
4414 address still always fails if the pattern is not itself empty.
4416 30. Exim went into a mad DNS loop when attempting to do a callout where the
4417 host was specified on an smtp transport, and looking it up yielded more
4418 than one IP address.
4420 31. Re-factored the code for checking spool and log partition space into a
4421 function that finds that data and another that does the check. The former
4422 is then used to implement four new variables: $spool_space, $log_space,
4423 $spool_inodes, and $log_inodes.
4425 32. The RFC2047 encoding function was originally intended for short strings
4426 such as real names; it was not keeping to the 75-character limit for
4427 encoded words that the RFC imposes. It now respects the limit, and
4428 generates multiple encoded words if necessary. To be on the safe side, I
4429 have increased the buffer size for the ${rfc2047: expansion operator from
4432 33. It is now permitted to omit both strings after an "if" condition; if the
4433 condition is true, the result is "true". As before, when the second string
4434 is omitted, a false condition yields an empty string. This makes it less
4435 cumbersome to write custom ACL and router conditions.
4437 34. Failure to deliver a bounce message always caused it to be frozen, even if
4438 there was an errors_to setting on the router. The errors_to setting is now
4441 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4442 canonical form (fully expanded) before being placed in
4443 $sender_host_address.
4445 36. The table in the code that translates DNS record types into text (T_A to
4446 "A" for instance) was missing entries for NS and CNAME. It is just possible
4447 that this could have caused confusion if both these types were looked up
4448 for the same domain, because the text type is used as part of Exim's
4449 per-process caching. But the chance of anyone hitting this buglet seems
4452 37. The dnsdb lookup has been extended in a number of ways.
4454 (1) There is a new type, "zns", which walks up the domain tree until it
4455 finds some nameserver records. It should be used with care.
4457 (2) There is a new type, "mxh", which is like "mx" except that it returns
4458 just the host names, not the priorities.
4460 (3) It is now possible to give a list of domains (or IP addresses) to be
4461 looked up. The behaviour when one of the lookups defers can be
4462 controlled by a keyword.
4464 (4) It is now possible to specify the separator character for use when
4465 multiple records are returned.
4467 38. The dnslists ACL condition has been extended: it is now possible to supply
4468 a list of IP addresses and/or domains to be looked up in a particular DNS
4471 39. Added log_selector=+queue_time_overall.
4473 40. When running the queue in the test harness, wait just a tad after forking a
4474 delivery process, to get repeatability of debugging output.
4476 41. Include certificate and key file names in error message when GnuTLS fails
4477 to set them up, because the GnuTLS error message doesn't include the name
4478 of the failing file when there is a problem reading it.
4480 42. Allow both -bf and -bF in the same test run.
4482 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4484 44. The "Exiscan patch" is now merged into the mainline Exim source.
4486 45. Sometimes the final signoff response after QUIT could fail to get
4487 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4488 < 0 before doing a fflush(). This bug looks as though it goes back to the
4489 introduction of TLS in release 3.20, but "sometimes" must have been rare
4490 because the tests only now provoked it.
4492 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4493 (this can affect the format of dates).
4495 47. exim_tidydb, when checking for the continued existence of a message for
4496 which it has found a message-specific retry record, was not finding
4497 messages that were in split spool directories. Consequently, it was
4498 deleting retry records that should have stayed in existence.
4500 48. Steve fixed some bugs in eximstats.
4502 49. The SPA authentication driver was not abandoning authentication and moving
4503 on to the next authenticator when an expansion was forced to fail,
4504 contradicting the general specification for all authenticators. Instead it
4505 was generating a temporary error. It now behaves as specified.
4507 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4508 (the order specifies the preference for clients). The order is now AES256,
4509 AES128, 3DES, ARCFOUR128.
4511 51. Small patch to Sieve code - explicitly set From: when generating an
4514 52. Exim crashed if a remote delivery caused a very long error message to be
4515 recorded - for instance if somebody sent an entire SpamAssassin report back
4516 as a large number of 550 error lines. This bug was coincidentally fixed by
4517 increasing the size of one of Exim's internal buffers (big_buffer) that
4518 happened as part of the Exiscan merge. However, to be on the safe side, I
4519 have made the code more robust (and fixed the comments that describe what
4522 53. Now that there can be additional text after "Completed" in log lines (if
4523 the queue_time_overall log selector is set), a one-byte patch to exigrep
4524 was needed to allow it to recognize "Completed" as not the last thing in
4527 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4528 patch that reportedly fixes this has been added. I am not expert enough to
4529 create a test for it. This is what the patch creator wrote:
4531 "I found a little strange behaviour of ldap code when working with
4532 Windows 2003 AD Domain, where users was placed in more than one
4533 Organization Units. When I tried to give exim partial DN, the exit code
4534 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4535 But simultaneously result of request was absolutely normal ldap result,
4536 so I produce this patch..."
4538 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4539 REFERENCE, so I have modified the code to exclude the patch when that macro
4542 55. Some experimental protocols are using DNS PTR records for new purposes. The
4543 keys for these records are domain names, not reversed IP addresses. The
4544 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4545 leaves it alone. Component reversal etc. now happens only for IP addresses.
4548 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4550 57. Double the size of the debug message buffer (to 2048) so that more of very
4551 long debug lines gets shown.
4553 58. The exicyclog utility now does better if the number of log files to keep
4554 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4556 59. Two changes related to the smtp_active_hostname option:
4558 (1) $smtp_active_hostname is now available as a variable.
4559 (2) The default for smtp_banner uses $smtp_active_hostname instead
4560 of $primary_hostname.
4562 60. The host_aton() function is supposed to be passed a string that is known
4563 to be a valid IP address. However, in the case of IPv6 addresses, it was
4564 not checking this. This is a hostage to fortune. Exim now panics and dies
4565 if the condition is not met. A case was found where this could be provoked
4566 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4567 components; fortuitously, this particular loophole had already been fixed
4568 by change 4.50/55 above.
4570 If there are any other similar loopholes, the new check in host_aton()
4571 itself should stop them being exploited. The report I received stated that
4572 data on the command line could provoke the exploit when Exim was running as
4573 exim, but did not say which command line option was involved. All I could
4574 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4575 running as the user.
4578 61. There was a buffer overflow vulnerability in the SPA authentication code
4579 (which came originally from the Samba project). I have added a test to the
4580 spa_base64_to_bits() function which I hope fixes it.
4583 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4584 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4586 63. The daemon start-up calls getloadavg() while still root for those OS that
4587 need the first call to be done as root, but it missed one case: when
4588 deliver_queue_load_max is set with deliver_drop_privilege. This is
4589 necessary for the benefit of the queue runner, because there is no re-exec
4590 when deliver_drop_privilege is set.
4592 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4593 This has been fixed.
4595 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4596 was in use, was not putting the data itself into the right store pool;
4597 consequently, it could be overwritten for a subsequent message in the same
4598 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4601 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4603 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4604 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4605 the function were treating the return as a boolean value, which happened to
4606 work because 0=false and not-0=true, but is not correct code.
4608 68. The host_aton() function was not handling scoped IPv6 addresses (those
4609 with, for example, "%eth0" on the end) correctly.
4611 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4612 negated items (that is, ~something) in unsigned ints. Some compilers
4613 apparently mutter when there is no cast.
4615 70. If an address verification called from an ACL failed, and did not produce a
4616 user-specific message (i.e. there was only a "system" message), nothing was
4617 put in $acl_verify_message. In this situation, it now puts the system
4620 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4621 session; change 4.31/43 added the unwanted input to the log line - except
4622 that it did not do this in the start of session case. It now does.
4624 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4625 This is wrong and can cause the other end to generate a synchronization
4626 error if it is another Exim or anything else that does the synchronization
4627 check. A QUIT command is no longer sent after a timeout.
4629 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4630 during host lookups.
4632 74. The defer_ok option of callout verification was not working if it was used
4633 when verifying addresses in header lines, that is, for this case:
4635 verify = header_sender/callout=defer_ok
4637 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4638 those file descriptors could be used for SMTP connections. If anything
4639 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4640 could be sent to the SMTP client, causing chaos. The daemon now opens
4641 stdin, stdout, and stderr to /dev/null when it puts itself into the
4644 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4645 log by default. The user can override this with suitable Perl magic.
4647 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4648 the log message when discard triggers, was not working for the DATA ACL or
4649 for the non-SMTP ACL.
4651 78. Error message wording change in sieve.c.
4653 79. If smtp_accept_max_per_host was set, the number of connections could be
4654 restricted to fewer than expected, because the daemon was trying to set up
4655 a new connection before checking whether the processes handling previous
4656 connections had finished. The check for completed processes is now done
4657 earlier. On busy systems, this bug wouldn't be noticed because something
4658 else would have woken the daemon, and it would have reaped the completed
4661 80. If a message was submitted locally by a user whose login name contained one
4662 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4663 It caused a spool format error. I have fixed the spool reading code. A
4664 related problem was that the "from" clause in the Received: line became
4665 illegal because of the space(s). It is now covered by ${quote_local_part.
4667 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4670 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4672 83. Patch from Sieve maintainer to fix unterminated string problem in
4673 "vacation" handling.
4675 84. Some minor changes to the Linux configuration files to help with other
4676 OS variants using glibc.
4678 85. One more patch for Sieve to update vacation handling to latest spec.
4681 ----------------------------------------------------
4682 See the note above about the 4.44 and 4.50 releases.
4683 ----------------------------------------------------
4689 1. Change 4.43/35 introduced a bug that caused file counts to be
4690 incorrectly computed when quota_filecount was set in an appendfile
4693 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4694 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4697 3. Give more explanation in the error message when the command for a transport
4698 filter fails to execute.
4700 4. There are several places where Exim runs a non-Exim command in a
4701 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4702 was being done only for the command run by the queryprogram router. It is
4703 now done for all such subprocesses. The other cases are: ${run, transport
4704 filters, and the commands run by the lmtp and pipe transports.
4706 5. Some older OS have a limit of 256 on the maximum number of file
4707 descriptors. Exim was using setrlimit() to set 1000 as a large value
4708 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4709 systems. I've change it so that if it can't get 1000, it tries for 256.
4711 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4712 was an oversight, and furthermore, ever since the addition of extra
4713 controls (e.g. 4.43/32), the checks on when to allow different forms of
4714 "control" were broken. There should now be diagnostics for all cases when a
4715 control that does not make sense is encountered.
4717 7. $recipients is now available in the predata ACL (oversight).
4719 8. Tidy the search cache before the fork to do a delivery from a message
4720 received from the command line. Otherwise the child will trigger a lookup
4721 failure and thereby defer the delivery if it tries to use (for example) a
4722 cached ldap connection that the parent has called unbind on.
4724 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4725 of $address_data from the recipient verification was clobbered by the
4726 sender verification.
4728 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4729 was its contents. (It was OK if the option was not defined at all.)
4731 11. A "Completed" log line is now written for messages that are removed from
4732 the spool by the -Mrm option.
4734 12. $host_address is now set to the target address during the checking of
4735 ignore_target_hosts.
4737 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4738 being passed; this would have caused $sender_host_name to have been used if
4739 matching the list had actually called for a host name (not very likely,
4740 since this list is usually IP addresses). A host name is now passed as
4743 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4744 code that specifies a non-restarting handler (typically sigaction() in
4745 modern systems) in an attempt to fix a rare and obscure crash bug.
4747 15. Narrowed the window for a race in the daemon that could cause it to ignore
4748 SIGCHLD signals. This is not a major problem, because they are used only to
4749 wake it up if nothing else does.
4751 16. A malformed maildirsize file could cause Exim to calculate negative values
4752 for the mailbox size or file count. Odd effects could occur as a result.
4753 The maildirsize information is now recalculated if the size or filecount
4756 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4757 support for a long time. Removed HAVE_SYS_VFS_H.
4759 18. Updated exipick to current release from John Jetmore.
4761 19. Allow an empty sender to be matched against a lookup in an address list.
4762 Previously the only cases considered were a regular expression, or an
4765 20. Exim went into a mad DNS lookup loop when doing a callout where the
4766 host was specified on the transport, if the DNS lookup yielded more than
4769 21. The RFC2047 encoding function was originally intended for short strings
4770 such as real names; it was not keeping to the 75-character limit for
4771 encoded words that the RFC imposes. It now respects the limit, and
4772 generates multiple encoded words if necessary. To be on the safe side, I
4773 have increased the buffer size for the ${rfc2047: expansion operator from
4776 22. Failure to deliver a bounce message always caused it to be frozen, even if
4777 there was an errors_to setting on the router. The errors_to setting is now
4780 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4781 canonical form (fully expanded) before being placed in
4782 $sender_host_address.
4784 24. Updated eximstats to version 1.33
4786 25. Include certificate and key file names in error message when GnuTLS fails
4787 to set them up, because the GnuTLS error message doesn't include the name
4788 of the failing file when there is a problem reading it.
4790 26. Expand error message when OpenSSL has problems setting up cert/key files.
4793 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4794 (this can affect the format of dates).
4796 28. exim_tidydb, when checking for the continued existence of a message for
4797 which it has found a message-specific retry record, was not finding
4798 messages that were in split spool directories. Consequently, it was
4799 deleting retry records that should have stayed in existence.
4801 29. eximstats updated to version 1.35
4802 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4803 1.35 - bugfix such that pie charts by volume are generated correctly
4805 30. The SPA authentication driver was not abandoning authentication and moving
4806 on to the next authenticator when an expansion was forced to fail,
4807 contradicting the general specification for all authenticators. Instead it
4808 was generating a temporary error. It now behaves as specified.
4810 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4811 (the order specifies the preference for clients). The order is now AES256,
4812 AES128, 3DES, ARCFOUR128.
4814 31. Small patch to Sieve code - explicitly set From: when generating an
4817 32. Exim crashed if a remote delivery caused a very long error message to be
4818 recorded - for instance if somebody sent an entire SpamAssassin report back
4819 as a large number of 550 error lines. This bug was coincidentally fixed by
4820 increasing the size of one of Exim's internal buffers (big_buffer) that
4821 happened as part of the Exiscan merge. However, to be on the safe side, I
4822 have made the code more robust (and fixed the comments that describe what
4825 33. Some experimental protocols are using DNS PTR records for new purposes. The
4826 keys for these records are domain names, not reversed IP addresses. The
4827 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4828 leaves it alone. Component reversal etc. now happens only for IP addresses.
4831 34. The host_aton() function is supposed to be passed a string that is known
4832 to be a valid IP address. However, in the case of IPv6 addresses, it was
4833 not checking this. This is a hostage to fortune. Exim now panics and dies
4834 if the condition is not met. A case was found where this could be provoked
4835 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4836 components; fortuitously, this particular loophole had already been fixed
4837 by change 4.50/55 or 4.44/33 above.
4839 If there are any other similar loopholes, the new check in host_aton()
4840 itself should stop them being exploited. The report I received stated that
4841 data on the command line could provoke the exploit when Exim was running as
4842 exim, but did not say which command line option was involved. All I could
4843 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4844 running as the user.
4847 35. There was a buffer overflow vulnerability in the SPA authentication code
4848 (which came originally from the Samba project). I have added a test to the
4849 spa_base64_to_bits() function which I hope fixes it.
4852 36. The daemon start-up calls getloadavg() while still root for those OS that
4853 need the first call to be done as root, but it missed one case: when
4854 deliver_queue_load_max is set with deliver_drop_privilege. This is
4855 necessary for the benefit of the queue runner, because there is no re-exec
4856 when deliver_drop_privilege is set.
4858 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4859 was in use, was not putting the data itself into the right store pool;
4860 consequently, it could be overwritten for a subsequent message in the same
4861 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4864 38. Sometimes the final signoff response after QUIT could fail to get
4865 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4866 < 0 before doing a fflush(). This bug looks as though it goes back to the
4867 introduction of TLS in release 3.20, but "sometimes" must have been rare
4868 because the tests only now provoked it.
4874 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4875 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4876 Now it returns an integer. A number of calls were still expecting a T/F
4877 return. Fortuitously, in all cases, the tests worked in OK situations,
4878 which is the norm. However, things would have gone wrong on any write
4879 failures on the smtp file descriptor. This function is used when sending
4880 messages over SMTP and also when doing verify callouts.
4882 2. When Exim is called to do synchronous delivery of a locally submitted
4883 message (the -odf or -odi options), it no longer closes stderr before doing
4886 3. Implemented the mua_wrapper option.
4888 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4890 5. Implemented the functions header_remove(), header_testname(),
4891 header_add_at_position(), and receive_remove_recipient(), and exported them
4894 6. If an ACL "warn" statement specified the addition of headers, Exim already
4895 inserted X-ACL-Warn: at the start if there was no header name. However, it
4896 was not making this test for the second and subsequent header lines if
4897 there were newlines in the string. This meant that an invalid header could
4898 be inserted if Exim was badly configured.
4900 7. Allow an ACL "warn" statement to add header lines at the start or after all
4901 the Received: headers, as well as at the end.
4903 8. Added the rcpt_4xx retry error code.
4905 9. Added postmaster_mailfrom=xxx to callout verification option.
4907 10. Added mailfrom=xxxx to the callout verification option, for verify=
4910 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4911 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4913 12. Inserted some casts to stop certain compilers warning when using pointer
4914 differences as field lengths or precisions in printf-type calls (mostly
4915 affecting debugging statements).
4917 13. Added optional readline() support for -be (dynamically loaded).
4919 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4920 same clock tick as a message's arrival, so that its received time was the
4921 same as the "first fail" time on the retry record, and that message
4922 remained on the queue past the ultimate address timeout, every queue runner
4923 would try a delivery (because it was past the ultimate address timeout) but
4924 after another failure, the ultimate address timeout, which should have then
4925 bounced the address, did not kick in. This was a "< instead of <=" error;
4926 in most cases the first failure would have been in the next clock tick
4927 after the received time, and all would be well.
4929 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4930 being recognized when the domain list was tested by the match_domain
4931 condition in an expansion string.
4933 16. Added the ${str2b64: operator.
4935 17. Exim was always calling setrlimit() to set a large limit for the number of
4936 processes, without checking whether the existing limit was already
4937 adequate. (It did check for the limit on file descriptors.) Furthermore,
4938 errors from getrlimit() and setrlimit() were being ignored. Now they are
4939 logged to the main and panic logs, but Exim does carry on, to try to do its
4940 job under whatever limits there are.
4942 18. Imported PCRE 5.0.
4944 19. Trivial typo in log message " temporarily refused connection" (the leading
4947 20. If the log selector return_path_on_delivery was set and an address was
4948 redirected to /dev/null, the delivery process crashed because it assumed
4949 that a return path would always be set for a "successful" delivery. In this
4950 case, the whole delivery is bypassed as an optimization, and therefore no
4953 21. Internal re-arrangement: the function for sending a challenge and reading
4954 a response while authentication was assuming a zero-terminated challenge
4955 string. It's now changed to take a pointer and a length, to allow for
4956 binary data in such strings.
4958 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4960 23. Exim was not respecting finduser_retries when seeking the login of the
4961 uid under which it was called; it was always trying 10 times. (The default
4962 setting of finduser_retries is zero.) Also, it was sleeping after the final
4963 failure, which is pointless.
4965 24. Implemented tls_on_connect_ports.
4967 25. Implemented acl_smtp_predata.
4969 26. If the domain in control=submission is set empty, Exim assumes that the
4970 authenticated id is a complete email address when it generates From: or
4971 Sender: header lines.
4973 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4974 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4975 chown and chgrp in /bin and hostname in /usr/bin.
4977 28. Exim was keeping the "process log" file open after each use, just as it
4978 does for the main log. This opens the possibility of it remaining open for
4979 long periods when the USR1 signal hits a daemon. Occasional processlog
4980 errors were reported, that could have been caused by this. Anyway, it seems
4981 much more sensible not to leave this file open at all, so that is what now
4984 29. The long-running daemon process does not normally write to the log once it
4985 has entered its main loop, and it closes the log before doing so. This is
4986 so that log files can straightforwardly be renamed and moved. However,
4987 there are a couple of unusual error situations where the daemon does write
4988 log entries, and I had neglected to close the log afterwards.
4990 30. The text of an SMTP error response that was received during a remote
4991 delivery was being truncated at 512 bytes. This is too short for some of
4992 the long messages that one sometimes sees. I've increased the limit to
4995 31. It is now possible to make retry rules that apply only when a message has a
4996 specific sender, in particular, an empty sender.
4998 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4999 it possible to be selective about when SMTP synchronization is enforced.
5001 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5003 32. Implemented hosts_connection_nolog.
5005 33. Added an ACL for QUIT.
5007 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5010 35. Added mailbox_size and mailbox_filecount to appendfile.
5012 36. Added control = no_multiline_responses to ACLs.
5014 37. There was a bug in the logic of the code that waits for the clock to tick
5015 in the case where the clock went backwards by a substantial amount such
5016 that the microsecond fraction of "now" was more than the microsecond
5017 fraction of "then" (but the whole seconds number was less).
5019 38. Added support for the libradius Radius client library this is found on
5020 FreeBSD (previously only the radiusclient library was supported).
5026 1. When certain lookups returned multiple values in the form name=value, the
5027 quoting of the values was not always being done properly. Specifically:
5028 (a) If the value started with a double quote, but contained no whitespace,
5030 (b) If the value contained whitespace other than a space character (i.e.
5031 tabs or newlines or carriage returns) it was not quoted.
5032 This fix has been applied to the mysql and pgsql lookups by writing a
5033 separate quoting function and calling it from the lookup code. The fix
5034 should probably also be applied to nisplus, ibase and oracle lookups, but
5035 since I cannot test any of those, I have not disturbed their existing code.
5037 2. A hit in the callout cache for a specific address caused a log line with no
5038 reason for rejecting RCPT. Now it says "Previous (cached) callout
5039 verification failure".
5041 3. There was an off-by-one bug in the queryprogram router. An over-long
5042 return line was truncated at 256 instead of 255 characters, thereby
5043 overflowing its buffer with the terminating zero. As well as fixing this, I
5044 have increased the buffer size to 1024 (and made a note to document this).
5046 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5047 when Exim is waiting for an SMTP response from a remote server, Exim
5048 restarts its select() call on the socket, thereby resetting its timeout.
5049 This is not a problem when such interrupts are rare. Somebody set up a cron
5050 job to run exiwhat every 2 minutes, which is less than the normal select()
5051 timeout (5 or 10 minutes). This meant that the select() timeout never
5052 kicked in because it was always reset. I have fixed this by comparing the
5053 time when an interrupt arrives with the time at the start of the first call
5054 to select(). If more time than the timeout has elapsed, the interrupt is
5055 treated as a timeout.
5057 5. Some internal re-factoring in preparation for the addition of Sieve
5058 extensions (by MH). In particular, the "personal" test is moved to a
5059 separate function, and given an option for scanning Cc: and Bcc: (which is
5060 not set for Exim filters).
5062 6. When Exim created an email address using the login of the caller as the
5063 local part (e.g. when creating a From: or Sender: header line), it was not
5064 quoting the local part when it contained special characters such as @.
5066 7. Installed new OpenBSD configuration files.
5068 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5069 try to make them clearer.
5071 9. Callout options, other than the timeout value, were being ignored when
5072 verifying sender addresses in header lines. For example, when using
5074 verify = header_sender/callout=no_cache
5076 the cache was (incorrectly) being used.
5078 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5079 only the Cygwin environment.
5081 11. When return_path_on_delivery was set as a log selector, if different remote
5082 addresses in the same message used different return paths and parallel
5083 remote delivery occurred, the wrong values would sometimes be logged.
5084 (Whenever a remote delivery process finished, the return path value from
5085 the most recently started remote delivery process was logged.)
5087 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5088 lines when AUTH and/or TLS are in use. This is the "received protocol"
5089 field. Exim used to use "asmtp" for authenticated SMTP, without any
5090 indication (in the protocol name) for TLS use. Now it follows the RFC and
5091 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5092 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5093 names appear in log lines as well as in Received: header lines.
5095 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5096 extensions, and comparison tests, and to fix some bugs.
5098 14. Changes to the "personal" filter test:
5100 (1) The test was buggy in that it was just doing the equivalent of
5101 "contains" tests on header lines. For example, if a user's address was
5102 anne@some.where, the "personal" test would incorrectly be true for
5104 To: susanne@some.where
5106 This test is now done by extracting each address from the header in turn,
5107 and checking the entire address. Other tests that are part of "personal"
5108 are now done using regular expressions (for example, to check local parts
5109 of addresses in From: header lines).
5111 (2) The list of non-personal local parts in From: addresses has been
5112 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5113 taken from the Sieve specification recommendations.
5115 (3) If the message contains any header line starting with "List-" it is
5116 treated as non-personal.
5118 (4) The test for "circular" in the Subject: header line has been removed
5119 because it now seems ill-conceived.
5121 15. Minor typos in src/EDITME comments corrected.
5123 16. Installed latest exipick from John Jetmore.
5125 17. If headers_add on a router specified a text string that was too long for
5126 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5127 of string_sprintf() is now avoided.
5129 18. $message_body_size was not set (it was always zero) when running the DATA
5130 ACL and the local_scan() function.
5132 19. For the "mail" command in an Exim filter, no default was being set for
5133 the once_repeat time, causing a random time value to be used if "once" was
5134 specified. (If the value happened to be <= 0, no repeat happened.) The
5135 default is now 0s, meaning "never repeat". The "vacation" command was OK
5136 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5137 (I found it when inspecting the code).
5139 20. There is now an overall timeout for performing a callout verification. It
5140 defaults to 4 times the callout timeout, which applies to individual SMTP
5141 commands during the callout. The overall timeout applies when there is more
5142 than one host that can be tried. The timeout is checked before trying the
5143 next host. This prevents very long delays if there are a large number of
5144 hosts and all are timing out (e.g. when the network connections are timing
5145 out). The value of the overall timeout can be changed by specifying an
5146 additional sub-option for "callout", called "maxwait". For example:
5148 verify = sender/callout=5s,maxwait=20s
5150 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5151 to the end before writing, but this should make it even safer).
5153 22. Exim was forgetting that it had advertised PIPELINING for the second and
5154 subsequent messages on an SMTP connection. It was also not resetting its
5155 memory on STARTTLS and an internal HELO.
5157 23. When Exim logs an SMTP synchronization error within a session, it now
5158 records whether PIPELINING has been advertised or not.
5160 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5161 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5162 rather than long int.
5164 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5166 26. Added the never_mail option to autoreply.
5172 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5173 crash if the getsockname() call failed; this can happen if a connection is
5174 closed very soon after it is established. The problem was simply in the
5175 order in which certain operations were done, causing Exim to try to write
5176 to the SMTP stream before it had set up the file descriptor. The bug has
5177 been fixed by making things happen in the correct order.
5183 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5184 before the connection was closed, thus losing the rejection response.
5186 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5187 some early Solaris releases, but causes trouble in current releases where
5188 socklen_t is defined.
5190 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5193 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5196 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5197 that releasing the top of it at the end releases what was used for sub-
5198 expansions (unless the block got too big). However, discard this block if
5199 the first thing is a variable or header, so that we can use its block when
5200 it is dynamic (useful for very large $message_headers, for example).
5202 6. Lookups now cache *every* query, not just the most recent. A new, separate
5203 store pool is used for this. It can be recovered when all lookup caches are
5204 flushed. Lookups now release memory at the end of their result strings.
5205 This has involved some general refactoring of the lookup sources.
5207 7. Some code has been added to the store_xxx() functions to reduce the amount
5208 of flapping under certain conditions.
5210 8. log_incoming_interface used to affect only the <= reception log lines. Now
5211 it causes the local interface and port to be added to several more SMTP log
5212 lines, for example "SMTP connection from", and rejection lines.
5214 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5216 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5218 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5219 could be overwritten at the end of the current message (or the start of a
5220 new message if it was set in a HELO ACL). The value is now preserved for
5221 the duration of the SMTP connection.
5223 12. If a transport had a headers_rewrite setting, and a matching header line
5224 contained an unqualified address, that address was qualified, even if it
5225 did not match any rewriting rules. The underlying bug was that the values
5226 of the flags that permit the existence of unqualified sender and recipient
5227 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5228 non-local messages, and by -bnq for local messages) were not being
5229 preserved with the message after it was received.
5231 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5232 "next input=" as part of the text comprising the host identity instead of
5233 the correct text. The code was using the same buffer for two different
5234 strings. However, depending on which order the printing function evaluated
5235 its arguments, the bug did not always show up. Under Linux, for example, my
5236 test suite worked just fine.
5238 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5239 doesn't work with some older versions of Perl. It has been changed to "my",
5240 which in any case is probably the better facility to use.
5242 15. A really picky compiler found some instances of statements for creating
5243 error messages that either had too many or two few arguments for the format
5246 16. The size of the buffer for calls to the DNS resolver has been increased
5247 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5248 for addresses that have a lot of PTR records. This alleviates a problem; it
5249 does not fully solve it.
5251 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5252 buffer now truncates the list and logs the incident, which is the same
5253 action as happens when Exim is looking up a host name and its aliases.
5254 Previously in this situation something unpredictable would happen;
5255 sometimes it was "internal error: store_reset failed".
5257 18. If a server dropped the connection unexpectedly when an Exim client was
5258 using GnuTLS and trying to read a response, the client delivery process
5259 crashed while trying to generate an error log message.
5261 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5262 string, for example:
5264 warn message = H1: something\nH2: something
5266 the text was added as a single header line from Exim's point of view
5267 though it ended up OK in the delivered message. However, searching for the
5268 second and subsequent header lines using $h_h2: did not work. This has been
5269 fixed. Similarly, if a system filter added multiple headers in this way,
5270 the routers could not see them.
5272 20. Expanded the error message when iplsearch is called with an invalid key to
5273 suggest using net-iplsearch in a host list.
5275 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5276 ACLs are no longer actually imposed (and a message to that effect is
5279 22. If a "gecos" field in a passwd entry contained escaped characters, in
5280 particular, if it contained a \" sequence, Exim got it wrong when building
5281 a From: or a Sender: header from that name. A second bug also caused
5282 incorrect handling when an unquoted " was present following a character
5283 that needed quoting.
5285 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5286 was not being matched caselessly.
5288 24. Arranged for all hyphens in the exim.8 source to be escaped with
5291 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5292 properly. Recipient callout cache records were still being keyed to include
5293 the sender, even when use_sender was set false. This led to far more
5294 callouts that were necessary. The sender is no longer included in the key
5295 when use_sender is false.
5297 26. Added "control = submission" modifier to ACLs.
5299 27. Added the ${base62d: operator to decode base 62 numbers.
5301 28. dnsdb lookups can now access SRV records.
5303 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5304 the configuration file.
5306 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5307 (-v) mode. This makes the output for a verbose queue run more intelligible.
5309 31. Added a use_postmaster feature to recipient callouts.
5311 32. Added the $body_zerocount variable, containing the number of binary zero
5312 bytes in the message body.
5314 33. The time of last modification of the "new" subdirectory is now used as the
5315 "mailbox time last read" when there is a quota error for a maildir
5318 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5320 35. Added +ignore_unknown as a special item in host lists.
5322 36. Code for decoding IPv6 addresses in host lists is now included, even if
5323 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5324 address was recognized as an IP address, but was then not correctly decoded
5325 into binary, causing unexpected and incorrect effects when compared with
5332 1. Very minor rewording of debugging text in manualroute to say "list of
5333 hosts" instead of "hostlist".
5335 2. If verify=header_syntax was set, and a header line with an unqualified
5336 address (no domain) and a large number of spaces between the end of the
5337 name and the colon was received, the reception process suffered a buffer
5338 overflow, and (when I tested it) crashed. This was caused by some obsolete
5339 code that should have been removed. The fix is to remove it!
5341 3. When running in the test harness, delay a bit after writing a bounce
5342 message to get a bit more predictability in the log output.
5344 4. Added a call to search_tidyup() just before forking a reception process. In
5345 theory, someone could use a lookup in the expansion of smtp_accept_max_
5346 per_host which, without the tidyup, could leave open a database connection.
5348 5. Added the variables $recipient_data and $sender_data which get set from a
5349 lookup success in an ACL "recipients" or "senders" condition, or a router
5350 "senders" option, similar to $domain_data and $local_part_data.
5352 6. Moved the writing of debug_print from before to after the "senders" test
5355 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5356 problems for message scanning, either using a data ACL, or using
5357 local_scan() because the Received: header was not generated till after they
5358 were called (in order to set the time as the time of reception completion).
5359 I have revised the way this works. The header is now generated after the
5360 body is received, but before the ACL or local_scan() are called. After they
5361 are run, the timestamp in the header is updated.
5367 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5368 before starting a queue runner without re-exec. This happened only when
5369 deliver_drop_privilege was set or when the Exim user was set to root. The
5370 effect of the bug was that timeouts during subsequent deliveries caused
5371 crashes instead of being properly handled. The handler is now left at its
5372 default (and expected) setting.
5374 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5375 message, again when deliver_drop_privilege is set or Exim is run as root.
5376 The bug described in (1) was not present in this case, but the tidying up
5377 of the other signals was missing. I have made the two cases consistent.
5379 3. The ignore_target_hosts setting on a manualroute router was being ignored
5380 for hosts that were looked up using the /MX notation.
5382 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5385 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5386 operated on the sender address. After changing the $sender_address to <>
5387 for the sender address verify, Exim was re-instated it as the original
5388 (before rewriting) address, but remembering that it had rewritten it, so it
5389 wasn't rewriting it again. This bug also had the effect of breaking the
5390 sender address verification caching when the sender address was rewritten.
5392 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5393 This has been changed so that if the ip literal address matches
5394 ignore_target_hosts, the router declines.
5396 7. Added expansion conditions match_domain, match_address, and match_local_
5397 part (NOT match_host).
5399 8. The placeholder for the Received: header didn't have a length field set.
5401 9. Added code to Exim itself and to exim_lock to test for a specific race
5402 condition that could lead to file corruption when using MBX delivery. The
5403 issue is with the lockfile that is created in /tmp. If this file is removed
5404 after a process has opened it but before that process has acquired a lock,
5405 there is the potential for a second process to recreate the file and also
5406 acquire a lock. This could lead to two Exim processes writing to the file
5407 at the same time. The added code performs the same test as UW imapd; it
5408 checks after acquiring the lock that its file descriptor still refers to
5409 the same named file.
5411 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5412 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5415 11. Added the smtp_active_hostname option. If used, this will typically be made
5416 to depend on the incoming interface address. Because $interface_address is
5417 not set up until the daemon has forked a reception process, error responses
5418 that can happen earlier (such as "too many connections") no longer contain
5421 12. If an expansion in a condition on a "warn" statement fails because a lookup
5422 defers, the "warn" statement is abandoned, and the next ACL statement is
5423 processed. Previously this caused the whole ACL to be aborted.
5425 13. Added the iplsearch lookup type.
5427 14. Added ident_timeout as a log selector.
5429 15. Added tls_certificate_verified as a log selector.
5431 16. Added a global option tls_require_ciphers (compare the smtp transport
5432 option of the same name). This controls incoming TLS connections.
5434 17. I finally figured out how to make tls_require_ciphers do a similar thing
5435 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5436 before starting the TLS session.
5438 18. Tabs are now shown as \t in -bP output.
5440 19. If the log selector return_path_on_delivery was set, Exim crashed when
5441 bouncing a message because it had too many Received: header lines.
5443 20. If two routers both had headers_remove settings, and the first one included
5444 a superfluous trailing colon, the final name in the first list and the
5445 first name in the second list were incorrectly joined into one item (with a
5446 colon in the middle).
5452 1. Added -C and -D options to the exinext utility, mainly to make it easier
5453 to include in the automated testing, but these could be helpful when
5454 multiple configurations are in use.
5456 2. The exinext utility was not formatting the output nicely when there was
5457 an alternate port involved in the retry record key, nor when there was a
5458 message id as well (for retries that were specific to a specific message
5459 and a specific host). It was also confused by IPv6 addresses, because of
5460 the additional colons they contain. I have fixed the IPv4 problem, and
5461 patched it up to do a reasonable job for IPv6.
5463 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5464 delivery, the log line now contains "pipelined" if PIPELINING was used.
5466 4. An SMTP transport process used to panic and die if the bind() call to set
5467 an explicit outgoing interface failed. This has been changed; it is now
5468 treated in the same way as a connect() failure.
5470 5. A reference to $sender_host_name in the part of a conditional expansion
5471 that was being skipped was still causing a DNS lookup. This no longer
5474 6. The def: expansion condition was not recognizing references to header lines
5475 that used bh_ and bheader_.
5477 7. Added the _cache feature to named lists.
5479 8. The code for checking quota_filecount in the appendfile transport was
5480 allowing one more file than it should have been.
5482 9. For compatibility with Sendmail, the command line option
5490 and sets the incoming protocol and host name (for trusted callers). The
5491 host name and its colon can be omitted when only the protocol is to be set.
5492 Note the Exim already has two private options, -pd and -ps, that refer to
5493 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5494 "s", but I don't think that's a major issue.
5496 10. A number of refactoring changes to the code, none of which should affect
5499 (a) The number of logging options was getting close to filling up the
5500 32-bit word that was used as a bit map. I have split them into two classes:
5501 those that are passed in the argument to log_write(), and those that are
5502 only ever tested independently outside of that function. These are now in
5503 separate 32-bit words, so there is plenty of room for expansion again.
5504 There is no change in the user interface or the logging behaviour.
5506 (b) When building, for example, log lines, the code previously used a
5507 macro that called string_cat() twice, in order to add two strings. This is
5508 not really sufficiently general. Furthermore, there was one instance where
5509 it was actually wrong because one of the argument was used twice, and in
5510 one call a function was used. (As it happened, calling the function twice
5511 did not affect the overall behaviour.) The macro has been replaced by a
5512 function that can join an arbitrary number of extra strings onto a growing
5515 (c) The code for expansion conditions now uses a table and a binary chop
5516 instead of a serial search (which was left over from when there were very
5517 few conditions). Also, it now recognizes conditions like "pam" even when
5518 the relevant support is not compiled in: a suitably worded error message is
5519 given if an attempt is made to use such a condition.
5521 11. Added ${time_interval:xxxxx}.
5523 12. A bug was causing one of the ddress fields not to be passed back correctly
5524 from remote delivery subprocesses. The field in question was not being
5525 subsequently used, so this caused to problems in practice.
5527 13. Added new log selectors queue_time and deliver_time.
5529 14. Might have fixed a bug in maildirsizefile handling that threw up
5530 "unexpected character" debug warnings, and recalculated the data
5531 unnecessarily. In any case, I expanded the warning message to give more
5534 15. Added the message "Restricted characters in address" to the statements in
5535 the default ACL that block characters like @ and % in local parts.
5537 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5538 Three changes have been made:
5540 (a) There was a serious bug; a negative response to MAIL caused the whole
5541 recipient domain to be cached as invalid, thereby blocking all messages
5542 to all local parts at the same domain, from all senders. This bug has
5543 been fixed. The domain is no longer cached after a negative response to
5544 MAIL if the sender used is not empty.
5546 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5549 (c) A new callout option, "use_sender" has been added for people who want
5550 the modified behaviour.
5556 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5559 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5560 indeed breaks things for older releases.
5562 3. Added additional logging to the case where there is a problem reading data
5563 from a filter that is running in a subprocess using a pipe, in order to
5564 try to track down a specific problem.
5566 4. Testing facility fudge: when running in the test harness and attempting
5567 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5568 getting "No route to host". Convert this to a timeout.
5570 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5573 6. Some OS don't have socklen_t but use size_t instead. This affects the
5574 fifth argument of getsockopt() amongst other things. This is now
5575 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5576 can be set for individual OS. I have set it for SunOS5, OSF1, and
5577 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5578 some earlier ones do not.
5580 7. Change 4.30/15 was not doing the test caselessly.
5582 8. The standard form for an IPv6 address literal was being rejected by address
5583 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5584 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5585 this, as well as the form without the "IPv6" on the front (but only when
5586 address literals are enabled, of course).
5588 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5590 10. Exim crashed if a message with an empty sender address specified by -f
5591 encountered a router with an errors_to setting. This could be provoked only
5592 by a command such as
5596 where an empty string was supplied; "<>" did not hit this bug.
5598 11. Installed PCRE release 4.5.
5600 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5601 remained set. It is now erased.
5603 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5604 times from message ids (which are base 36 rather than the normal 62).
5606 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5607 were being counted as actual protocol errors, and logged if the log
5608 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5609 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5610 and DATA following a set of rejected RCPTs do not count as protocol errors.
5611 In other words, Exim assumes they were pipelined, though this may not
5612 actually be the case. Of course, in all cases the client gets an
5613 appropriate error code.
5615 15. If a lookup fails in an ACL condition, a message about the failure may
5616 be available; it is used if testing the ACL cannot continue, because most
5617 such messages specify what the cause of the deferral is. However, some
5618 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5619 that caused an old message to be retained and used if a later statement
5620 caused a defer, replacing the real cause of the deferral.
5622 16. If an IP address had so many PTR records that the DNS lookup buffer
5623 was not large enough to hold them, Exim could crash while trying to process
5624 the truncated data. It now detects and logs this case.
5626 17. Further to 4.21/58, another change has been made: if (and only if) the
5627 first line of a message (the first header line) ends with CRLF, a bare LF
5628 in a subsequent header line has a space inserted after it, so as not to
5629 terminate the header.
5631 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5632 unnecessarily, used atoi() instead of strtol(), and didn't check the
5633 termination when getting file sizes from file names by regex.
5635 19. Completely re-implemented the support for maildirsize files, in the light
5636 of a number of problems with the previous contributed implementation
5637 (4.30/29). In particular:
5639 . If the quota is zero, the maildirsize file is maintained, but no quota is
5642 . If the maildir directory does not exist, it is created before any attempt
5643 to write a maildirsize file.
5645 . The quota value in the file is just a cache; if the quota is changed in
5646 the transport, the new value overrides.
5648 . A regular expression is available for excluding directories from the
5651 20. The autoreply transport checks the characters in options that define the
5652 message's headers; it allows continued headers, but it was checking with
5653 isspace() after an embedded newline instead of explicitly looking for a
5656 21. If all the "regular" hosts to which an address was routed had passed their
5657 expiry times, and had not reached their retry times, the address was
5658 bounced, even if fallback hosts were defined. Now Exim should go on to try
5661 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5662 equivalent code in the SMTP transport. Some hosts send humungous responses
5663 to HELO/EHLO, more than 1024 it seems.
5665 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5666 gives compiler warnings in some environments. I've now done it "properly",
5669 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5670 (because of problems with the built-in one) was declared to return uschar *
5671 instead of char *, causing compiler failure.
5673 25. Fixed a file descriptor leak when processing alias/forward files.
5675 26. Fixed a minor format string issue in dbfn.c.
5677 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5679 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5680 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5683 29. When a sender address is verified, it is cached, to save repeating the test
5684 when there is more than one recipient in a message. However, when the
5685 verification involves a callout, it is possible for different callout
5686 options to be set for different recipients. It is too complicated to keep
5687 track of this in the cache, so now Exim always runs a verification when a
5688 callout is required, relying on the callout cache for the optimization.
5689 The overhead is duplication of the address routing, but this should not be
5692 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5693 to be verified with callout=postmaster, and the main callout worked but the
5694 postmaster check failed, the verification correctly failed. However, if a
5695 subsequent RCPT command asked for sender verification *without* the
5696 postmaster check, incorrect caching caused this verification also to fail,
5699 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5700 it was not caching the DNS options (qualify_single, search_parents) that
5701 were used when the lookup failed. A subsequent lookup with different
5702 options therefore always gave the same answer, though there were cases
5703 where it should not have. (Example: a "domains = !$mx_any" option on a
5704 dnslookup router: the "domains" option is always processed without any
5705 widening, but the router might have qualify_single set.) Now Exim uses the
5706 cached value only when the same options are set.
5708 32. Added John Jetmore's "exipick" utility to the distribution.
5710 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5711 than a timeout (e.g. a certificate is required, and is not provided), an
5712 Exim server now closes the connection immediately. Previously it waited for
5713 the client to close - but if the client is SSL, it seems that they each
5714 wait for each other, leading to a delay before one of them times out.
5716 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5717 maintained 0.8.x compatibility because I don't think many are using it, and
5718 it is clearly obsolete.
5720 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5723 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5724 client certificate was expired. A simple patch fixes this, though I don't
5725 understand the full logic of why the verify callback is called multiple
5728 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5729 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5730 which causes problems with some clients (such as the Certicom SSL Plus
5731 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5732 disables the coutermeasure allowing Eudora to connect."
5734 38. Exim was not checking that a write() to a log file succeeded. This could
5735 lead to Bad Things if a log got too big, in particular if it hit a file
5736 size limit. Exim now panics and dies if it cannot write to a log file, just
5737 as it does if it cannot open a log file.
5739 39. Modified OS/Makefile-Linux so that it now contains
5741 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5743 The two -D definitions ensure that Exim is compiled with large file
5744 support, which makes it possible to handle log files that are bigger than
5747 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5748 instance) a domain was checked against a named list that involved a lookup,
5749 causing $domain_data to be set, then another domain was checked against the
5750 same list, then the first domain was re-checked, the value of $domain_data
5751 after the final check could be wrong. In particular, if the second check
5752 failed, it could be set empty. This bug probably also applied to
5755 41. The strip_trailing_dot option was not being applied to the address given
5756 with the -f command-line option.
5758 42. The code for reading a message's header from the spool was incrementing
5759 $received_count, but never initializing it. This meant that the value was
5760 incorrect (doubled) while delivering a message in the same process in which
5761 it was received. In the most common configuration of Exim, this never
5762 happens - a fresh exec is done - but it can happen when
5763 deliver_drop_privilege is set.
5765 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5766 it now includes up to 150 characters of the unexpected data in the log
5769 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5770 and building data strings. The size of both of these buffers was 10 000
5771 bytes - far larger than anybody would *ever* want, thought I. Needless to
5772 say, somebody hit the limit. I have increased the maximum line length to
5773 20 000 and the maximum data length of concatenated lines to 100 000. I have
5774 also fixed two bugs, because there was no checking on these buffers. Tsk,
5775 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5776 buffer is too small.
5778 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5779 lsearch lookups. Now it does.
5781 46. When parsing a route_list item in a manualroute router, a fixed-length
5782 buffer was used for the list of hosts. I made this 1024 bytes long,
5783 thinking that nobody would ever have a list of hosts that long. Wrong.
5784 Somebody had a whole pile of complicated expansion conditions, and the
5785 string was silently truncated, leading to an expansion error. It turns out
5786 that it is easier to change to an unlimited length (owing to other changes
5787 that have happened since this code was originally written) than to build
5788 structure for giving a limitation error. The length of the item that
5789 expands into the list of hosts is now unlimited.
5791 47. The lsearch lookup could not handle data where the length of text line was
5792 more than 4095 characters. Such lines were truncated, leading to shortened
5793 data being returned. It should now handle lines of any length.
5795 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5796 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5799 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5800 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5801 no longer happens in comment lines. A list of replacements is now placed
5802 at the head of all of the source files, except those whose only change is
5803 to replace PERL_COMMAND in the very first #! line.
5805 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5806 messages on the queue, with a bottom-up merge sort, using code contributed
5807 by Michael Haardt. This should make operations like -bp somewhat faster on
5808 large queues. It won't affect queue runners, except when queue_run_in_order
5811 51. Installed eximstats 1.31 in the distribution.
5813 52. Added support for SRV lookups to the dnslookup router.
5815 53. If an ACL referred to $message_body or $message_body_end, the value was not
5816 reset for any messages that followed in the same SMTP session.
5818 54. The store-handling optimization for building very long strings was not
5819 differentiating between the different store pools. I don't think this
5820 actually made any difference in practice, but I've tidied it.
5822 55. While running the routers to verify a sender address, $sender_address
5823 was still set to the sender address. This is wrong, because when routing to
5824 send a bounce to the sender, it would be empty. Therefore, I have changed
5825 it so that, while verifying a sender address, $sender_address is set to <>.
5826 (There is no change to what happens when verifying a recipient address.)
5828 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5829 target A or AAAA records (if not already returned) without resetting the
5830 qualify_single or search_parents options of the DNS resolver. These are
5831 inappropriate in this case because the targets of MX and SRV records must
5832 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5833 target that, when qualified, matched something in the local domain. These
5834 two options are now turned off when doing these lookups.
5836 57. It seems that at least some releases of Reiserfs (which does not have the
5837 concept of a fixed number of inodes) returns zero and not -1 for the
5838 number of available inodes. This interacted badly with check_spool_inodes,
5839 which assumed that -1 was the "no such thing" setting. What I have done is
5840 to check that the total number of inodes is greater than zero before doing
5841 the test of how many are available.
5843 58. When a "warn" ACL statement has a log_message modifier, the message is
5844 remembered, and not repeated. This is to avoid a lot of repetition when a
5845 message has many recipients that cause the same warning to be written.
5846 However, Exim was preserving the list of already written lines for an
5847 entire SMTP session, which doesn't seem right. The memory is now reset if a
5848 new message is started.
5850 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5851 debugging output unless log_rewrite was also set.
5853 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5854 of (int)(handle) when we know that handle contains (void *)(-1).
5856 61. The Exim daemon panic-logs an error return when it closes the incoming
5857 connection. However "connection reset by peer" seems to be common, and
5858 isn't really an error worthy of noting specially, so that particular error
5861 62. When Exim is trying to find all the local interfaces, it used to panic and
5862 die if the ioctl to get the interface flags failed. However, it seems that
5863 on at least one OS (Solaris 9) it is possible to have an interface that is
5864 included in the list of interfaces, but for which you get a failure error
5865 for this call. This happens when the interface is not "plumbed" into a
5866 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5867 failure of the "get flags" call assumes that the interface is down.
5869 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5870 makes life easier for people who are doing arithmetic on fields extracted
5871 from dates, where you often get leading zeros that should not be
5872 interpreted as octal.
5874 64. Added qualify_domain to the redirect router, to override the global
5877 65. If a pathologically long header line contained very many addresses (the
5878 report of this problem mentioned 10 000) and each of them was rewritten,
5879 Exim could use up a very large amount of memory. (It kept on making new
5880 copies of the header line as it rewrote, and never released the old ones.)
5881 At the expense of a bit more processing, the header rewriting function has
5882 been changed so that it no longer eats memory in this way.
5884 66. The generation of the Received: header has been moved from the time that a
5885 message starts to be received, to the time that it finishes. The timestamp
5886 in the Received: header should now be very close to that of the <= log
5887 line. There are two side-effects of this change:
5889 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5890 logged header lines no longer include the local Received: line, because
5891 it has not yet been created. The same applies to a copy of the message
5892 that is returned to a non-SMTP sender when a message is rejected.
5894 (b) When a filter file is tested using -bf, no additional Received: header
5895 is added to the test message. After some thought, I decided that this
5898 This change does not affect the value of $received_for. It is still set
5899 after address rewriting, but before local_scan() is called.
5901 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5903 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5904 gave an unhelpful panic error message, and a defer error. I have managed to
5905 change this behaviour so that it now rejects any supplied certificate,
5906 which seems right, as the list of acceptable certificates is empty.
5908 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5909 gave an unhelpful defer error. I have not managed to make this reject any
5910 supplied certificates, but the error message it gives is "no certificate
5911 supplied", which is not helpful.
5913 70. exigrep's output now also includes lines that are not associated with any
5914 message, but which match the given pattern. Implemented by a patch from
5915 Martin Sluka, which also tidied up the Perl a bit.
5917 71. Recipient callout verification, like sender verification, was using <> in
5918 the MAIL FROM command. This isn't really the right thing, since the actual
5919 sender may affect whether the remote host accepts the recipient or not. I
5920 have changed it to use the actual sender in the callout; this means that
5921 the cache record is now keyed on a recipient/sender pair, not just the
5922 recipient address. There doesn't seem to be a real danger of callout loops,
5923 since a callout by the remote host to check the sender would use <>.
5924 [SEE ABOVE: changed after hitting problems.]
5926 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5927 temporary errors. However, in the case of such a code being given after
5928 the end of a data transmission (i.e. after ".") Exim was failing to write
5929 a retry record for the message. (Yes, there was some broken host that was
5930 actually sending 8xx at this point.)
5932 73. An unknown lookup type in a host list could cause Exim to panic-die when
5933 the list was checked. (An example that provoked this was putting <; in the
5934 middle of a list instead of at the start.) If this happened during a DATA
5935 ACL check, a -D file could be left lying around. This kind of configuration
5936 error no longer causes Exim to die; instead it causes a defer error. The
5937 incident is still logged to the main and panic logs.
5939 74. Buglet left over from Exim 3 conversion. The message "too many messages
5940 in one connection" was written to the rejectlog but not the mainlog, except
5941 when address rewriting (yes!) was being logged.
5943 75. Added write_rejectlog option.
5945 76. When a system filter was run not as root (that is, when system_filter_user
5946 was set), the values of the $n variables were not being returned to the
5947 main process; thus, they were not subsequently available in the $sn
5950 77. Added +return_path_on_delivery log selector.
5952 78. A connection timeout was being treated differently from recipients deferred
5953 when testing hosts_max_try with a message that was older than the host's
5954 retry timeout. (The host should not be counted, thus allowing all hosts to
5955 be tried at least once before bouncing.) This may have been the cause of an
5956 occasionally reported bug whereby a message would remain on the queue
5957 longer than the retry timeout, but would be bounced if a delivery was
5958 forced. I say "may" because I never totally pinned down the problem;
5959 setting up timeout/retry tests is difficult. See also the next item.
5961 79. The ultimate address timeout was not being applied to errors that involved
5962 a combination of host plus message (for example, a timeout on a MAIL
5963 command). When an address resolved to a number of possible hosts, and they
5964 were not all tried for each delivery (e.g. because of hosts_max_try), a
5965 message could remain on the queue longer than the retry timeout.
5967 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5970 81. Fixed an obscure SMTP outgoing bug which required at least the following
5971 conditions: (a) there was another message waiting for the same server;
5972 (b) the server returned 5xx to all RCPT commands in the first message so
5973 that the message was not completed; (c) the server dropped the connection
5974 or gave a negative response to the RSET that Exim sends to abort the
5975 transaction. The observed case was a dropped connection after DATA that had
5976 been sent in pipelining mode. That is, the server had advertised PIPELINING
5977 but was not implementing it correctly. The effect of the bug was incorrect
5978 behaviour, such as trying another host, and this could lead to a crash.
5984 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5985 and daemon.c were passed as pointers to ints; they should have been
5986 pointers to socklen_t variables (which are typically unsigned ints).
5988 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5991 3. Fixed a really odd bug that affected only the testing scheme; patching a
5992 certain fixed string in the binary changed the value of another string that
5993 happened to be identical to the end of the original first string.
5995 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5996 name", it returns that address as the IP address. On some operating
5997 systems (e.g. Solaris), it also passes back the IP address string as the
5998 "host name". However, on others (e.g. Linux), it passes back an empty
5999 string. Exim wasn't checking for this, and was changing the host name to an
6000 empty string, assuming it had been canonicalized.
6002 5. Although rare, it is permitted to have more than one PTR record for a given
6003 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6004 all the names associated with an address, because they do in Solaris.
6005 However, it seems that they do not in Linux for data that comes from the
6006 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6007 I found this out when I moved to a new Linux workstation and tried to run
6008 the Exim test suite.
6010 To get round this problem I have changed the code so that it now does its
6011 own call to the DNS to look up PTR records when searching for a host name.
6012 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6013 addresses that are only in /etc/hosts are still found.
6015 This behaviour is, however, controlled by an option called host_lookup_
6016 order, which defaults to "bydns:byaddr". If people want to use the other
6017 order, or indeed, just use one or the other means of lookup, they can
6018 specify it in this variable.
6020 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6021 some operating systems, this comes back from gethostbyaddr() as an empty
6022 string, and this is what Exim used to test for. However, it seems that in
6023 other systems, "." is yielded. Exim now tests for this case too.
6025 7. The values of check_spool_space and check_log_space are now held internally
6026 as a number of kilobytes instead of an absolute number of bytes. If a
6027 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6028 kilobyte. This means that much larger values can be stored.
6030 8. Exim monitor: an attempt to get the action menu when not actually pointing
6031 at a message produces an empty menu entitled "No message selected". This
6032 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6033 no entries in it ("Shell widget menu has zero width and/or height"). So I
6034 have added a single, blank menu entry in this case.
6036 9. Added ${quote_local_part.
6038 10. MIME decoding is now applied to the contents of Subject: header lines when
6041 11. Now that a reference to $sender_host_address automatically causes a reverse
6042 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6043 host lookup before query-style lookups in lists that might use this
6044 variable. This has therefore been abolished, and the "net-" prefix is no
6045 longer necessary for query-style lookups.
6047 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6048 have been a typo for LFLAGS, so it has been changed.
6050 13. The install script calls Exim with "-C /dev/null" in order to find the
6051 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6052 to be output. However, since Exim outputs its version number before the
6053 error, it didn't break the script. It just looked ugly. I fixed this by
6054 always allowing "-C /dev/null" if the caller is root.
6056 14. Ignore overlarge ACL variable number when reading spool file - insurance
6057 against a later release with more variables having written the file.
6059 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6060 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6061 this, as well as the form without the "IPv6" on the front.
6063 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6064 OS/Makefile-Darwin file.
6066 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6067 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6069 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6070 variable. However, some OS use ints for inodes. I've added cast to long int
6071 to get rid of the compiler warning.
6073 19. I had forgotten to lock out "/../" in configuration file names when
6074 ALT_CONFIG_PREFIX was set.
6076 20. Routers used for verification do not need to specify transports. However,
6077 if such a router generated a host list, and callout was configured, Exim
6078 crashed, because it could not find a port number from the (non-existent)
6079 transport. It now assumes port 25 in this circumstance.
6081 21. Added the -t option to exigrep.
6083 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6084 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6085 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6086 them is set, LOOKUP_LSEARCH is forced.
6088 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6090 24. Added sender and host information to the "rejected by local_scan()" log
6091 line; previously there was no indication of these.
6093 25. Added .include_if_exists.
6095 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6096 Linux. It turns out that not all file systems support this. Apparently some
6097 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6098 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6099 not supported on the file descriptor, is now ignored when Exim is trying to
6100 sync a directory. This applies only to Linux.
6102 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6104 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6105 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6108 29. Added support for maildirsize files from supplied patch (modified a bit).
6110 30. The use of :fail: followed by an empty string could lead Exim to respond to
6111 sender verification failures with (e.g.):
6113 550 Verification failed for <xxx>
6114 550 Sender verify failed
6116 where the first response line was missing the '-' that indicates it is not
6117 the final line of the response.
6119 31. The loop for finding the name of the user that called Exim had a hardwired
6120 limit of 10; it now uses the value of finduser_retries, which is used for
6121 all other user lookups.
6123 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6126 33. Exim was neglecting to zero errno before one call of strtol() when
6127 expanding a string and expecting an integer value. On some systems this
6128 resulted in spurious "integer overflow" errors. Also, it was casting the
6129 result into an int without checking.
6131 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6132 did not work. The code looks as if it has *never* worked, though it appears
6133 to have been documented since at least release 1.62. I have made it work.
6135 35. The "timeout_DNS" error in retry rules, also documented since at least
6136 1.62, also never worked. As it isn't clear exactly what this means, and
6137 clearly it isn't a major issue, I have abolished the feature by treating it
6138 as "timeout", and writing a warning to the main and panic logs.
6140 36. The display of retry rules for -brt wasn't always showing the error code
6143 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6144 timeout_connect_A, timeout_connect_MX.
6146 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6147 to the empty sender.
6149 39. The daemon was not analysing the content of -oX till after it had closed
6150 stderr and disconnected from the controlling terminal. This meant that any
6151 syntax errors were only noted on the panic log, and the return code from
6152 the command was 0. By re-arranging the code a little, I've made the
6153 decoding happen first, so such errors now appear on stderr, and the return
6154 code is 1. However, the actual setting up of the sockets still happens in
6155 the disconnected process, so errors there are still only recorded on the
6158 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6159 connections (as happens on some IP stacks) was logged at start up time as
6160 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6161 it from "IPv6 and IPv4", which means that two separate sockets are being
6164 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6165 says whether AF_INET or AF_INET6 was passed as an argument.
6167 42. Exiwhat output was messed up when time zones were included in log
6170 43. Exiwhat now gives more information about the daemon's listening ports,
6171 and whether -tls-on-connect was used.
6173 44. The "port" option of the smtp transport is now expanded.
6175 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6176 silently ignored. Now an error message is written to the main and panic
6179 46. There's a new ACL modifier called "logwrite" which writes to a log file
6180 as soon as it is encountered.
6182 47. Added $local_user_uid and $local_user_gid at routing time.
6184 48. Exim crashed when trying to verify a sender address that was being
6187 49. Exim was recognizing only a space character after ".include". It now also
6188 recognizes a tab character.
6190 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6191 extracting the relevant information from the specification. The man page no
6192 longer contains scrambled data for the -d option, and I've added a section
6193 at the front about calling Exim under different names.
6195 51. Added "extra_headers" argument to the "mail" command in filter files.
6197 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6200 53. Installed eximstats 1.29.
6202 54. Added transport_filter_timeout as a generic transport option.
6204 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6205 Cc: header lines. This was required by RFC 822, but it not required by RFC
6208 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6209 incoming messages that did not have them. Now it does so only if the
6210 message originates locally, that is, if there is no associated remote host
6211 address. When Resent- header lines are present, this applies to the Resent-
6212 lines rather than the non-Resent- lines.
6214 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6215 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6217 58. Messages for configuration errors now include the name of the main
6218 configuration file - useful now that there may be more than one file in a
6219 list (.included file names were always shown).
6221 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6222 for those rare installations that do not start the daemon as root or run it
6223 setuid root. I've cut out the call to initgroups() if the daemon is not
6226 60. The Exim user and group can now be bound into the binary as text strings
6227 that are looked up at the start of Exim's processing.
6229 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6231 62. Added $mailstore_basename variable.
6233 63. Installed patch to sieve.c from Michael Haardt.
6235 64. When Exim failed to open the panic log after failing to open the main log,
6236 the original message it was trying to log was written to stderr and debug
6237 output, but if they were not available (the usual case in production), it
6238 was lost. Now it is written to syslog before the two lines that record the
6239 failures to open the logs.
6241 65. Users' Exim filters run in subprocesses under the user's uid. It is
6242 possible for a "deliver" command or an alias in a "personal" command to
6243 provoke an address rewrite. If logging of address rewriting is configured,
6244 this fails because the process is not running as root or exim. There may be
6245 a better way of dealing with this, but for the moment (because 4.30 needs
6246 to be released), I have disabled address rewrite logging when running a
6247 filter in a non-root, non-exim process.
6253 1. The buildconfig auxiliary program wasn't quoting the value set for
6254 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6255 not defined. This bug was masked in 4.22 by the effect that was fixed in
6258 2. Some messages that were rejected after a message id was allocated were
6259 shown as "incomplete" by exigrep. It no longer does this for messages that
6260 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6262 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6263 have allow_domain_literals set, the ID did not get logged in the <= line.
6264 Domain literals are now always recognized in Message-ID: header lines.
6266 4. The first argument for a ${extract expansion item is the key name or field
6267 number. Leading and trailing spaces in this item were not being ignored,
6268 causing some misleading effects.
6270 5. When deliver_drop_privilege was set, single queue runner processes started
6271 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6272 same command in the process it spins off) were not dropping privilege.
6274 6. When the daemon running as "exim" started a queue runner, it always
6275 re-executed Exim in the spun-off process. This is a waste of effort when
6276 deliver_drop_privilege is set. The new process now just calls the
6277 queue-runner function directly.
6283 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6286 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6287 ignored. Though the use of -oP was forcing the writing of a pid file, it
6288 was always written to the default place.
6290 3. If the message "no IP address found for host xxxx" is generated during
6291 incoming verification, it is now followed by identification of the incoming
6292 connection (so you can more easily find what provoked it).
6294 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6296 5. Added some features to "harden" Exim a bit more against certain attacks:
6298 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6299 be put in Local/Makefile. This is like the never_users runtime option,
6300 but it cannot be overridden. The default setting is "root".
6302 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6303 prefix string with which any file named in a -C command line option
6306 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6307 is retained for -C and -D only if the caller of Exim is root. Without
6308 it, the exim user may also use -C and -D and retain privilege.
6310 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6311 command line option is disabled.
6313 6. Macro names set by the -D option must start with an upper case letter, just
6314 like macro names defined in the configuration file.
6316 7. Added "dereference=" facility to LDAP.
6318 8. Two instances of the typo "uknown" in the source files are fixed.
6320 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6321 the Configure-Makefile script screwed up while processing it.
6323 10. Incorporated PCRE 4.4.
6325 11. The SMTP synchronization check was not operating right at the start of an
6326 SMTP session. For example, it could not catch a HELO sent before the client
6327 waited for the greeting. There is now a check for outstanding input at the
6328 point when the greeting is written. Because of the duplex, asynchronous
6329 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6330 way, but not yet received, when the check is performed.
6332 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6333 on TCP/IP sockets, because this apparently causes some broken clients to
6336 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6337 unchanged) from the Cygwin maintainer.
6339 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6340 was supported instead of testing for mbx. Effectively a typo.
6342 15. The spa authenticator server code was not checking that the input it
6343 received was valid base64.
6345 16. The debug output line for the "set" modifier in ACLs was not showing the
6346 name of the variable that was being set.
6348 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6350 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6351 lookup on its own. Something else was needed to trigger the lookup. For
6352 example, a match in host_lookup or the need for a host name in a host list.
6353 Now, if $sender_host_name is referenced and the host name has not yet been
6354 looked up, a lookup is performed. If the lookup fails, the variable remains
6355 empty, and $host_lookup_failed is set to "1".
6357 19. Added "eqi" as a case-independent comparison operator.
6359 20. The saslauthd authentication condition could segfault if neither service
6360 nor realm was specified.
6362 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6363 error message that was logged was misleading, and incoming SMTP
6364 connections were dropped. The message is now more accurate, and temporary
6365 errors are given to SMTP connections.
6367 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6368 (except RSET) with a 421 error, until QUIT is received. However, it was
6369 failing to send a response to QUIT.
6371 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6372 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6373 ACL is now run after the helo_try_verify_hosts code.
6375 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6376 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6377 case-independent, so other case variants are also recognized). Apparently
6378 some systems use these upper case variants.
6380 25. If more than two messages were waiting for the same host, and a transport
6381 filter was specified for the transport, Exim sent two messages over the
6382 same TCP/IP connection, and then failed with "socket operation on non-
6383 socket" when it tried to send the third.
6385 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6387 27. The extern definition of crypt16() in expand.c was not being excluded when
6388 the OS had its own crypt16() function.
6390 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6391 as a preferred synonym for return_size_limit, both as an option and as an
6394 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6396 30. Changed the default configuration ACL to relax the local part checking rule
6397 for addresses that are not in any local domains. For these addresses,
6398 slashes and pipe symbols are allowed within local parts, but the sequence
6399 /../ is explicitly forbidden.
6401 31. SPA server authentication was not clearing the challenge buffer before
6404 32. log_message in a "warn" ACL statement was writing to the reject log as
6405 well as to the main log, which contradicts the documentation and doesn't
6406 seem right (because no rejection is happening). So I have stopped it.
6408 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6409 However, I am unable to do any testing of this.
6411 34. Fixed an infelicity in the appendfile transport. When checking directories
6412 for a mailbox, to see if any needed to be created, it was accidentally
6413 using path names with one or more superfluous leading slashes; tracing
6414 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6416 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6417 contents are added to the log line that is written for every discarded
6418 recipient. (Previously a log_message setting was ignored.)
6420 36. The ${quote: operator now quotes the string if it is empty.
6422 37. The install script runs exim in order to find its version number. If for
6423 some reason other than non-existence or emptiness, which it checks, it
6424 could not run './exim', it was installing it with an empty version number,
6425 i.e. as "exim-". This error state is now caught, and the installation is
6428 38. An argument was missing from the function that creates an error message
6429 when Exim fails to connect to the socket for saslauthd authentication.
6430 This could cause Exim to crash, or give a corrupted message.
6432 39. Added isip, isip4, and isip6 to ${if conditions.
6434 40. The ACL variables $acl_xx are now saved with the message, and can be
6435 accessed later in routers, transports, and filters.
6437 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6438 strings in the file are not string-expanded.
6440 42. If a MAIL command specified a SIZE value that was too large to fit into an
6441 int variable, the check against message_size_limit failed. Such values are
6442 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6443 day this will have to be increased, but I don't think I want to be around
6444 when emails are that large.
6451 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6452 iconv() is not standard in FreeBSD.
6454 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6455 IPv6 enabled. The observed symptom was a segmentation fault on return from
6456 the function os_common_find_running_interfaces() in src/os.c.
6458 3. In the check_special_case() function in daemon.c I had used "errno" as an
6459 argument name, which causes warnings on some systems. This was basically a
6460 typo, since it was named "eno" in the comments!
6462 4. The code that waits for the clock to tick (at a resolution of some fraction
6463 of a second) so as to ensure message-id uniqueness was always waiting for
6464 at least one whole tick, when it could have waited for less. [This is
6465 almost certainly not relevant at current processor speeds, where it is
6466 unlikely to ever wait at all. But we try to future-proof.]
6468 5. The function that sleeps for a time interval that includes fractions of a
6469 second contained a race. It did not block SIGALRM between setting the
6470 timer, and suspending (a couple of lines later). If the interval was short
6471 and the sigsuspend() was delayed until after it had expired, the suspension
6472 never ended. On busy systems this could lead to processes getting stuck for
6475 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6476 process, before it forks any delivery processes. The open lookup caching
6477 mechanism meant that the open file or database connection was passed into
6478 the delivery process. The problem was that delivery processes always tidy
6479 up cached lookup data. This could cause a problem for the next delivery
6480 process started by the queue runner, because the external queue runner
6481 process does not know about the closure. So the next delivery process
6482 still has data in the lookup cache. In the case of a file lookup, there was
6483 no problem because closing a file descriptor in a subprocess doesn't affect
6484 the parent. However, if the lookup was caching a connection to a database,
6485 the connection was closed, and the second delivery process was likely to
6486 see errors such as "PGSQL: query failed: server closed the connection
6487 unexpectedly". The problem has been fixed by closing all cached lookups
6488 in a queue runner before running a delivery process.
6490 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6491 seem to have the "const" qualifier which it has on other OS. I've
6494 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6495 *of the same type* (client or server) with the same public name that an
6496 error should be diagnosed.
6498 9. When Exim looked up a host name for an IP address, but failed to find the
6499 original IP address when looking up the host name (a safety check), it
6500 output the message "<ip address> does not match any IP for NULL", which was
6501 confusing, to say the least. The bug was that the host name should have
6502 appeared instead of "NULL".
6504 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6505 user that is built into the binary, and the -C or -D options is used, root
6506 privilege is dropped before the configuration file is read. In addition,
6507 logging is switched to stderr instead of the normal log files. If the
6508 configuration then re-defines the Exim user, the unprivileged environment
6509 is probably not what is expected, so Exim logs a panic warning message (but
6512 However, if deliver_drop_privilege is set, the unprivileged state may well
6513 be exactly what is intended, so the warning has been cut out in that case,
6514 and Exim is allowed to try to write to its normal log files.
6520 1. smtp_return_error_details was not giving details for temporary sender
6521 or receiver verification errors.
6523 2. Diagnose a configuration error if two authenticators have the same public
6526 3. Exim used not to create the message log file for a message until the first
6527 delivery attempt. This could be confusing when incoming messages were held
6528 for policy or load reasons. The message log file is now created at the time
6529 the message is received, and an initial "Received" line is written to it.
6531 4. The automatically generated man page for command line options had a minor
6532 bug that caused no ill effects; however, a more serious problem was that
6533 the procedure for building the man page automatically didn't always
6534 operate. Consequently, release 4.20 contains an out-of-date version. This
6535 shouldn't happen again.
6537 5. When building Exim with embedded Perl support, the script that builds the
6538 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6539 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6541 6. The freeze_tell option was not being used for messages that were frozen on
6542 arrival, either by an ACL or by local_scan().
6544 7. Added the smtp_incomplete_transaction log selector.
6546 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6547 was accepting AUTH without a new EHLO.
6549 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6550 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6553 10. Logging of TCP/IP connections (when configured) now happens in the main
6554 daemon process instead of the child process, so that the TCP/IP connection
6555 count is more accurate (but it can never be perfect).
6557 11. The use of "drop" in a nested ACL was not being handled correctly in the
6558 outer ACL. Now, if condition failure induced by the nested "drop" causes
6559 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6560 or "require"), the connection is dropped.
6562 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6563 that yield "discard" can now be used with an "accept" or a "discard" verb,
6564 but an error is generated for any others (because I can't see a useful way
6565 to define what should happen).
6567 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6568 are now processed in the same way as lines in the Exim configuration file.
6569 In particular, continuation lines are supported.
6571 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6573 15. Added -ti meaning -t -i.
6575 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6576 domains, and warn by logging if others are found.
6578 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6579 returned from GIFCONF when Exim is trying to find the list of interfaces on
6580 a host. The code in os.c has been modified to copy each ifreq to an aligned
6581 structure in all cases.
6583 Also, in some cases, the returned ifreq's were being copied to a 'struct
6584 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6585 means the last couple of bytes of an IPv6 address could be chopped if the
6586 ifreq contained only a normal sockaddr (14 bytes storage).
6588 18. Named domain lists were not supported in the hosts_treat_as_local option.
6589 An entry such as +xxxx was not recognized, and was treated as a literal
6592 19. Ensure that header lines added by a DATA ACL are included in the reject log
6593 if the ACL subsequently rejects the message.
6595 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6596 MD5 (which is deprecated).
6598 21. When testing a filter file using -bf, Exim was writing a message when it
6599 took the sender from a "From " line in the message, but it was not doing so
6600 when it took $return_path from a Return-Path: header line. It now does.
6602 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6603 with a valid header line field name (a series of printing characters
6604 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6606 23. Changed "disc" in the source to "disk" to conform to the documentation and
6607 the book and for uniformity.
6609 24. Ignore Sendmail's -Ooption=value command line item.
6611 25. When execve() failed while trying to run a command in a pipe transport,
6612 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6613 could be confused with a return value of 69 from the command itself. This
6614 has been changed to 127, the value the shell returns if it is asked to run
6615 a non-existent command. The wording for the related log line suggests a
6616 non-existent command as the problem.
6618 26. If received_header_text expands to an empty string, do not add a Received:
6619 header line to the message. (Well, it adds a token one on the spool, but
6620 marks it "old" so that it doesn't get used or transmitted.)
6622 27. Installed eximstats 1.28 (addition of -nt option).
6624 28. There was no check for failure on the call to getsockname() in the daemon
6625 code. This can fail if there is a shortage of resources on the system, with
6626 ENOMEM, for example. A temporary error is now given on failure.
6628 29. Contrary to the C standard, it seems that in some environments, the
6629 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6630 program. Exim now does this explicitly; it affects the formatting of
6631 timestamps using strftime().
6633 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6634 complaints. I've now initialized all the variables, to avoid this.
6636 32. Header lines added by a system filter were not being "seen" during
6637 transport-time rewrites.
6639 33. The info_callback() function passed to OpenSSL is set up with type void
6640 (*)(SSL *, int, int), as described somewhere. However, when calling the
6641 function (actually a macro) that sets it up, the type void(*)() is
6642 expected. I've put in a cast to prevent warnings from picky compilers.
6644 34. If a DNS black list lookup found a CNAME record, but there were no A
6645 records associated with the domain it pointed at, Exim crashed.
6647 35. If a DNS black list lookup returned more than one A record, Exim ignored
6648 all but the first. It now scans all returned addresses if a particular IP
6649 value is being sought. In this situation, the contents of the
6650 $dnslist_value variable are a list of all the addresses, separated by a
6653 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6654 to accept a host name and all its aliases if the forward lookup for any of
6655 them yielded the IP address of the incoming connection. Now it accepts only
6656 those names whose forward lookup yields the correct IP address. Any other
6657 names are discarded. This closes a loophole whereby a rogue DNS
6658 administrator could create reverse DNS records to break through a
6659 wildcarded host restriction in an ACL.
6661 37. If a user filter or a system filter that ran in a subprocess used any of
6662 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6663 the wrong values were passed to the pipe command ($thisaddress had the
6664 value of $0, $0 had the value of $1, etc). This bug was introduced by
6665 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6667 38. Improved the line breaking for long SMTP error messages from ACLs.
6668 Previously, if there was no break point between 40 and 75 characters, Exim
6669 left the rest of the message alone. Two changes have been made: (a) I've
6670 reduced the minimum length to 35 characters; (b) if it can't find a break
6671 point between 35 and 75 characters, it looks ahead and uses the first one
6672 that it finds. This may give the occasional overlong line, but at least the
6673 remaining text gets split now.
6675 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6676 file descriptors might be low, and that setting 1000 would always raise it.
6677 It turns out that in some environments, the limit is already over 1000 and
6678 that lowering it causes trouble. So now Exim takes care not to decrease it.
6680 40. When delivering a message, the value of $return_path is set to $sender_
6681 address at the start of routing (routers may change the value). By an
6682 oversight, this default was not being set up when an address was tested by
6683 -bt or -bv, which affected the outcome if any router or filter referred to
6686 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6687 log only when "message" or "log_message" are set. However, if one of the
6688 conditions was an address verification, or a call to a nested ACL, the
6689 messages generated by the underlying test were being passed through. This
6690 no longer happens. The underlying message is available in $acl_verify_
6691 message for both "message" and "log_message" expansions, so it can be
6692 passed through if needed.
6694 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6695 new expansion $bh_ to give the encoded byte string without charset
6696 translation. Translation happens only if iconv() is available; HAVE_ICONV
6697 indicates this at build time. HEADERS_CHARSET gives the charset to
6698 translate to; headers_charset can change it in the configuration, and
6699 "headers charset" can change it in an individual filter file.
6701 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6702 that creates RFC 2047 encoded "words" labels them as that charset instead
6703 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6704 expansion operator; (ii) when Exim creates a From: line for a local
6705 message; (iii) when a header line is rewritten to include a "phrase" part.
6707 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6708 buggy, causing it to skip the first lines of messages whose message ID
6709 ended in 'D'. This would not have bitten before Exim release 4.14, because
6710 message IDs were unlikely to end in 'D' before then. The effect was to have
6711 incorrect size information for certain domains.
6713 45. #include "config.h" was missing at the start of the crypt16.c module. This
6714 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6717 46. If there was a timeout during a "random" callout check, Exim treated it as
6718 a failure of the random address, and carried on sending RSET and the real
6719 address. If the delay was just some slowness somewhere, the response to the
6720 original RCPT would be taken as a response to RSET and so on, causing
6721 mayhem of various kinds.
6723 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6724 when I implemented it. It didn't allow for the fact that some option values
6725 may legitimately be negative (e.g. size_addition), and it didn't even do
6726 the right test for positive values.
6728 48. Domain names in DNS records are case-independent. Exim always looks them up
6729 in lower case. Some resolvers return domain names in exactly the case they
6730 appear in the zone file, that is, they may contain uppercase letters. Not
6731 all resolvers do this - some return always lower case. Exim was treating a
6732 change of case by a resolver as a change of domain, similar to a widening
6733 of a domain abbreviation. This triggered its re-routing code and so it was
6734 trying to route what was effectively the same domain again. This normally
6735 caused routing to fail (because the router wouldn't handle the domain
6736 twice). Now Exim checks for this case specially, and just changes the
6737 casing of the domain that it ultimately uses when it transmits the message
6740 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6743 50. If a filter generated a file delivery with a non-absolute name (possible if
6744 no home directory exists for the router), the forbid_file option was not
6747 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6748 the existing equality matching.
6750 52. Exim was using ints instead of ino_t variables in some places where it was
6751 dealing with inode numbers.
6753 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6754 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6755 called TMPDIR, and if it finds it is different, it changes its value.
6757 54. The smtp_printf() function is now made available to local_scan() so
6758 additional output lines can be written before returning. There is also an
6759 smtp_fflush() function to enable the detection of a dropped connection.
6760 The variables smtp_input and smtp_batched_input are exported to
6763 55. Changed the default runtime configuration: the message "Unknown user"
6764 has been removed from the ACL, and instead placed on the localuser router,
6765 using the cannot_route_message feature. This means that any verification
6766 failures that generate their own messages won't get overridden. Similarly,
6767 the "Unrouteable address" message that was in the ACL for unverifiable
6768 relay addresses has also been removed.
6770 56. Added hosts_avoid_esmtp to the smtp transport.
6772 57. The exicyclog script was not checking for the esoteric option
6773 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6774 will work only if exicyclog is run under the appropriate euid.
6776 58. Following a discussion on the list, the rules by which Exim recognises line
6777 endings on incoming messages have been changed. The -dropcr and drop_cr
6778 options are now no-ops, retained only for backwards compatibility. The
6779 following line terminators are recognized: LF CRLF CR. However, special
6780 processing applies to CR:
6782 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6783 nor a local message in the state where . is a terminator.
6785 (ii) If a bare CR is encountered in a header line, an extra space is added
6786 after the line terminator so as not to end the header. The reasoning
6787 behind this is that bare CRs in header lines are most likely either
6788 to be mistakes, or people trying to play silly games.
6790 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6791 was being incorrectly given as 18 bytes larger than it should have been.
6792 This is a VOB (very old bug).
6794 60. This may never have affected anything current, but just in case it has:
6795 When the local host is found other than at the start of a list of hosts,
6796 the local host, those with the same MX, and any that follow, are discarded.
6797 When the list in question was part of a longer list of hosts, the following
6798 hosts (not currently being processed) were also being discarded. This no
6799 longer happens. I'm not sure if this situation could ever has previously
6802 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6805 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6806 header. This is something that is recommended in a new Internet Draft, and
6807 is something that is documented as being done by Sendmail. There are two
6808 possible values. For messages generated by the autoreply transport, Exim
6811 Auto-Submitted: auto-replied
6813 whereas for all other generated messages (e.g. bounces) it adds
6815 Auto-Submitted: auto-generated
6817 63. The "personal" condition in filters now includes a test for the
6818 Auto-Submitted: header. If it contains the string "auto-" the message it
6819 not considered personal.
6821 64. Added rcpt_include_affixes as a generic transport option.
6823 65. Added queue_only_override (default true).
6825 66. Added the syslog_duplication option.
6827 67. If what should have been the first header line of a message consisted of
6828 a space followed by a colon, Exim was mis-interpreting it as a header line.
6829 It isn't of course - it is syntactically invalid and should therefore be
6830 treated as the start of the message body. The misbehaviour could have
6831 caused a number of strange effects, including loss of data in subsequent
6832 header lines, and spool format errors.
6834 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6835 client host had authenticated. This control can now be exercised by an ACL
6836 for more flexibility.
6838 69. By default, callouts do not happen when testing with -bh. There is now a
6839 variant, -bhc, which does actually run the callout code, including
6840 consulting and updating the callout cache.
6842 70. Added support for saslauthd authentication, courtesy of Alexander
6845 71. If statvfs() failed on the spool or log directories while checking their
6846 size for availability, Exim confusingly gave the error "space shortage".
6847 Furthermore, in debugging mode it crashed with a floating point exception.
6848 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6849 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6850 serious problem, Exim now writes to the main and panic logs when this
6851 happens, with details of the failure. It then refuses to accept the
6852 incoming message, giving the message "spool directory problem" or "log
6853 directory problem" with a 421 code for SMTP messages.
6855 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6856 0, 1, and 2 exist, because some OS complain for execs without them (see
6857 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6858 descriptors. However, the code omitted to check that the open succeeded,
6859 causing mysterious errors if for some reason the permissions on /dev/null
6860 got screwed. Now Exim writes a message to the main and panic logs, and
6861 bombs out if it can't open /dev/null.
6863 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6864 interact so that it is all more flexible. It is supposed to remain
6865 backwards compatible. Also added extra_local_interfaces.
6867 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6868 to bomb out with an assertion failure - to the client this appears as a
6869 connection drop. This problem occurs in the part of the code that was taken
6870 from the Samba project. Fortunately, the assertion is in a very simple
6871 function, so I have fixed this by reproducing the function inline in the
6872 one place where it is called, and arranging for authentication to fail
6873 instead of killing the process with assert().
6875 75. The SPA client code was not working when the server requested OEM rather
6876 than Unicode encoding.
6878 76. Added code to make require_files with a specific uid setting more usable in
6879 the case where statting the file as root fails - usually a non-root-mounted
6880 NFS file system. When this happens and the failure is EACCES, Exim now
6881 forks a subprocess and does the per-uid checking as the relevant uid.
6883 77. Added process_log_path.
6885 78. If log_file_path was not explicitly set, a setting of check_log_space or
6886 check_log_inodes was ignored.
6888 79. If a space check for the spool or log partitions fails, the incident is now
6889 logged. Of course, in the latter case the data may get lost...
6891 80. Added the %p formatting code to string_format() so that it can be used to
6892 print addresses in debug_print(). Adjusted all the address printing in the
6893 debugging in store.c to use %p rather than %d.
6895 81. There was a concern that a line of code in smtp_in.c could overflow a
6896 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6897 initially expressed, the concern was not well-founded, because trailing
6898 spaces are removed early. However, if the trailing spaces were followed by
6899 a NULL, they did not get removed, so the overflow was possible. Two fixes
6902 (a) I re-wrote the offending code in a cleaner fashion.
6903 (b) If an incoming SMTP command contains a NULL character, it is rejected
6906 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6907 runs initgroups(), so that if the Exim user is in any additional groups,
6908 they will be used during message reception.
6914 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob